Chiebot-Mirror
/
grpc
mirror of https://github.com/grpc/grpc.git
8
0
Fork 0
Code Issues Projects Releases Wiki Activity
The C based gRPC (C++, Python, Ruby, Objective-C, PHP, C#) https://grpc.io/
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
72 Commits
274 Branches
377 Tags
1.2 GiB
Tag: Branch: Tree: c5abe59fe1
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from 'c5abe59fe1'
${ noResults }
grpc/Makefile

8144 lines
422 KiB
Raw Normal View History Unescape

Initial import.
10 years ago
# GRPC global makefile
# This currently builds C and C++ code.
# General settings.
# You may want to change these depending on your system.
prefix ?= /usr/local
PROTOC = protoc
CC = gcc
CXX = g++
LD = gcc
LDXX = g++
AR = ar
STRIP = strip --strip-unneeded
INSTALL = install -D
RM = rm -f
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
HOST_CC = $(CC)
HOST_CXX = $(CXX)
HOST_LD = $(LD)
HOST_LDXX = $(LDXX)
Initial import.
10 years ago
ifeq ($(DEBUG),)
CPPFLAGS += -O2
DEFINES += NDEBUG
else
CPPFLAGS += -O0
DEFINES += _DEBUG DEBUG
endif
CFLAGS += -std=c89 -pedantic
CXXFLAGS += -std=c++11
CPPFLAGS += -g -fPIC -Wall -Werror -Wno-long-long
LDFLAGS += -g -pthread -fPIC
INCLUDES = . include gens
LIBS = rt m z event event_pthreads pthread
LIBSXX = protobuf
LIBS_SECURE = ssl crypto dl
Fixing C++ and ruby code generators for the opensource build. Change on 2014/12/11 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81921764
10 years ago
LIBS_PROTOC = protoc protobuf
Initial import.
10 years ago
ifneq ($(wildcard /usr/src/gtest/src/gtest-all.cc),)
GTEST_LIB = /usr/src/gtest/src/gtest-all.cc -I/usr/src/gtest
else
GTEST_LIB = -lgtest
endif
Build CPP interop client in Git-on-[] as well as in []. To build client in Git-on-[]: # regenerate Makefile net/grpc/tools/buildgen/generate_projects.sh # generate .pb.h and .pb.cc protoc --cpp_out=. --grpc_out=. --plugin=protoc-gen-grpc=[]-bin/net/grpc/compiler/cpp_plugin net/grpc/cpp/test/interop/test.proto net/grpc/cpp/test/interop/empty.proto net/grpc/cpp/test/interop/messages.proto # Complile and link net/grpc/tools/build_grpc_dev.sh bins/interop_client To test against GFE/ESF: # bring up server [] build net/grpc/testing/interop:server_components_env []-bin/net/grpc/testing/interop/server_components_env --manual --rpc_port=25000 # start client /tmp/grpc-codebase/bins/interop_client --enable_ssl=true --server_port="server ssl port listening port" To test [] build against GFE/ESF: [] run net/grpc/cpp:interop_client -- --enable_ssl=true --server_port="server ssl port listening port" Change on 2014/12/10 by chenw <chenw@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81821921
10 years ago
GTEST_LIB += -lgflags
Initial import.
10 years ago
ifeq ($(V),1)
E = @:
Q =
else
E = @echo
Q = @
endif
VERSION = 0.8.0.0
CPPFLAGS_NO_ARCH += $(addprefix -I, $(INCLUDES)) $(addprefix -D, $(DEFINES))
CPPFLAGS += $(CPPFLAGS_NO_ARCH) $(ARCH_FLAGS)
LDFLAGS += $(ARCH_FLAGS)
LDLIBS += $(addprefix -l, $(LIBS))
LDLIBSXX += $(addprefix -l, $(LIBSXX))
LDLIBS_SECURE += $(addprefix -l, $(LIBS_SECURE))
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
HOST_LDLIBS_PROTOC += $(addprefix -l, $(LIBS_PROTOC))
HOST_CPPFLAGS = $(CPPFLAGS)
HOST_CFLAGS = $(CFLAGS)
HOST_CXXFLAGS = $(CXXFLAGS)
HOST_LDFLAGS = $(LDFLAGS)
HOST_LDLIBS = $(LDLIBS)
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
# These are automatically computed variables.
# There shouldn't be any need to change anything from now on.
HOST_SYSTEM = $(shell uname | cut -f 1 -d_)
ifeq ($(SYSTEM),)
SYSTEM = $(HOST_SYSTEM)
endif
ifeq ($(wildcard .git),)
IS_GIT_FOLDER = false
else
IS_GIT_FOLDER = true
endif
EVENT2_CHECK_CMD = $(CC) $(CFLAGS) $(CPPFLAGS) -o /dev/null test/build/event2.c -levent $(LDFLAGS)
OPENSSL_ALPN_CHECK_CMD = $(CC) $(CFLAGS) $(CPPFLAGS) -o /dev/null test/build/openssl-alpn.c -levent $(LDFLAGS) $(LDLIBS_SECURE)
ZLIB_CHECK_CMD = $(CC) $(CFLAGS) $(CPPFLAGS) -o /dev/null test/build/event2.c -levent $(LDFLAGS)
HAS_SYSTEM_EVENT2 = $(shell $(EVENT2_CHECK_CMD) >& /dev/null && echo true || echo false)
HAS_SYSTEM_OPENSSL_ALPN = $(shell $(OPENSSL_ALPN_CHECK_CMD) >& /dev/null && echo true || echo false)
HAS_SYSTEM_ZLIB = $(shell $(ZLIB_CHECK_CMD) >& /dev/null && echo true || echo false)
ifeq ($(wildcard third_party/libevent/include/event2/event.h),)
HAS_EMBEDDED_EVENT2 = false
else
HAS_EMBEDDED_EVENT2 = true
endif
ifeq ($(wildcard third_party/openssl/ssl/ssl.h),)
HAS_EMBEDDED_OPENSSL_ALPN = false
else
HAS_EMBEDDED_OPENSSL_ALPN = true
endif
ifeq ($(wildcard third_party/zlib/zlib.h),)
HAS_EMBEDDED_ZLIB = false
else
HAS_EMBEDDED_ZLIB = true
endif
ifneq ($(SYSTEM),MINGW32)
ifeq ($(HAS_SYSTEM_EVENT2),false)
DEP_MISSING += libevent
endif
endif
ifeq ($(HAS_SYSTEM_ZLIB),false)
ifeq ($(HAS_EMBEDDED_ZLIB),true)
ZLIB_DEP = third_party/zlib/libz.a
CPPFLAGS += -Ithird_party/zlib
LDFLAGS += -Lthird_party/zlib
else
DEP_MISSING += zlib
endif
endif
ifeq ($(HAS_SYSTEM_OPENSSL_ALPN),false)
ifeq ($(HAS_EMBEDDED_OPENSSL_ALPN),true)
OPENSSL_DEP = third_party/openssl/libssl.a
CPPFLAGS += -Ithird_party/openssl/include
LDFLAGS += -Lthird_party/openssl
else
NO_SECURE = true
endif
endif
ifneq ($(DEP_MISSING),)
NO_DEPS = true
endif
ifneq ($(MAKECMDGOALS),clean)
NO_DEPS = true
endif
Initial import.
10 years ago
.SECONDARY = %.pb.h %.pb.cc
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(DEP_MISSING),)
Initial import.
10 years ago
all: static shared
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
dep_error:
@echo "You shouldn't see this message - all of your dependencies are correct."
else
all: dep_error git_update stop
dep_error:
@echo
@echo "DEPENDENCY ERROR"
@echo
@echo "You are missing system dependencies that are essential to build grpc,"
@echo "and the third_party directory doesn't have them:"
@echo
@echo " $(DEP_MISSING)"
@echo
@echo "Installing the development packages for your system will solve"
@echo "this issue. Please consult INSTALL to get more information."
@echo
@echo "If you need information about why these tests failed, run:"
@echo
@echo " make run_dep_checks"
@echo
endif
git_update:
ifeq ($(IS_GIT_FOLDER),true)
@echo "Additionally, since you are in a git clone, you can download the"
@echo "missing dependencies in third_party by running the following command:"
@echo
Fix Makefile help text. Change on 2014/12/15 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82155747
10 years ago
@echo " git submodule update --init"
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
@echo
endif
openssl_dep_error: openssl_dep_message git_update stop
openssl_dep_message:
@echo
@echo "DEPENDENCY ERROR"
@echo
@echo "The target you are trying to run requires OpenSSL with ALPN support."
@echo "Your system doesn't have it, and neither does the third_party directory."
@echo
@echo "Please consult INSTALL to get more information."
@echo
@echo "If you need information about why these tests failed, run:"
@echo
@echo " make run_dep_checks"
@echo
stop:
@false
run_dep_checks:
$(EVENT2_CHECK_CMD) || true
$(OPENSSL_ALPN_CHECK_CMD) || true
$(ZLIB_CHECK_CMD) || true
third_party/zlib/libz.a:
(cd third_party/zlib ; CFLAGS="-fPIC -fvisibility=hidden" ./configure --static)
$(MAKE) -C third_party/zlib
third_party/openssl/libssl.a:
(cd third_party/openssl ; CC="$(CC) -fPIC -fvisibility=hidden" ./config)
$(MAKE) -C third_party/openssl build_crypto build_ssl
Better C/C++ split in the Makefile targets, to better prepare slicing it into two different Makefiles. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81076626
10 years ago
static: static_c static_cxx
Initial import.
10 years ago
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
static_c: dep_c libs/libgpr.a libs/libgrpc.a libs/libgrpc_unsecure.a
Initial import.
10 years ago
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
static_cxx: dep_cxx libs/libgrpc++.a
Initial import.
10 years ago
Better C/C++ split in the Makefile targets, to better prepare slicing it into two different Makefiles. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81076626
10 years ago
shared: shared_c shared_cxx
Initial import.
10 years ago
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
shared_c: dep_c libs/libgpr.so.$(VERSION) libs/libgrpc.so.$(VERSION) libs/libgrpc_unsecure.so.$(VERSION)
Initial import.
10 years ago
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
shared_cxx: dep_cxx libs/libgrpc++.so.$(VERSION)
Initial import.
10 years ago
Better C/C++ split in the Makefile targets, to better prepare slicing it into two different Makefiles. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81076626
10 years ago
privatelibs: privatelibs_c privatelibs_cxx
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
privatelibs_c: dep_c libs/libgrpc_test_util.a libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_cancel_after_accept.a libs/libend2end_test_cancel_after_accept_and_writes_closed.a libs/libend2end_test_cancel_after_invoke.a libs/libend2end_test_cancel_before_invoke.a libs/libend2end_test_cancel_in_a_vacuum.a libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.a libs/libend2end_test_early_server_shutdown_finishes_tags.a libs/libend2end_test_invoke_large_request.a libs/libend2end_test_max_concurrent_streams.a libs/libend2end_test_no_op.a libs/libend2end_test_ping_pong_streaming.a libs/libend2end_test_request_response_with_binary_metadata_and_payload.a libs/libend2end_test_request_response_with_metadata_and_payload.a libs/libend2end_test_request_response_with_payload.a libs/libend2end_test_simple_delayed_request.a libs/libend2end_test_simple_request.a libs/libend2end_test_thread_stress.a libs/libend2end_test_writes_done_hangs_with_pending_read.a libs/libend2end_certs.a
Better C/C++ split in the Makefile targets, to better prepare slicing it into two different Makefiles. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81076626
10 years ago
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
privatelibs_cxx: dep_cxx libs/libgrpc_test_util.a libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_cancel_after_accept.a libs/libend2end_test_cancel_after_accept_and_writes_closed.a libs/libend2end_test_cancel_after_invoke.a libs/libend2end_test_cancel_before_invoke.a libs/libend2end_test_cancel_in_a_vacuum.a libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.a libs/libend2end_test_early_server_shutdown_finishes_tags.a libs/libend2end_test_invoke_large_request.a libs/libend2end_test_max_concurrent_streams.a libs/libend2end_test_no_op.a libs/libend2end_test_ping_pong_streaming.a libs/libend2end_test_request_response_with_binary_metadata_and_payload.a libs/libend2end_test_request_response_with_metadata_and_payload.a libs/libend2end_test_request_response_with_payload.a libs/libend2end_test_simple_delayed_request.a libs/libend2end_test_simple_request.a libs/libend2end_test_thread_stress.a libs/libend2end_test_writes_done_hangs_with_pending_read.a libs/libend2end_certs.a
Better C/C++ split in the Makefile targets, to better prepare slicing it into two different Makefiles. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81076626
10 years ago
buildtests: buildtests_c buildtests_cxx
Fix a bug in base-log, and add test. The tests compute usable space in the log, but do so using subtraction on unsigned values and did not correctly account for number of blocks used per-core. This could lead underflow, and an incorrect space calculation. In addition, testing showed that with current defaults, tests would fail on a system with > ~64 cores. This can be alleviated through use of a larger log. An explicit check in the log initialization has been added for this case, and the log size increased if necessary. Change on 2014/12/15 by aveitch <aveitch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82154432
10 years ago
buildtests_c: bin_dep_c privatelibs_c bins/grpc_byte_buffer_reader_test bins/gpr_cancellable_test bins/gpr_log_test bins/gpr_useful_test bins/gpr_cmdline_test bins/gpr_histogram_test bins/gpr_host_port_test bins/gpr_slice_buffer_test bins/gpr_slice_test bins/gpr_string_test bins/gpr_sync_test bins/gpr_thd_test bins/gpr_time_test bins/murmur_hash_test bins/grpc_stream_op_test bins/alpn_test bins/chttp2_stream_encoder_test bins/hpack_table_test bins/chttp2_stream_map_test bins/hpack_parser_test bins/transport_metadata_test bins/chttp2_status_conversion_test bins/chttp2_transport_end2end_test bins/tcp_posix_test bins/dualstack_socket_test bins/no_server_test bins/resolve_address_test bins/sockaddr_utils_test bins/tcp_server_posix_test bins/tcp_client_posix_test bins/grpc_channel_stack_test bins/metadata_buffer_test bins/grpc_completion_queue_test bins/census_window_stats_test bins/census_statistics_quick_test bins/census_statistics_small_log_test bins/census_statistics_performance_test bins/census_statistics_multiple_writers_test bins/census_statistics_multiple_writers_circular_buffer_test bins/census_stub_test bins/census_hash_table_test bins/fling_server bins/fling_client bins/fling_test bins/echo_server bins/echo_client bins/echo_test bins/message_compress_test bins/bin_encoder_test bins/secure_endpoint_test bins/httpcli_format_request_test bins/httpcli_parser_test bins/httpcli_test bins/grpc_credentials_test bins/grpc_base64_test bins/grpc_json_token_test bins/timeout_encoding_test bins/fd_posix_test bins/fling_stream_test bins/lame_client_test bins/alarm_test bins/time_test bins/chttp2_fake_security_cancel_after_accept_test bins/chttp2_fake_security_cancel_after_accept_and_writes_closed_test bins/chttp2_fake_security_cancel_after_invoke_test bins/chttp2_fake_security_cancel_before_invoke_test bins/chttp2_fake_security_cancel_in_a_vacuum_test bins/chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test bins/chttp2_fake_security_early_server_shutdown_finishes_tags_test bins/chttp2_fake_security_invoke_large_request_test bins/chttp2_fake_security_max_concurrent_streams_test bins/chttp2_fake_security_no_op_test bins/chttp2_fake_security_ping_pong_streaming_test bins/chttp2_fake_security_request_response_with_binary_metadata_and_payload_test bins/chttp2_fake_security_request_response_with_metadata_and_payload_test bins/chttp2_fake_security_request_response_with_payload_test bins/chttp2_fake_security_simple_delayed_request_test bins/chttp2_fake_security_simple_request_test bins/chttp2_fake_security_thread_stress_test bins/chttp2_fake_security_writes_done_hangs_with_pending_read_test bins/chttp2_fullstack_cancel_after_accept_test bins/chttp2_fullstack_cancel_after_accept_and_writes_closed_test bins/chttp2_fullstack_cancel_after_invoke_test bins/chttp2_fullstack_cancel_before_invoke_test bins/chttp2_fullstack_cancel_in_a_vacuum_test bins/chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test bins/chttp2_fullstack_early_server_shutdown_finishes_tags_test bins/chttp2_fullstack_invoke_large_request_test bins/chttp2_fullstack_max_concurrent_streams_test bins/chttp2_fullstack_no_op_test bins/chttp2_fullstack_ping_pong_streaming_test bins/chttp2_fullstack_request_response_with_binary_metadata_and_payload_test bins/chttp2_fullstack_request_response_with_metadata_and_payload_test bins/chttp2_fullstack_request_response_with_payload_test bins/chttp2_fullstack_simple_delayed_request_test bins/chttp2_fullstack_simple_request_test bins/chttp2_fullstack_thread_stress_test bins/chttp2_fullstack_writes_done_hangs_with_pending_read_test bins/chttp2_simple_ssl_fullstack_cancel_after_accept_test bins/chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test bins/chttp2_simple_ssl_fullstack_cancel_after_invoke_test bins/chttp2_simple_ssl_fullstack_cancel_before_invoke_test bins/chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_tags_test bins/chttp2_simple_s
Better C/C++ split in the Makefile targets, to better prepare slicing it into two different Makefiles. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81076626
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
buildtests_cxx: bin_dep_cxx privatelibs_cxx bins/thread_pool_test bins/status_test bins/sync_client_async_server_test bins/qps_client bins/qps_server bins/interop_server bins/interop_client bins/end2end_test
Better C/C++ split in the Makefile targets, to better prepare slicing it into two different Makefiles. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81076626
10 years ago
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
test: test_c test_cxx
Better C/C++ split in the Makefile targets, to better prepare slicing it into two different Makefiles. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81076626
10 years ago
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
test_c: buildtests_c
Initial import.
10 years ago
$(E) "[RUN] Testing grpc_byte_buffer_reader_test"
$(Q) ./bins/grpc_byte_buffer_reader_test || ( echo test grpc_byte_buffer_reader_test failed ; exit 1 )
$(E) "[RUN] Testing gpr_cancellable_test"
$(Q) ./bins/gpr_cancellable_test || ( echo test gpr_cancellable_test failed ; exit 1 )
$(E) "[RUN] Testing gpr_log_test"
$(Q) ./bins/gpr_log_test || ( echo test gpr_log_test failed ; exit 1 )
Add a test for useful.h. Additionally add rotl, rotr which will be needed to implement some hashing functions shortly. Change on 2014/12/15 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82149559
10 years ago
$(E) "[RUN] Testing gpr_useful_test"
$(Q) ./bins/gpr_useful_test || ( echo test gpr_useful_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing gpr_cmdline_test"
$(Q) ./bins/gpr_cmdline_test || ( echo test gpr_cmdline_test failed ; exit 1 )
$(E) "[RUN] Testing gpr_histogram_test"
$(Q) ./bins/gpr_histogram_test || ( echo test gpr_histogram_test failed ; exit 1 )
$(E) "[RUN] Testing gpr_host_port_test"
$(Q) ./bins/gpr_host_port_test || ( echo test gpr_host_port_test failed ; exit 1 )
$(E) "[RUN] Testing gpr_slice_buffer_test"
$(Q) ./bins/gpr_slice_buffer_test || ( echo test gpr_slice_buffer_test failed ; exit 1 )
$(E) "[RUN] Testing gpr_slice_test"
$(Q) ./bins/gpr_slice_test || ( echo test gpr_slice_test failed ; exit 1 )
$(E) "[RUN] Testing gpr_string_test"
$(Q) ./bins/gpr_string_test || ( echo test gpr_string_test failed ; exit 1 )
$(E) "[RUN] Testing gpr_sync_test"
$(Q) ./bins/gpr_sync_test || ( echo test gpr_sync_test failed ; exit 1 )
$(E) "[RUN] Testing gpr_thd_test"
$(Q) ./bins/gpr_thd_test || ( echo test gpr_thd_test failed ; exit 1 )
$(E) "[RUN] Testing gpr_time_test"
$(Q) ./bins/gpr_time_test || ( echo test gpr_time_test failed ; exit 1 )
$(E) "[RUN] Testing murmur_hash_test"
$(Q) ./bins/murmur_hash_test || ( echo test murmur_hash_test failed ; exit 1 )
$(E) "[RUN] Testing grpc_stream_op_test"
$(Q) ./bins/grpc_stream_op_test || ( echo test grpc_stream_op_test failed ; exit 1 )
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(E) "[RUN] Testing alpn_test"
$(Q) ./bins/alpn_test || ( echo test alpn_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_stream_encoder_test"
$(Q) ./bins/chttp2_stream_encoder_test || ( echo test chttp2_stream_encoder_test failed ; exit 1 )
$(E) "[RUN] Testing hpack_table_test"
$(Q) ./bins/hpack_table_test || ( echo test hpack_table_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_stream_map_test"
$(Q) ./bins/chttp2_stream_map_test || ( echo test chttp2_stream_map_test failed ; exit 1 )
$(E) "[RUN] Testing hpack_parser_test"
$(Q) ./bins/hpack_parser_test || ( echo test hpack_parser_test failed ; exit 1 )
$(E) "[RUN] Testing transport_metadata_test"
$(Q) ./bins/transport_metadata_test || ( echo test transport_metadata_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_status_conversion_test"
$(Q) ./bins/chttp2_status_conversion_test || ( echo test chttp2_status_conversion_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_transport_end2end_test"
$(Q) ./bins/chttp2_transport_end2end_test || ( echo test chttp2_transport_end2end_test failed ; exit 1 )
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
$(E) "[RUN] Testing tcp_posix_test"
$(Q) ./bins/tcp_posix_test || ( echo test tcp_posix_test failed ; exit 1 )
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(E) "[RUN] Testing dualstack_socket_test"
$(Q) ./bins/dualstack_socket_test || ( echo test dualstack_socket_test failed ; exit 1 )
$(E) "[RUN] Testing no_server_test"
$(Q) ./bins/no_server_test || ( echo test no_server_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing resolve_address_test"
$(Q) ./bins/resolve_address_test || ( echo test resolve_address_test failed ; exit 1 )
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
$(E) "[RUN] Testing sockaddr_utils_test"
$(Q) ./bins/sockaddr_utils_test || ( echo test sockaddr_utils_test failed ; exit 1 )
$(E) "[RUN] Testing tcp_server_posix_test"
$(Q) ./bins/tcp_server_posix_test || ( echo test tcp_server_posix_test failed ; exit 1 )
$(E) "[RUN] Testing tcp_client_posix_test"
$(Q) ./bins/tcp_client_posix_test || ( echo test tcp_client_posix_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing grpc_channel_stack_test"
$(Q) ./bins/grpc_channel_stack_test || ( echo test grpc_channel_stack_test failed ; exit 1 )
$(E) "[RUN] Testing metadata_buffer_test"
$(Q) ./bins/metadata_buffer_test || ( echo test metadata_buffer_test failed ; exit 1 )
$(E) "[RUN] Testing grpc_completion_queue_test"
$(Q) ./bins/grpc_completion_queue_test || ( echo test grpc_completion_queue_test failed ; exit 1 )
$(E) "[RUN] Testing census_window_stats_test"
$(Q) ./bins/census_window_stats_test || ( echo test census_window_stats_test failed ; exit 1 )
$(E) "[RUN] Testing census_statistics_quick_test"
$(Q) ./bins/census_statistics_quick_test || ( echo test census_statistics_quick_test failed ; exit 1 )
Fix a bug in base-log, and add test. The tests compute usable space in the log, but do so using subtraction on unsigned values and did not correctly account for number of blocks used per-core. This could lead underflow, and an incorrect space calculation. In addition, testing showed that with current defaults, tests would fail on a system with > ~64 cores. This can be alleviated through use of a larger log. An explicit check in the log initialization has been added for this case, and the log size increased if necessary. Change on 2014/12/15 by aveitch <aveitch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82154432
10 years ago
$(E) "[RUN] Testing census_statistics_small_log_test"
$(Q) ./bins/census_statistics_small_log_test || ( echo test census_statistics_small_log_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing census_statistics_performance_test"
$(Q) ./bins/census_statistics_performance_test || ( echo test census_statistics_performance_test failed ; exit 1 )
$(E) "[RUN] Testing census_statistics_multiple_writers_test"
$(Q) ./bins/census_statistics_multiple_writers_test || ( echo test census_statistics_multiple_writers_test failed ; exit 1 )
$(E) "[RUN] Testing census_statistics_multiple_writers_circular_buffer_test"
$(Q) ./bins/census_statistics_multiple_writers_circular_buffer_test || ( echo test census_statistics_multiple_writers_circular_buffer_test failed ; exit 1 )
$(E) "[RUN] Testing census_stub_test"
$(Q) ./bins/census_stub_test || ( echo test census_stub_test failed ; exit 1 )
$(E) "[RUN] Testing census_hash_table_test"
$(Q) ./bins/census_hash_table_test || ( echo test census_hash_table_test failed ; exit 1 )
$(E) "[RUN] Testing fling_test"
$(Q) ./bins/fling_test || ( echo test fling_test failed ; exit 1 )
$(E) "[RUN] Testing echo_test"
$(Q) ./bins/echo_test || ( echo test echo_test failed ; exit 1 )
$(E) "[RUN] Testing message_compress_test"
$(Q) ./bins/message_compress_test || ( echo test message_compress_test failed ; exit 1 )
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(E) "[RUN] Testing bin_encoder_test"
$(Q) ./bins/bin_encoder_test || ( echo test bin_encoder_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing secure_endpoint_test"
$(Q) ./bins/secure_endpoint_test || ( echo test secure_endpoint_test failed ; exit 1 )
$(E) "[RUN] Testing httpcli_format_request_test"
$(Q) ./bins/httpcli_format_request_test || ( echo test httpcli_format_request_test failed ; exit 1 )
$(E) "[RUN] Testing httpcli_parser_test"
$(Q) ./bins/httpcli_parser_test || ( echo test httpcli_parser_test failed ; exit 1 )
$(E) "[RUN] Testing httpcli_test"
$(Q) ./bins/httpcli_test || ( echo test httpcli_test failed ; exit 1 )
$(E) "[RUN] Testing grpc_credentials_test"
$(Q) ./bins/grpc_credentials_test || ( echo test grpc_credentials_test failed ; exit 1 )
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
$(E) "[RUN] Testing grpc_base64_test"
$(Q) ./bins/grpc_base64_test || ( echo test grpc_base64_test failed ; exit 1 )
$(E) "[RUN] Testing grpc_json_token_test"
$(Q) ./bins/grpc_json_token_test || ( echo test grpc_json_token_test failed ; exit 1 )
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
$(E) "[RUN] Testing timeout_encoding_test"
$(Q) ./bins/timeout_encoding_test || ( echo test timeout_encoding_test failed ; exit 1 )
$(E) "[RUN] Testing fd_posix_test"
$(Q) ./bins/fd_posix_test || ( echo test fd_posix_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing fling_stream_test"
$(Q) ./bins/fling_stream_test || ( echo test fling_stream_test failed ; exit 1 )
$(E) "[RUN] Testing lame_client_test"
$(Q) ./bins/lame_client_test || ( echo test lame_client_test failed ; exit 1 )
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
$(E) "[RUN] Testing alarm_test"
$(Q) ./bins/alarm_test || ( echo test alarm_test failed ; exit 1 )
$(E) "[RUN] Testing time_test"
$(Q) ./bins/time_test || ( echo test time_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_fake_security_cancel_after_accept_test"
$(Q) ./bins/chttp2_fake_security_cancel_after_accept_test || ( echo test chttp2_fake_security_cancel_after_accept_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_cancel_after_accept_and_writes_closed_test"
$(Q) ./bins/chttp2_fake_security_cancel_after_accept_and_writes_closed_test || ( echo test chttp2_fake_security_cancel_after_accept_and_writes_closed_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_cancel_after_invoke_test"
$(Q) ./bins/chttp2_fake_security_cancel_after_invoke_test || ( echo test chttp2_fake_security_cancel_after_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_cancel_before_invoke_test"
$(Q) ./bins/chttp2_fake_security_cancel_before_invoke_test || ( echo test chttp2_fake_security_cancel_before_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_cancel_in_a_vacuum_test"
$(Q) ./bins/chttp2_fake_security_cancel_in_a_vacuum_test || ( echo test chttp2_fake_security_cancel_in_a_vacuum_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test"
$(Q) ./bins/chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test || ( echo test chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_early_server_shutdown_finishes_tags_test"
$(Q) ./bins/chttp2_fake_security_early_server_shutdown_finishes_tags_test || ( echo test chttp2_fake_security_early_server_shutdown_finishes_tags_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_invoke_large_request_test"
$(Q) ./bins/chttp2_fake_security_invoke_large_request_test || ( echo test chttp2_fake_security_invoke_large_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_max_concurrent_streams_test"
$(Q) ./bins/chttp2_fake_security_max_concurrent_streams_test || ( echo test chttp2_fake_security_max_concurrent_streams_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_no_op_test"
$(Q) ./bins/chttp2_fake_security_no_op_test || ( echo test chttp2_fake_security_no_op_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_ping_pong_streaming_test"
$(Q) ./bins/chttp2_fake_security_ping_pong_streaming_test || ( echo test chttp2_fake_security_ping_pong_streaming_test failed ; exit 1 )
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
$(E) "[RUN] Testing chttp2_fake_security_request_response_with_binary_metadata_and_payload_test"
$(Q) ./bins/chttp2_fake_security_request_response_with_binary_metadata_and_payload_test || ( echo test chttp2_fake_security_request_response_with_binary_metadata_and_payload_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_fake_security_request_response_with_metadata_and_payload_test"
$(Q) ./bins/chttp2_fake_security_request_response_with_metadata_and_payload_test || ( echo test chttp2_fake_security_request_response_with_metadata_and_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_request_response_with_payload_test"
$(Q) ./bins/chttp2_fake_security_request_response_with_payload_test || ( echo test chttp2_fake_security_request_response_with_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_simple_delayed_request_test"
$(Q) ./bins/chttp2_fake_security_simple_delayed_request_test || ( echo test chttp2_fake_security_simple_delayed_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fake_security_simple_request_test"
$(Q) ./bins/chttp2_fake_security_simple_request_test || ( echo test chttp2_fake_security_simple_request_test failed ; exit 1 )
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(E) "[RUN] Testing chttp2_fake_security_thread_stress_test"
$(Q) ./bins/chttp2_fake_security_thread_stress_test || ( echo test chttp2_fake_security_thread_stress_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_fake_security_writes_done_hangs_with_pending_read_test"
$(Q) ./bins/chttp2_fake_security_writes_done_hangs_with_pending_read_test || ( echo test chttp2_fake_security_writes_done_hangs_with_pending_read_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_cancel_after_accept_test"
$(Q) ./bins/chttp2_fullstack_cancel_after_accept_test || ( echo test chttp2_fullstack_cancel_after_accept_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_cancel_after_accept_and_writes_closed_test"
$(Q) ./bins/chttp2_fullstack_cancel_after_accept_and_writes_closed_test || ( echo test chttp2_fullstack_cancel_after_accept_and_writes_closed_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_cancel_after_invoke_test"
$(Q) ./bins/chttp2_fullstack_cancel_after_invoke_test || ( echo test chttp2_fullstack_cancel_after_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_cancel_before_invoke_test"
$(Q) ./bins/chttp2_fullstack_cancel_before_invoke_test || ( echo test chttp2_fullstack_cancel_before_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_cancel_in_a_vacuum_test"
$(Q) ./bins/chttp2_fullstack_cancel_in_a_vacuum_test || ( echo test chttp2_fullstack_cancel_in_a_vacuum_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test"
$(Q) ./bins/chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test || ( echo test chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_early_server_shutdown_finishes_tags_test"
$(Q) ./bins/chttp2_fullstack_early_server_shutdown_finishes_tags_test || ( echo test chttp2_fullstack_early_server_shutdown_finishes_tags_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_invoke_large_request_test"
$(Q) ./bins/chttp2_fullstack_invoke_large_request_test || ( echo test chttp2_fullstack_invoke_large_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_max_concurrent_streams_test"
$(Q) ./bins/chttp2_fullstack_max_concurrent_streams_test || ( echo test chttp2_fullstack_max_concurrent_streams_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_no_op_test"
$(Q) ./bins/chttp2_fullstack_no_op_test || ( echo test chttp2_fullstack_no_op_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_ping_pong_streaming_test"
$(Q) ./bins/chttp2_fullstack_ping_pong_streaming_test || ( echo test chttp2_fullstack_ping_pong_streaming_test failed ; exit 1 )
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
$(E) "[RUN] Testing chttp2_fullstack_request_response_with_binary_metadata_and_payload_test"
$(Q) ./bins/chttp2_fullstack_request_response_with_binary_metadata_and_payload_test || ( echo test chttp2_fullstack_request_response_with_binary_metadata_and_payload_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_fullstack_request_response_with_metadata_and_payload_test"
$(Q) ./bins/chttp2_fullstack_request_response_with_metadata_and_payload_test || ( echo test chttp2_fullstack_request_response_with_metadata_and_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_request_response_with_payload_test"
$(Q) ./bins/chttp2_fullstack_request_response_with_payload_test || ( echo test chttp2_fullstack_request_response_with_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_simple_delayed_request_test"
$(Q) ./bins/chttp2_fullstack_simple_delayed_request_test || ( echo test chttp2_fullstack_simple_delayed_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_fullstack_simple_request_test"
$(Q) ./bins/chttp2_fullstack_simple_request_test || ( echo test chttp2_fullstack_simple_request_test failed ; exit 1 )
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(E) "[RUN] Testing chttp2_fullstack_thread_stress_test"
$(Q) ./bins/chttp2_fullstack_thread_stress_test || ( echo test chttp2_fullstack_thread_stress_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_fullstack_writes_done_hangs_with_pending_read_test"
$(Q) ./bins/chttp2_fullstack_writes_done_hangs_with_pending_read_test || ( echo test chttp2_fullstack_writes_done_hangs_with_pending_read_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_cancel_after_accept_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_cancel_after_accept_test || ( echo test chttp2_simple_ssl_fullstack_cancel_after_accept_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test || ( echo test chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_cancel_after_invoke_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_cancel_after_invoke_test || ( echo test chttp2_simple_ssl_fullstack_cancel_after_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_cancel_before_invoke_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_cancel_before_invoke_test || ( echo test chttp2_simple_ssl_fullstack_cancel_before_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test || ( echo test chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test || ( echo test chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_tags_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_tags_test || ( echo test chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_tags_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_invoke_large_request_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_invoke_large_request_test || ( echo test chttp2_simple_ssl_fullstack_invoke_large_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_max_concurrent_streams_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_max_concurrent_streams_test || ( echo test chttp2_simple_ssl_fullstack_max_concurrent_streams_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_no_op_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_no_op_test || ( echo test chttp2_simple_ssl_fullstack_no_op_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_ping_pong_streaming_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_ping_pong_streaming_test || ( echo test chttp2_simple_ssl_fullstack_ping_pong_streaming_test failed ; exit 1 )
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_request_response_with_binary_metadata_and_payload_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_request_response_with_binary_metadata_and_payload_test || ( echo test chttp2_simple_ssl_fullstack_request_response_with_binary_metadata_and_payload_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_request_response_with_metadata_and_payload_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_request_response_with_metadata_and_payload_test || ( echo test chttp2_simple_ssl_fullstack_request_response_with_metadata_and_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_request_response_with_payload_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_request_response_with_payload_test || ( echo test chttp2_simple_ssl_fullstack_request_response_with_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_simple_delayed_request_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_simple_delayed_request_test || ( echo test chttp2_simple_ssl_fullstack_simple_delayed_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_simple_request_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_simple_request_test || ( echo test chttp2_simple_ssl_fullstack_simple_request_test failed ; exit 1 )
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_thread_stress_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_thread_stress_test || ( echo test chttp2_simple_ssl_fullstack_thread_stress_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_simple_ssl_fullstack_writes_done_hangs_with_pending_read_test"
$(Q) ./bins/chttp2_simple_ssl_fullstack_writes_done_hangs_with_pending_read_test || ( echo test chttp2_simple_ssl_fullstack_writes_done_hangs_with_pending_read_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_and_writes_closed_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_and_writes_closed_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_and_writes_closed_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_invoke_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_invoke_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_cancel_before_invoke_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_before_invoke_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_cancel_before_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_cancel_in_a_vacuum_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_in_a_vacuum_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_cancel_in_a_vacuum_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_inflight_calls_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_inflight_calls_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_inflight_calls_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_tags_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_tags_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_tags_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_invoke_large_request_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_invoke_large_request_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_invoke_large_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_max_concurrent_streams_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_max_concurrent_streams_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_max_concurrent_streams_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_no_op_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_no_op_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_no_op_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_ping_pong_streaming_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_ping_pong_streaming_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_ping_pong_streaming_test failed ; exit 1 )
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_binary_metadata_and_payload_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_binary_metadata_and_payload_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_binary_metadata_and_payload_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_metadata_and_payload_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_metadata_and_payload_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_metadata_and_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_payload_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_payload_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_simple_delayed_request_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_simple_delayed_request_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_simple_delayed_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_simple_request_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_simple_request_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_simple_request_test failed ; exit 1 )
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_thread_stress_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_thread_stress_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_thread_stress_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_simple_ssl_with_oauth2_fullstack_writes_done_hangs_with_pending_read_test"
$(Q) ./bins/chttp2_simple_ssl_with_oauth2_fullstack_writes_done_hangs_with_pending_read_test || ( echo test chttp2_simple_ssl_with_oauth2_fullstack_writes_done_hangs_with_pending_read_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_cancel_after_accept_test"
$(Q) ./bins/chttp2_socket_pair_cancel_after_accept_test || ( echo test chttp2_socket_pair_cancel_after_accept_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_cancel_after_accept_and_writes_closed_test"
$(Q) ./bins/chttp2_socket_pair_cancel_after_accept_and_writes_closed_test || ( echo test chttp2_socket_pair_cancel_after_accept_and_writes_closed_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_cancel_after_invoke_test"
$(Q) ./bins/chttp2_socket_pair_cancel_after_invoke_test || ( echo test chttp2_socket_pair_cancel_after_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_cancel_before_invoke_test"
$(Q) ./bins/chttp2_socket_pair_cancel_before_invoke_test || ( echo test chttp2_socket_pair_cancel_before_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_cancel_in_a_vacuum_test"
$(Q) ./bins/chttp2_socket_pair_cancel_in_a_vacuum_test || ( echo test chttp2_socket_pair_cancel_in_a_vacuum_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_early_server_shutdown_finishes_inflight_calls_test"
$(Q) ./bins/chttp2_socket_pair_early_server_shutdown_finishes_inflight_calls_test || ( echo test chttp2_socket_pair_early_server_shutdown_finishes_inflight_calls_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_early_server_shutdown_finishes_tags_test"
$(Q) ./bins/chttp2_socket_pair_early_server_shutdown_finishes_tags_test || ( echo test chttp2_socket_pair_early_server_shutdown_finishes_tags_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_invoke_large_request_test"
$(Q) ./bins/chttp2_socket_pair_invoke_large_request_test || ( echo test chttp2_socket_pair_invoke_large_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_max_concurrent_streams_test"
$(Q) ./bins/chttp2_socket_pair_max_concurrent_streams_test || ( echo test chttp2_socket_pair_max_concurrent_streams_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_no_op_test"
$(Q) ./bins/chttp2_socket_pair_no_op_test || ( echo test chttp2_socket_pair_no_op_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_ping_pong_streaming_test"
$(Q) ./bins/chttp2_socket_pair_ping_pong_streaming_test || ( echo test chttp2_socket_pair_ping_pong_streaming_test failed ; exit 1 )
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
$(E) "[RUN] Testing chttp2_socket_pair_request_response_with_binary_metadata_and_payload_test"
$(Q) ./bins/chttp2_socket_pair_request_response_with_binary_metadata_and_payload_test || ( echo test chttp2_socket_pair_request_response_with_binary_metadata_and_payload_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_socket_pair_request_response_with_metadata_and_payload_test"
$(Q) ./bins/chttp2_socket_pair_request_response_with_metadata_and_payload_test || ( echo test chttp2_socket_pair_request_response_with_metadata_and_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_request_response_with_payload_test"
$(Q) ./bins/chttp2_socket_pair_request_response_with_payload_test || ( echo test chttp2_socket_pair_request_response_with_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_simple_delayed_request_test"
$(Q) ./bins/chttp2_socket_pair_simple_delayed_request_test || ( echo test chttp2_socket_pair_simple_delayed_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_simple_request_test"
$(Q) ./bins/chttp2_socket_pair_simple_request_test || ( echo test chttp2_socket_pair_simple_request_test failed ; exit 1 )
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(E) "[RUN] Testing chttp2_socket_pair_thread_stress_test"
$(Q) ./bins/chttp2_socket_pair_thread_stress_test || ( echo test chttp2_socket_pair_thread_stress_test failed ; exit 1 )
Initial import.
10 years ago
$(E) "[RUN] Testing chttp2_socket_pair_writes_done_hangs_with_pending_read_test"
$(Q) ./bins/chttp2_socket_pair_writes_done_hangs_with_pending_read_test || ( echo test chttp2_socket_pair_writes_done_hangs_with_pending_read_test failed ; exit 1 )
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_and_writes_closed_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_and_writes_closed_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_and_writes_closed_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_cancel_after_invoke_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_invoke_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_cancel_after_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_cancel_before_invoke_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_cancel_before_invoke_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_cancel_before_invoke_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_cancel_in_a_vacuum_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_cancel_in_a_vacuum_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_cancel_in_a_vacuum_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_inflight_calls_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_inflight_calls_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_inflight_calls_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_tags_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_tags_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_tags_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_invoke_large_request_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_invoke_large_request_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_invoke_large_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_max_concurrent_streams_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_max_concurrent_streams_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_max_concurrent_streams_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_no_op_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_no_op_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_no_op_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_ping_pong_streaming_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_ping_pong_streaming_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_ping_pong_streaming_test failed ; exit 1 )
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_request_response_with_binary_metadata_and_payload_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_binary_metadata_and_payload_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_request_response_with_binary_metadata_and_payload_test failed ; exit 1 )
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_request_response_with_metadata_and_payload_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_metadata_and_payload_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_request_response_with_metadata_and_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_request_response_with_payload_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_payload_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_request_response_with_payload_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_simple_delayed_request_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_simple_delayed_request_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_simple_delayed_request_test failed ; exit 1 )
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_simple_request_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_simple_request_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_simple_request_test failed ; exit 1 )
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_thread_stress_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_thread_stress_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_thread_stress_test failed ; exit 1 )
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(E) "[RUN] Testing chttp2_socket_pair_one_byte_at_a_time_writes_done_hangs_with_pending_read_test"
$(Q) ./bins/chttp2_socket_pair_one_byte_at_a_time_writes_done_hangs_with_pending_read_test || ( echo test chttp2_socket_pair_one_byte_at_a_time_writes_done_hangs_with_pending_read_test failed ; exit 1 )
Initial import.
10 years ago
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
test_cxx: buildtests_cxx
Better C/C++ split in the Makefile targets, to better prepare slicing it into two different Makefiles. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81076626
10 years ago
$(E) "[RUN] Testing thread_pool_test"
$(Q) ./bins/thread_pool_test || ( echo test thread_pool_test failed ; exit 1 )
$(E) "[RUN] Testing status_test"
$(Q) ./bins/status_test || ( echo test status_test failed ; exit 1 )
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
$(E) "[RUN] Testing sync_client_async_server_test"
$(Q) ./bins/sync_client_async_server_test || ( echo test sync_client_async_server_test failed ; exit 1 )
$(E) "[RUN] Testing qps_client"
$(Q) ./bins/qps_client || ( echo test qps_client failed ; exit 1 )
$(E) "[RUN] Testing qps_server"
$(Q) ./bins/qps_server || ( echo test qps_server failed ; exit 1 )
$(E) "[RUN] Testing end2end_test"
$(Q) ./bins/end2end_test || ( echo test end2end_test failed ; exit 1 )
Better C/C++ split in the Makefile targets, to better prepare slicing it into two different Makefiles. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81076626
10 years ago
Initial import.
10 years ago
tools: privatelibs bins/gen_hpack_tables
First pass at making the compilers open-sourcable. Change on 2014/12/10 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81829127
10 years ago
protoc_plugins: bins/cpp_plugin bins/ruby_plugin
Initial import.
10 years ago
buildbenchmarks: privatelibs bins/grpc_completion_queue_benchmark bins/low_level_ping_pong_benchmark
benchmarks: buildbenchmarks
strip: strip-static strip-shared
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
strip-static_c: static_c
Initial import.
10 years ago
$(E) "[STRIP] Stripping libgpr.a"
$(Q) $(STRIP) libs/libgpr.a
$(E) "[STRIP] Stripping libgrpc.a"
$(Q) $(STRIP) libs/libgrpc.a
$(E) "[STRIP] Stripping libgrpc_unsecure.a"
$(Q) $(STRIP) libs/libgrpc_unsecure.a
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
strip-static_cxx: static_cxx
$(E) "[STRIP] Stripping libgrpc++.a"
$(Q) $(STRIP) libs/libgrpc++.a
strip-shared_c: shared_c
Initial import.
10 years ago
$(E) "[STRIP] Stripping libgpr.so"
$(Q) $(STRIP) libs/libgpr.so.$(VERSION)
$(E) "[STRIP] Stripping libgrpc.so"
$(Q) $(STRIP) libs/libgrpc.so.$(VERSION)
$(E) "[STRIP] Stripping libgrpc_unsecure.so"
$(Q) $(STRIP) libs/libgrpc_unsecure.so.$(VERSION)
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
strip-shared_cxx: shared_cxx
$(E) "[STRIP] Stripping libgrpc++.so"
$(Q) $(STRIP) libs/libgrpc++.so.$(VERSION)
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
deps/gens/test/cpp/interop/empty.pb.dep:
$(Q) mkdir -p `dirname $@`
$(Q) touch $@
gens/test/cpp/interop/empty.pb.cc: test/cpp/interop/empty.proto protoc_plugins
Initial import.
10 years ago
$(E) "[PROTOC] Generating protobuf CC file from $<"
$(Q) mkdir -p `dirname $@`
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
$(Q) $(PROTOC) --cpp_out=gens --grpc_out=gens --plugin=protoc-gen-grpc=bins/cpp_plugin $<
Initial import.
10 years ago
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
deps/gens/test/cpp/interop/messages.pb.dep:
Initial import.
10 years ago
$(Q) mkdir -p `dirname $@`
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
$(Q) touch $@
gens/test/cpp/interop/messages.pb.cc: test/cpp/interop/messages.proto protoc_plugins
$(E) "[PROTOC] Generating protobuf CC file from $<"
$(Q) mkdir -p `dirname $@`
$(Q) $(PROTOC) --cpp_out=gens --grpc_out=gens --plugin=protoc-gen-grpc=bins/cpp_plugin $<
deps/gens/test/cpp/interop/test.pb.dep:
$(Q) mkdir -p `dirname $@`
$(Q) touch $@
gens/test/cpp/interop/test.pb.cc: test/cpp/interop/test.proto protoc_plugins
$(E) "[PROTOC] Generating protobuf CC file from $<"
$(Q) mkdir -p `dirname $@`
$(Q) $(PROTOC) --cpp_out=gens --grpc_out=gens --plugin=protoc-gen-grpc=bins/cpp_plugin $<
deps/gens/test/cpp/util/echo.pb.dep:
$(Q) mkdir -p `dirname $@`
$(Q) touch $@
gens/test/cpp/util/echo.pb.cc: test/cpp/util/echo.proto protoc_plugins
$(E) "[PROTOC] Generating protobuf CC file from $<"
$(Q) mkdir -p `dirname $@`
$(Q) $(PROTOC) --cpp_out=gens --grpc_out=gens --plugin=protoc-gen-grpc=bins/cpp_plugin $<
Initial import.
10 years ago
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
deps/%.dep : %.c
Initial import.
10 years ago
$(E) "[DEP] Generating dependencies for $<"
$(Q) mkdir -p `dirname $@`
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS_NO_ARCH) -MG -M $< > $@
Initial import.
10 years ago
deps/%.dep : %.cc
$(E) "[DEP] Generating dependencies for $<"
$(Q) mkdir -p `dirname $@`
$(Q) $(CXX) $(CXXFLAGS) $(CPPFLAGS_NO_ARCH) -MG -M $< > $@
objs/%.o : %.c
$(E) "[C] Compiling $<"
$(Q) mkdir -p `dirname $@`
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) -c -o $@ $<
objs/%.o : gens/%.pb.cc
$(E) "[CXX] Compiling $<"
$(Q) mkdir -p `dirname $@`
$(Q) $(CXX) $(CXXFLAGS) $(CPPFLAGS) -c -o $@ $<
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
objs/src/compiler/%.o : src/compiler/%.cc
$(E) "[HOSTCXX] Compiling $<"
$(Q) mkdir -p `dirname $@`
$(Q) $(HOST_CXX) $(HOST_CXXFLAGS) $(HOST_CPPFLAGS) -c -o $@ $<
Initial import.
10 years ago
objs/%.o : %.cc
$(E) "[CXX] Compiling $<"
$(Q) mkdir -p `dirname $@`
$(Q) $(CXX) $(CXXFLAGS) $(CPPFLAGS) -c -o $@ $<
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
dep: dep_c dep_cxx
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
dep_c: deps_libgpr deps_libgrpc deps_libgrpc_test_util deps_libend2end_fixture_chttp2_fake_security deps_libend2end_fixture_chttp2_fullstack deps_libend2end_fixture_chttp2_simple_ssl_fullstack deps_libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack deps_libend2end_fixture_chttp2_socket_pair deps_libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time deps_libend2end_test_cancel_after_accept deps_libend2end_test_cancel_after_accept_and_writes_closed deps_libend2end_test_cancel_after_invoke deps_libend2end_test_cancel_before_invoke deps_libend2end_test_cancel_in_a_vacuum deps_libend2end_test_early_server_shutdown_finishes_inflight_calls deps_libend2end_test_early_server_shutdown_finishes_tags deps_libend2end_test_invoke_large_request deps_libend2end_test_max_concurrent_streams deps_libend2end_test_no_op deps_libend2end_test_ping_pong_streaming deps_libend2end_test_request_response_with_binary_metadata_and_payload deps_libend2end_test_request_response_with_metadata_and_payload deps_libend2end_test_request_response_with_payload deps_libend2end_test_simple_delayed_request deps_libend2end_test_simple_request deps_libend2end_test_thread_stress deps_libend2end_test_writes_done_hangs_with_pending_read deps_libend2end_certs deps_libgrpc_unsecure
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Fix a bug in base-log, and add test. The tests compute usable space in the log, but do so using subtraction on unsigned values and did not correctly account for number of blocks used per-core. This could lead underflow, and an incorrect space calculation. In addition, testing showed that with current defaults, tests would fail on a system with > ~64 cores. This can be alleviated through use of a larger log. An explicit check in the log initialization has been added for this case, and the log size increased if necessary. Change on 2014/12/15 by aveitch <aveitch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82154432
10 years ago
bins_dep_c: deps_gen_hpack_tables deps_grpc_byte_buffer_reader_test deps_gpr_cancellable_test deps_gpr_log_test deps_gpr_useful_test deps_gpr_cmdline_test deps_gpr_histogram_test deps_gpr_host_port_test deps_gpr_slice_buffer_test deps_gpr_slice_test deps_gpr_string_test deps_gpr_sync_test deps_gpr_thd_test deps_gpr_time_test deps_murmur_hash_test deps_grpc_stream_op_test deps_alpn_test deps_chttp2_stream_encoder_test deps_hpack_table_test deps_chttp2_stream_map_test deps_hpack_parser_test deps_transport_metadata_test deps_chttp2_status_conversion_test deps_chttp2_transport_end2end_test deps_tcp_posix_test deps_dualstack_socket_test deps_no_server_test deps_resolve_address_test deps_sockaddr_utils_test deps_tcp_server_posix_test deps_tcp_client_posix_test deps_grpc_channel_stack_test deps_metadata_buffer_test deps_grpc_completion_queue_test deps_grpc_completion_queue_benchmark deps_census_window_stats_test deps_census_statistics_quick_test deps_census_statistics_small_log_test deps_census_statistics_performance_test deps_census_statistics_multiple_writers_test deps_census_statistics_multiple_writers_circular_buffer_test deps_census_stub_test deps_census_hash_table_test deps_fling_server deps_fling_client deps_fling_test deps_echo_server deps_echo_client deps_echo_test deps_low_level_ping_pong_benchmark deps_message_compress_test deps_bin_encoder_test deps_secure_endpoint_test deps_httpcli_format_request_test deps_httpcli_parser_test deps_httpcli_test deps_grpc_credentials_test deps_grpc_base64_test deps_grpc_json_token_test deps_timeout_encoding_test deps_fd_posix_test deps_fling_stream_test deps_lame_client_test deps_alarm_test deps_time_test deps_chttp2_fake_security_cancel_after_accept_test deps_chttp2_fake_security_cancel_after_accept_and_writes_closed_test deps_chttp2_fake_security_cancel_after_invoke_test deps_chttp2_fake_security_cancel_before_invoke_test deps_chttp2_fake_security_cancel_in_a_vacuum_test deps_chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test deps_chttp2_fake_security_early_server_shutdown_finishes_tags_test deps_chttp2_fake_security_invoke_large_request_test deps_chttp2_fake_security_max_concurrent_streams_test deps_chttp2_fake_security_no_op_test deps_chttp2_fake_security_ping_pong_streaming_test deps_chttp2_fake_security_request_response_with_binary_metadata_and_payload_test deps_chttp2_fake_security_request_response_with_metadata_and_payload_test deps_chttp2_fake_security_request_response_with_payload_test deps_chttp2_fake_security_simple_delayed_request_test deps_chttp2_fake_security_simple_request_test deps_chttp2_fake_security_thread_stress_test deps_chttp2_fake_security_writes_done_hangs_with_pending_read_test deps_chttp2_fullstack_cancel_after_accept_test deps_chttp2_fullstack_cancel_after_accept_and_writes_closed_test deps_chttp2_fullstack_cancel_after_invoke_test deps_chttp2_fullstack_cancel_before_invoke_test deps_chttp2_fullstack_cancel_in_a_vacuum_test deps_chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test deps_chttp2_fullstack_early_server_shutdown_finishes_tags_test deps_chttp2_fullstack_invoke_large_request_test deps_chttp2_fullstack_max_concurrent_streams_test deps_chttp2_fullstack_no_op_test deps_chttp2_fullstack_ping_pong_streaming_test deps_chttp2_fullstack_request_response_with_binary_metadata_and_payload_test deps_chttp2_fullstack_request_response_with_metadata_and_payload_test deps_chttp2_fullstack_request_response_with_payload_test deps_chttp2_fullstack_simple_delayed_request_test deps_chttp2_fullstack_simple_request_test deps_chttp2_fullstack_thread_stress_test deps_chttp2_fullstack_writes_done_hangs_with_pending_read_test deps_chttp2_simple_ssl_fullstack_cancel_after_accept_test deps_chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test deps_chttp2_simple_ssl_fullstack_cancel_after_invoke_test deps_chttp2_simple_ssl_fullstack_cancel_before_invoke_test deps_chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test deps_chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test deps_chttp2_simple_ssl_ful
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
dep_cxx: deps_libgrpc++ deps_libgrpc++_test_util
bins_dep_cxx: deps_cpp_plugin deps_ruby_plugin deps_thread_pool_test deps_status_test deps_sync_client_async_server_test deps_qps_client deps_qps_server deps_interop_server deps_interop_client deps_end2end_test
Initial import.
10 years ago
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
install: install_c install_cxx
install_c: install-headers_c install-static_c install-shared_c
install_cxx: install-headers_cxx install-static_cxx install-shared_cxx
install-headers: install-headers_c install-headers_cxx
Initial import.
10 years ago
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
install-headers_c:
$(E) "[INSTALL] Installing public C headers"
$(Q) $(foreach h, $(PUBLIC_HEADERS_C), $(INSTALL) $(h) $(prefix)/$(h) && ) exit 0 || exit 1
Initial import.
10 years ago
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
install-headers_cxx:
$(E) "[INSTALL] Installing public C++ headers"
$(Q) $(foreach h, $(PUBLIC_HEADERS_CXX), $(INSTALL) $(h) $(prefix)/$(h) && ) exit 0 || exit 1
install-static: install-static_c install-static_cxx
install-static_c: static_c strip-static_c
Initial import.
10 years ago
$(E) "[INSTALL] Installing libgpr.a"
$(Q) $(INSTALL) libs/libgpr.a $(prefix)/lib/libgpr.a
$(E) "[INSTALL] Installing libgrpc.a"
$(Q) $(INSTALL) libs/libgrpc.a $(prefix)/lib/libgrpc.a
$(E) "[INSTALL] Installing libgrpc_unsecure.a"
$(Q) $(INSTALL) libs/libgrpc_unsecure.a $(prefix)/lib/libgrpc_unsecure.a
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
install-static_cxx: static_cxx strip-static_cxx
$(E) "[INSTALL] Installing libgrpc++.a"
$(Q) $(INSTALL) libs/libgrpc++.a $(prefix)/lib/libgrpc++.a
install-shared_c: shared_c strip-shared_c
Initial import.
10 years ago
$(E) "[INSTALL] Installing libgpr.so"
$(Q) $(INSTALL) libs/libgpr.so.$(VERSION) $(prefix)/lib/libgpr.so.$(VERSION)
$(E) "[INSTALL] Installing libgrpc.so"
$(Q) $(INSTALL) libs/libgrpc.so.$(VERSION) $(prefix)/lib/libgrpc.so.$(VERSION)
$(E) "[INSTALL] Installing libgrpc_unsecure.so"
$(Q) $(INSTALL) libs/libgrpc_unsecure.so.$(VERSION) $(prefix)/lib/libgrpc_unsecure.so.$(VERSION)
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
install-shared_cxx: shared_cxx strip-shared_cxx
$(E) "[INSTALL] Installing libgrpc++.so"
$(Q) $(INSTALL) libs/libgrpc++.so.$(VERSION) $(prefix)/lib/libgrpc++.so.$(VERSION)
Fix a bug in base-log, and add test. The tests compute usable space in the log, but do so using subtraction on unsigned values and did not correctly account for number of blocks used per-core. This could lead underflow, and an incorrect space calculation. In addition, testing showed that with current defaults, tests would fail on a system with > ~64 cores. This can be alleviated through use of a larger log. An explicit check in the log initialization has been added for this case, and the log size increased if necessary. Change on 2014/12/15 by aveitch <aveitch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82154432
10 years ago
clean: clean_libgpr clean_libgrpc clean_libgrpc_test_util clean_libgrpc++ clean_libgrpc++_test_util clean_libend2end_fixture_chttp2_fake_security clean_libend2end_fixture_chttp2_fullstack clean_libend2end_fixture_chttp2_simple_ssl_fullstack clean_libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack clean_libend2end_fixture_chttp2_socket_pair clean_libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time clean_libend2end_test_cancel_after_accept clean_libend2end_test_cancel_after_accept_and_writes_closed clean_libend2end_test_cancel_after_invoke clean_libend2end_test_cancel_before_invoke clean_libend2end_test_cancel_in_a_vacuum clean_libend2end_test_early_server_shutdown_finishes_inflight_calls clean_libend2end_test_early_server_shutdown_finishes_tags clean_libend2end_test_invoke_large_request clean_libend2end_test_max_concurrent_streams clean_libend2end_test_no_op clean_libend2end_test_ping_pong_streaming clean_libend2end_test_request_response_with_binary_metadata_and_payload clean_libend2end_test_request_response_with_metadata_and_payload clean_libend2end_test_request_response_with_payload clean_libend2end_test_simple_delayed_request clean_libend2end_test_simple_request clean_libend2end_test_thread_stress clean_libend2end_test_writes_done_hangs_with_pending_read clean_libend2end_certs clean_libgrpc_unsecure clean_gen_hpack_tables clean_cpp_plugin clean_ruby_plugin clean_grpc_byte_buffer_reader_test clean_gpr_cancellable_test clean_gpr_log_test clean_gpr_useful_test clean_gpr_cmdline_test clean_gpr_histogram_test clean_gpr_host_port_test clean_gpr_slice_buffer_test clean_gpr_slice_test clean_gpr_string_test clean_gpr_sync_test clean_gpr_thd_test clean_gpr_time_test clean_murmur_hash_test clean_grpc_stream_op_test clean_alpn_test clean_chttp2_stream_encoder_test clean_hpack_table_test clean_chttp2_stream_map_test clean_hpack_parser_test clean_transport_metadata_test clean_chttp2_status_conversion_test clean_chttp2_transport_end2end_test clean_tcp_posix_test clean_dualstack_socket_test clean_no_server_test clean_resolve_address_test clean_sockaddr_utils_test clean_tcp_server_posix_test clean_tcp_client_posix_test clean_grpc_channel_stack_test clean_metadata_buffer_test clean_grpc_completion_queue_test clean_grpc_completion_queue_benchmark clean_census_window_stats_test clean_census_statistics_quick_test clean_census_statistics_small_log_test clean_census_statistics_performance_test clean_census_statistics_multiple_writers_test clean_census_statistics_multiple_writers_circular_buffer_test clean_census_stub_test clean_census_hash_table_test clean_fling_server clean_fling_client clean_fling_test clean_echo_server clean_echo_client clean_echo_test clean_low_level_ping_pong_benchmark clean_message_compress_test clean_bin_encoder_test clean_secure_endpoint_test clean_httpcli_format_request_test clean_httpcli_parser_test clean_httpcli_test clean_grpc_credentials_test clean_grpc_base64_test clean_grpc_json_token_test clean_timeout_encoding_test clean_fd_posix_test clean_fling_stream_test clean_lame_client_test clean_thread_pool_test clean_status_test clean_sync_client_async_server_test clean_qps_client clean_qps_server clean_interop_server clean_interop_client clean_end2end_test clean_alarm_test clean_time_test clean_chttp2_fake_security_cancel_after_accept_test clean_chttp2_fake_security_cancel_after_accept_and_writes_closed_test clean_chttp2_fake_security_cancel_after_invoke_test clean_chttp2_fake_security_cancel_before_invoke_test clean_chttp2_fake_security_cancel_in_a_vacuum_test clean_chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test clean_chttp2_fake_security_early_server_shutdown_finishes_tags_test clean_chttp2_fake_security_invoke_large_request_test clean_chttp2_fake_security_max_concurrent_streams_test clean_chttp2_fake_security_no_op_test clean_chttp2_fake_security_ping_pong_streaming_test clean_chttp2_fake_security_request_response_with_binary_metadata_and_payload_test clean_chttp2_fake_security_request_response_with_metadata_and_payload_test clean_chttp2_fake_security_request_response_with_
Initial import.
10 years ago
$(Q) $(RM) -r deps objs libs bins gens
# The various libraries
LIBGPR_SRC = \
src/core/support/alloc.c \
src/core/support/cancellable.c \
src/core/support/cmdline.c \
src/core/support/cpu_posix.c \
src/core/support/histogram.c \
src/core/support/host_port.c \
src/core/support/log.c \
src/core/support/log_posix.c \
src/core/support/log_linux.c \
src/core/support/log_android.c \
src/core/support/log_win32.c \
src/core/support/murmur_hash.c \
src/core/support/slice.c \
src/core/support/slice_buffer.c \
src/core/support/string.c \
src/core/support/string_posix.c \
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
src/core/support/string_win32.c \
Initial import.
10 years ago
src/core/support/sync.c \
src/core/support/sync_posix.c \
Fixes for GPR library on Windows (+ getting rid of warnings under MS C++ compiler) Change on 2014/12/09 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81704538
10 years ago
src/core/support/sync_win32.c \
Initial import.
10 years ago
src/core/support/thd_posix.c \
src/core/support/thd_win32.c \
src/core/support/time.c \
src/core/support/time_posix.c \
src/core/support/time_win32.c \
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
PUBLIC_HEADERS_C += \
Initial import.
10 years ago
include/grpc/support/alloc.h \
include/grpc/support/atm_gcc_atomic.h \
include/grpc/support/atm_gcc_sync.h \
include/grpc/support/atm.h \
include/grpc/support/atm_win32.h \
include/grpc/support/cancellable_platform.h \
include/grpc/support/cmdline.h \
include/grpc/support/histogram.h \
include/grpc/support/host_port.h \
include/grpc/support/log.h \
include/grpc/support/port_platform.h \
include/grpc/support/slice_buffer.h \
include/grpc/support/slice.h \
include/grpc/support/string.h \
include/grpc/support/sync_generic.h \
include/grpc/support/sync.h \
include/grpc/support/sync_posix.h \
include/grpc/support/sync_win32.h \
include/grpc/support/thd.h \
include/grpc/support/thd_posix.h \
include/grpc/support/thd_win32.h \
include/grpc/support/time.h \
include/grpc/support/time_posix.h \
include/grpc/support/time_win32.h \
include/grpc/support/useful.h \
LIBGPR_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBGPR_SRC))))
LIBGPR_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBGPR_SRC))))
libs/libgpr.a: $(LIBGPR_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libgpr.a $(LIBGPR_OBJS)
libs/libgpr.so.$(VERSION): $(LIBGPR_OBJS)
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) -shared -Wl,-soname,libgpr.so.0 -o libs/libgpr.so.$(VERSION) $(LIBGPR_OBJS) $(LDLIBS)
deps_libgpr: $(LIBGPR_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBGPR_DEPS)
endif
clean_libgpr:
$(E) "[CLEAN] Cleaning libgpr files"
$(Q) $(RM) $(LIBGPR_OBJS)
$(Q) $(RM) $(LIBGPR_DEPS)
$(Q) $(RM) libs/libgpr.a
$(Q) $(RM) libs/libgpr.so.$(VERSION)
LIBGRPC_SRC = \
src/core/channel/call_op_string.c \
src/core/channel/census_filter.c \
src/core/channel/channel_args.c \
src/core/channel/channel_stack.c \
Split client_channel into client_channel & child_channel. Safely managing disconnection and goaways needs reference counting at the child channel level, which was near impossible to provide with the previous embedding of the child channel in the client channel. This is a (hopefully) no-op refactoring to provide that split. The next CL in this series will actually get disconnection and goaway somewhat right. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81985590
10 years ago
src/core/channel/child_channel.c \
Initial import.
10 years ago
src/core/channel/client_channel.c \
src/core/channel/client_setup.c \
src/core/channel/connected_channel.c \
src/core/channel/http_client_filter.c \
src/core/channel/http_filter.c \
src/core/channel/http_server_filter.c \
src/core/channel/metadata_buffer.c \
src/core/channel/noop_filter.c \
src/core/compression/algorithm.c \
src/core/compression/message_compress.c \
src/core/endpoint/endpoint.c \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
src/core/endpoint/secure_endpoint.c \
src/core/httpcli/format_request.c \
src/core/httpcli/httpcli.c \
src/core/httpcli/httpcli_security_context.c \
src/core/httpcli/parser.c \
src/core/iomgr/endpoint_pair_posix.c \
src/core/iomgr/iomgr_libevent.c \
src/core/iomgr/iomgr_libevent_use_threads.c \
src/core/iomgr/resolve_address_posix.c \
src/core/iomgr/sockaddr_utils.c \
src/core/iomgr/socket_utils_common_posix.c \
src/core/iomgr/socket_utils_linux.c \
src/core/iomgr/socket_utils_posix.c \
src/core/iomgr/tcp_client_posix.c \
src/core/iomgr/tcp_posix.c \
src/core/iomgr/tcp_server_posix.c \
src/core/security/auth.c \
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
src/core/security/base64.c \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
src/core/security/credentials.c \
src/core/security/google_root_certs.c \
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
src/core/security/json_token.c \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
src/core/security/secure_transport_setup.c \
src/core/security/security_context.c \
src/core/security/server_secure_chttp2.c \
src/core/statistics/census_init.c \
src/core/statistics/census_rpc_stats.c \
src/core/statistics/census_tracing.c \
src/core/statistics/hash_table.c \
Tweaking log calls a bit. -) Introducing gpr_vlog so to spare a few vsprintf later (at least one for now) -) Renaming statistics/log.* to statistics/census_log.* to avoid collisions. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81995756
10 years ago
src/core/statistics/census_log.c \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
src/core/statistics/window_stats.c \
Initial import.
10 years ago
src/core/surface/byte_buffer.c \
src/core/surface/byte_buffer_reader.c \
src/core/surface/call.c \
src/core/surface/channel.c \
src/core/surface/channel_create.c \
src/core/surface/client.c \
src/core/surface/completion_queue.c \
src/core/surface/event_string.c \
src/core/surface/init.c \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
src/core/surface/lame_client.c \
src/core/surface/secure_channel_create.c \
src/core/surface/secure_server_create.c \
Initial import.
10 years ago
src/core/surface/server.c \
src/core/surface/server_chttp2.c \
src/core/surface/server_create.c \
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
src/core/transport/chttp2/alpn.c \
src/core/transport/chttp2/bin_encoder.c \
Initial import.
10 years ago
src/core/transport/chttp2/frame_data.c \
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
src/core/transport/chttp2/frame_goaway.c \
Initial import.
10 years ago
src/core/transport/chttp2/frame_ping.c \
src/core/transport/chttp2/frame_rst_stream.c \
src/core/transport/chttp2/frame_settings.c \
src/core/transport/chttp2/frame_window_update.c \
src/core/transport/chttp2/hpack_parser.c \
src/core/transport/chttp2/hpack_table.c \
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
src/core/transport/chttp2/huffsyms.c \
Initial import.
10 years ago
src/core/transport/chttp2/status_conversion.c \
src/core/transport/chttp2/stream_encoder.c \
src/core/transport/chttp2/stream_map.c \
src/core/transport/chttp2/timeout_encoding.c \
src/core/transport/chttp2_transport.c \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
src/core/transport/chttp2/varint.c \
Initial import.
10 years ago
src/core/transport/metadata.c \
src/core/transport/stream_op.c \
src/core/transport/transport.c \
src/core/tsi/fake_transport_security.c \
src/core/tsi/ssl_transport_security.c \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
src/core/tsi/transport_security.c \
Initial import.
10 years ago
third_party/cJSON/cJSON.c \
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
PUBLIC_HEADERS_C += \
Initial import.
10 years ago
include/grpc/byte_buffer.h \
include/grpc/byte_buffer_reader.h \
include/grpc/grpc.h \
include/grpc/grpc_security.h \
include/grpc/status.h \
LIBGRPC_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBGRPC_SRC))))
LIBGRPC_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBGRPC_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
LIBGRPC_OBJS += $(OPENSSL_DEP)
ifeq ($(NO_SECURE),true)
libs/libgrpc.a: openssl_dep_error
else
Initial import.
10 years ago
libs/libgrpc.a: $(LIBGRPC_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libgrpc.a $(LIBGRPC_OBJS)
libs/libgrpc.so.$(VERSION): $(LIBGRPC_OBJS)
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) -shared -Wl,-soname,libgrpc.so.0 -o libs/libgrpc.so.$(VERSION) $(LIBGRPC_OBJS) $(LDLIBS) $(LDLIBS_SECURE)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_libgrpc: $(LIBGRPC_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBGRPC_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_libgrpc:
$(E) "[CLEAN] Cleaning libgrpc files"
$(Q) $(RM) $(LIBGRPC_OBJS)
$(Q) $(RM) $(LIBGRPC_DEPS)
$(Q) $(RM) libs/libgrpc.a
$(Q) $(RM) libs/libgrpc.so.$(VERSION)
LIBGRPC_TEST_UTIL_SRC = \
test/core/util/grpc_profiler.c \
Fix ipv6.c build breakage on Windows. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81910401
10 years ago
test/core/util/ipv6_posix.c \
Initial import.
10 years ago
test/core/util/parse_hexstring.c \
Add mako templates to generate MS Visual Studio projects. Current state for Windows platform: -- gpr and grpc_test_util can be compiled -- gpr_log_test and gpr_cmdline_test pass Change on 2014/12/08 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81615574
10 years ago
test/core/util/port_posix.c \
Initial import.
10 years ago
test/core/util/slice_splitter.c \
test/core/util/test_config.c \
test/core/end2end/end2end_tests.c \
Build CPP interop client in Git-on-[] as well as in []. To build client in Git-on-[]: # regenerate Makefile net/grpc/tools/buildgen/generate_projects.sh # generate .pb.h and .pb.cc protoc --cpp_out=. --grpc_out=. --plugin=protoc-gen-grpc=[]-bin/net/grpc/compiler/cpp_plugin net/grpc/cpp/test/interop/test.proto net/grpc/cpp/test/interop/empty.proto net/grpc/cpp/test/interop/messages.proto # Complile and link net/grpc/tools/build_grpc_dev.sh bins/interop_client To test against GFE/ESF: # bring up server [] build net/grpc/testing/interop:server_components_env []-bin/net/grpc/testing/interop/server_components_env --manual --rpc_port=25000 # start client /tmp/grpc-codebase/bins/interop_client --enable_ssl=true --server_port="server ssl port listening port" To test [] build against GFE/ESF: [] run net/grpc/cpp:interop_client -- --enable_ssl=true --server_port="server ssl port listening port" Change on 2014/12/10 by chenw <chenw@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81821921
10 years ago
test/core/end2end/data/server1_cert.c \
test/core/end2end/data/server1_key.c \
test/core/end2end/data/ca_cert.c \
Initial import.
10 years ago
test/core/end2end/cq_verifier.c \
test/core/endpoint/endpoint_tests.c \
test/core/transport/transport_end2end_tests.c \
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
test/core/statistics/census_log_tests.c \
Initial import.
10 years ago
LIBGRPC_TEST_UTIL_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBGRPC_TEST_UTIL_SRC))))
LIBGRPC_TEST_UTIL_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBGRPC_TEST_UTIL_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
LIBGRPC_TEST_UTIL_OBJS += $(OPENSSL_DEP)
ifeq ($(NO_SECURE),true)
libs/libgrpc_test_util.a: openssl_dep_error
else
Initial import.
10 years ago
libs/libgrpc_test_util.a: $(LIBGRPC_TEST_UTIL_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libgrpc_test_util.a $(LIBGRPC_TEST_UTIL_OBJS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_libgrpc_test_util: $(LIBGRPC_TEST_UTIL_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBGRPC_TEST_UTIL_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_libgrpc_test_util:
$(E) "[CLEAN] Cleaning libgrpc_test_util files"
$(Q) $(RM) $(LIBGRPC_TEST_UTIL_OBJS)
$(Q) $(RM) $(LIBGRPC_TEST_UTIL_DEPS)
$(Q) $(RM) libs/libgrpc_test_util.a
$(Q) $(RM) libs/libgrpc_test_util.so.$(VERSION)
LIBGRPC++_SRC = \
src/cpp/server/server.cc \
src/cpp/server/server_rpc_handler.cc \
src/cpp/server/thread_pool.cc \
src/cpp/server/async_server_context.cc \
src/cpp/server/async_server.cc \
src/cpp/server/completion_queue.cc \
src/cpp/server/server_builder.cc \
src/cpp/stream/stream_context.cc \
src/cpp/client/create_channel.cc \
src/cpp/client/channel.cc \
src/cpp/client/client_context.cc \
src/cpp/client/internal_stub.cc \
src/cpp/util/time.cc \
src/cpp/util/status.cc \
src/cpp/proto/proto_utils.cc \
src/cpp/rpc_method.cc \
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
PUBLIC_HEADERS_CXX += \
Initial import.
10 years ago
include/grpc++/channel_interface.h \
include/grpc++/async_server.h \
include/grpc++/create_channel.h \
include/grpc++/server_builder.h \
include/grpc++/thread_pool_interface.h \
include/grpc++/stream_context_interface.h \
include/grpc++/status.h \
include/grpc++/config.h \
include/grpc++/completion_queue.h \
include/grpc++/stream.h \
include/grpc++/async_server_context.h \
include/grpc++/server.h \
include/grpc++/client_context.h \
LIBGRPC++_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBGRPC++_SRC))))
LIBGRPC++_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBGRPC++_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
LIBGRPC++_OBJS += $(OPENSSL_DEP)
ifeq ($(NO_SECURE),true)
libs/libgrpc++.a: openssl_dep_error
else
Initial import.
10 years ago
libs/libgrpc++.a: $(LIBGRPC++_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libgrpc++.a $(LIBGRPC++_OBJS)
libs/libgrpc++.so.$(VERSION): $(LIBGRPC++_OBJS)
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LDXX) $(LDFLAGS) -shared -Wl,-soname,libgrpc++.so.0 -o libs/libgrpc++.so.$(VERSION) $(LIBGRPC++_OBJS) $(LDLIBS) $(LDLIBS_SECURE)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_libgrpc++: $(LIBGRPC++_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBGRPC++_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_libgrpc++:
$(E) "[CLEAN] Cleaning libgrpc++ files"
$(Q) $(RM) $(LIBGRPC++_OBJS)
$(Q) $(RM) $(LIBGRPC++_DEPS)
$(Q) $(RM) libs/libgrpc++.a
$(Q) $(RM) libs/libgrpc++.so.$(VERSION)
LIBGRPC++_TEST_UTIL_SRC = \
Build CPP interop client in Git-on-[] as well as in []. To build client in Git-on-[]: # regenerate Makefile net/grpc/tools/buildgen/generate_projects.sh # generate .pb.h and .pb.cc protoc --cpp_out=. --grpc_out=. --plugin=protoc-gen-grpc=[]-bin/net/grpc/compiler/cpp_plugin net/grpc/cpp/test/interop/test.proto net/grpc/cpp/test/interop/empty.proto net/grpc/cpp/test/interop/messages.proto # Complile and link net/grpc/tools/build_grpc_dev.sh bins/interop_client To test against GFE/ESF: # bring up server [] build net/grpc/testing/interop:server_components_env []-bin/net/grpc/testing/interop/server_components_env --manual --rpc_port=25000 # start client /tmp/grpc-codebase/bins/interop_client --enable_ssl=true --server_port="server ssl port listening port" To test [] build against GFE/ESF: [] run net/grpc/cpp:interop_client -- --enable_ssl=true --server_port="server ssl port listening port" Change on 2014/12/10 by chenw <chenw@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81821921
10 years ago
test/cpp/util/test_ssl_channel.cc \
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
gens/test/cpp/util/echo.pb.cc \
test/cpp/end2end/async_test_server.cc \
Initial import.
10 years ago
LIBGRPC++_TEST_UTIL_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBGRPC++_TEST_UTIL_SRC))))
LIBGRPC++_TEST_UTIL_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBGRPC++_TEST_UTIL_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
LIBGRPC++_TEST_UTIL_OBJS += $(OPENSSL_DEP)
ifeq ($(NO_SECURE),true)
libs/libgrpc++_test_util.a: openssl_dep_error
else
Initial import.
10 years ago
libs/libgrpc++_test_util.a: $(LIBGRPC++_TEST_UTIL_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libgrpc++_test_util.a $(LIBGRPC++_TEST_UTIL_OBJS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_libgrpc++_test_util: $(LIBGRPC++_TEST_UTIL_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBGRPC++_TEST_UTIL_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_libgrpc++_test_util:
$(E) "[CLEAN] Cleaning libgrpc++_test_util files"
$(Q) $(RM) $(LIBGRPC++_TEST_UTIL_OBJS)
$(Q) $(RM) $(LIBGRPC++_TEST_UTIL_DEPS)
$(Q) $(RM) libs/libgrpc++_test_util.a
$(Q) $(RM) libs/libgrpc++_test_util.so.$(VERSION)
LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_SRC = \
test/core/end2end/fixtures/chttp2_fake_security.c \
LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_SRC))))
LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_OBJS += $(OPENSSL_DEP)
ifeq ($(NO_SECURE),true)
libs/libend2end_fixture_chttp2_fake_security.a: openssl_dep_error
else
Initial import.
10 years ago
libs/libend2end_fixture_chttp2_fake_security.a: $(LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_fixture_chttp2_fake_security.a $(LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_OBJS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_libend2end_fixture_chttp2_fake_security: $(LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_libend2end_fixture_chttp2_fake_security:
$(E) "[CLEAN] Cleaning libend2end_fixture_chttp2_fake_security files"
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_OBJS)
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_FAKE_SECURITY_DEPS)
$(Q) $(RM) libs/libend2end_fixture_chttp2_fake_security.a
$(Q) $(RM) libs/libend2end_fixture_chttp2_fake_security.so.$(VERSION)
LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_SRC = \
test/core/end2end/fixtures/chttp2_fullstack.c \
LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_SRC))))
LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_OBJS += $(OPENSSL_DEP)
ifeq ($(NO_SECURE),true)
libs/libend2end_fixture_chttp2_fullstack.a: openssl_dep_error
else
Initial import.
10 years ago
libs/libend2end_fixture_chttp2_fullstack.a: $(LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_fixture_chttp2_fullstack.a $(LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_OBJS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_libend2end_fixture_chttp2_fullstack: $(LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_libend2end_fixture_chttp2_fullstack:
$(E) "[CLEAN] Cleaning libend2end_fixture_chttp2_fullstack files"
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_OBJS)
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_FULLSTACK_DEPS)
$(Q) $(RM) libs/libend2end_fixture_chttp2_fullstack.a
$(Q) $(RM) libs/libend2end_fixture_chttp2_fullstack.so.$(VERSION)
LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_SRC = \
test/core/end2end/fixtures/chttp2_simple_ssl_fullstack.c \
LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_SRC))))
LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_OBJS += $(OPENSSL_DEP)
ifeq ($(NO_SECURE),true)
libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a: openssl_dep_error
else
Initial import.
10 years ago
libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a: $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_OBJS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_libend2end_fixture_chttp2_simple_ssl_fullstack: $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_libend2end_fixture_chttp2_simple_ssl_fullstack:
$(E) "[CLEAN] Cleaning libend2end_fixture_chttp2_simple_ssl_fullstack files"
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_OBJS)
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_FULLSTACK_DEPS)
$(Q) $(RM) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a
$(Q) $(RM) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.so.$(VERSION)
LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SRC = \
test/core/end2end/fixtures/chttp2_simple_ssl_with_oauth2_fullstack.c \
LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SRC))))
LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_OBJS += $(OPENSSL_DEP)
ifeq ($(NO_SECURE),true)
libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a: openssl_dep_error
else
Initial import.
10 years ago
libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a: $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_OBJS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack: $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack:
$(E) "[CLEAN] Cleaning libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack files"
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_OBJS)
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_DEPS)
$(Q) $(RM) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a
$(Q) $(RM) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.so.$(VERSION)
LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_SRC = \
test/core/end2end/fixtures/chttp2_socket_pair.c \
LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_SRC))))
LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_OBJS += $(OPENSSL_DEP)
ifeq ($(NO_SECURE),true)
libs/libend2end_fixture_chttp2_socket_pair.a: openssl_dep_error
else
Initial import.
10 years ago
libs/libend2end_fixture_chttp2_socket_pair.a: $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_fixture_chttp2_socket_pair.a $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_OBJS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_libend2end_fixture_chttp2_socket_pair: $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_libend2end_fixture_chttp2_socket_pair:
$(E) "[CLEAN] Cleaning libend2end_fixture_chttp2_socket_pair files"
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_OBJS)
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_DEPS)
$(Q) $(RM) libs/libend2end_fixture_chttp2_socket_pair.a
$(Q) $(RM) libs/libend2end_fixture_chttp2_socket_pair.so.$(VERSION)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SRC = \
test/core/end2end/fixtures/chttp2_socket_pair_one_byte_at_a_time.c \
LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SRC))))
LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_OBJS += $(OPENSSL_DEP)
ifeq ($(NO_SECURE),true)
libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a: $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(AR) rcs libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_OBJS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time: $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time:
$(E) "[CLEAN] Cleaning libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time files"
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_OBJS)
$(Q) $(RM) $(LIBEND2END_FIXTURE_CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_DEPS)
$(Q) $(RM) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a
$(Q) $(RM) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.so.$(VERSION)
Initial import.
10 years ago
LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_SRC = \
test/core/end2end/tests/cancel_after_accept.c \
LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_SRC))))
LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_SRC))))
libs/libend2end_test_cancel_after_accept.a: $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_cancel_after_accept.a $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_OBJS)
deps_libend2end_test_cancel_after_accept: $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_DEPS)
endif
clean_libend2end_test_cancel_after_accept:
$(E) "[CLEAN] Cleaning libend2end_test_cancel_after_accept files"
$(Q) $(RM) $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_DEPS)
$(Q) $(RM) libs/libend2end_test_cancel_after_accept.a
$(Q) $(RM) libs/libend2end_test_cancel_after_accept.so.$(VERSION)
LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_SRC = \
test/core/end2end/tests/cancel_after_accept_and_writes_closed.c \
LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_SRC))))
LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_SRC))))
libs/libend2end_test_cancel_after_accept_and_writes_closed.a: $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_cancel_after_accept_and_writes_closed.a $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_OBJS)
deps_libend2end_test_cancel_after_accept_and_writes_closed: $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_DEPS)
endif
clean_libend2end_test_cancel_after_accept_and_writes_closed:
$(E) "[CLEAN] Cleaning libend2end_test_cancel_after_accept_and_writes_closed files"
$(Q) $(RM) $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_DEPS)
$(Q) $(RM) libs/libend2end_test_cancel_after_accept_and_writes_closed.a
$(Q) $(RM) libs/libend2end_test_cancel_after_accept_and_writes_closed.so.$(VERSION)
LIBEND2END_TEST_CANCEL_AFTER_INVOKE_SRC = \
test/core/end2end/tests/cancel_after_invoke.c \
LIBEND2END_TEST_CANCEL_AFTER_INVOKE_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_CANCEL_AFTER_INVOKE_SRC))))
LIBEND2END_TEST_CANCEL_AFTER_INVOKE_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_CANCEL_AFTER_INVOKE_SRC))))
libs/libend2end_test_cancel_after_invoke.a: $(LIBEND2END_TEST_CANCEL_AFTER_INVOKE_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_cancel_after_invoke.a $(LIBEND2END_TEST_CANCEL_AFTER_INVOKE_OBJS)
deps_libend2end_test_cancel_after_invoke: $(LIBEND2END_TEST_CANCEL_AFTER_INVOKE_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_CANCEL_AFTER_INVOKE_DEPS)
endif
clean_libend2end_test_cancel_after_invoke:
$(E) "[CLEAN] Cleaning libend2end_test_cancel_after_invoke files"
$(Q) $(RM) $(LIBEND2END_TEST_CANCEL_AFTER_INVOKE_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_CANCEL_AFTER_INVOKE_DEPS)
$(Q) $(RM) libs/libend2end_test_cancel_after_invoke.a
$(Q) $(RM) libs/libend2end_test_cancel_after_invoke.so.$(VERSION)
LIBEND2END_TEST_CANCEL_BEFORE_INVOKE_SRC = \
test/core/end2end/tests/cancel_before_invoke.c \
LIBEND2END_TEST_CANCEL_BEFORE_INVOKE_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_CANCEL_BEFORE_INVOKE_SRC))))
LIBEND2END_TEST_CANCEL_BEFORE_INVOKE_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_CANCEL_BEFORE_INVOKE_SRC))))
libs/libend2end_test_cancel_before_invoke.a: $(LIBEND2END_TEST_CANCEL_BEFORE_INVOKE_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_cancel_before_invoke.a $(LIBEND2END_TEST_CANCEL_BEFORE_INVOKE_OBJS)
deps_libend2end_test_cancel_before_invoke: $(LIBEND2END_TEST_CANCEL_BEFORE_INVOKE_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_CANCEL_BEFORE_INVOKE_DEPS)
endif
clean_libend2end_test_cancel_before_invoke:
$(E) "[CLEAN] Cleaning libend2end_test_cancel_before_invoke files"
$(Q) $(RM) $(LIBEND2END_TEST_CANCEL_BEFORE_INVOKE_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_CANCEL_BEFORE_INVOKE_DEPS)
$(Q) $(RM) libs/libend2end_test_cancel_before_invoke.a
$(Q) $(RM) libs/libend2end_test_cancel_before_invoke.so.$(VERSION)
LIBEND2END_TEST_CANCEL_IN_A_VACUUM_SRC = \
test/core/end2end/tests/cancel_in_a_vacuum.c \
LIBEND2END_TEST_CANCEL_IN_A_VACUUM_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_CANCEL_IN_A_VACUUM_SRC))))
LIBEND2END_TEST_CANCEL_IN_A_VACUUM_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_CANCEL_IN_A_VACUUM_SRC))))
libs/libend2end_test_cancel_in_a_vacuum.a: $(LIBEND2END_TEST_CANCEL_IN_A_VACUUM_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_cancel_in_a_vacuum.a $(LIBEND2END_TEST_CANCEL_IN_A_VACUUM_OBJS)
deps_libend2end_test_cancel_in_a_vacuum: $(LIBEND2END_TEST_CANCEL_IN_A_VACUUM_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_CANCEL_IN_A_VACUUM_DEPS)
endif
clean_libend2end_test_cancel_in_a_vacuum:
$(E) "[CLEAN] Cleaning libend2end_test_cancel_in_a_vacuum files"
$(Q) $(RM) $(LIBEND2END_TEST_CANCEL_IN_A_VACUUM_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_CANCEL_IN_A_VACUUM_DEPS)
$(Q) $(RM) libs/libend2end_test_cancel_in_a_vacuum.a
$(Q) $(RM) libs/libend2end_test_cancel_in_a_vacuum.so.$(VERSION)
LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_SRC = \
test/core/end2end/tests/early_server_shutdown_finishes_inflight_calls.c \
LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_SRC))))
LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_SRC))))
libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.a: $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.a $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_OBJS)
deps_libend2end_test_early_server_shutdown_finishes_inflight_calls: $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_DEPS)
endif
clean_libend2end_test_early_server_shutdown_finishes_inflight_calls:
$(E) "[CLEAN] Cleaning libend2end_test_early_server_shutdown_finishes_inflight_calls files"
$(Q) $(RM) $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_DEPS)
$(Q) $(RM) libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.a
$(Q) $(RM) libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.so.$(VERSION)
LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_SRC = \
test/core/end2end/tests/early_server_shutdown_finishes_tags.c \
LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_SRC))))
LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_SRC))))
libs/libend2end_test_early_server_shutdown_finishes_tags.a: $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_early_server_shutdown_finishes_tags.a $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_OBJS)
deps_libend2end_test_early_server_shutdown_finishes_tags: $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_DEPS)
endif
clean_libend2end_test_early_server_shutdown_finishes_tags:
$(E) "[CLEAN] Cleaning libend2end_test_early_server_shutdown_finishes_tags files"
$(Q) $(RM) $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_DEPS)
$(Q) $(RM) libs/libend2end_test_early_server_shutdown_finishes_tags.a
$(Q) $(RM) libs/libend2end_test_early_server_shutdown_finishes_tags.so.$(VERSION)
LIBEND2END_TEST_INVOKE_LARGE_REQUEST_SRC = \
test/core/end2end/tests/invoke_large_request.c \
LIBEND2END_TEST_INVOKE_LARGE_REQUEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_INVOKE_LARGE_REQUEST_SRC))))
LIBEND2END_TEST_INVOKE_LARGE_REQUEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_INVOKE_LARGE_REQUEST_SRC))))
libs/libend2end_test_invoke_large_request.a: $(LIBEND2END_TEST_INVOKE_LARGE_REQUEST_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_invoke_large_request.a $(LIBEND2END_TEST_INVOKE_LARGE_REQUEST_OBJS)
deps_libend2end_test_invoke_large_request: $(LIBEND2END_TEST_INVOKE_LARGE_REQUEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_INVOKE_LARGE_REQUEST_DEPS)
endif
clean_libend2end_test_invoke_large_request:
$(E) "[CLEAN] Cleaning libend2end_test_invoke_large_request files"
$(Q) $(RM) $(LIBEND2END_TEST_INVOKE_LARGE_REQUEST_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_INVOKE_LARGE_REQUEST_DEPS)
$(Q) $(RM) libs/libend2end_test_invoke_large_request.a
$(Q) $(RM) libs/libend2end_test_invoke_large_request.so.$(VERSION)
LIBEND2END_TEST_MAX_CONCURRENT_STREAMS_SRC = \
test/core/end2end/tests/max_concurrent_streams.c \
LIBEND2END_TEST_MAX_CONCURRENT_STREAMS_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_MAX_CONCURRENT_STREAMS_SRC))))
LIBEND2END_TEST_MAX_CONCURRENT_STREAMS_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_MAX_CONCURRENT_STREAMS_SRC))))
libs/libend2end_test_max_concurrent_streams.a: $(LIBEND2END_TEST_MAX_CONCURRENT_STREAMS_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_max_concurrent_streams.a $(LIBEND2END_TEST_MAX_CONCURRENT_STREAMS_OBJS)
deps_libend2end_test_max_concurrent_streams: $(LIBEND2END_TEST_MAX_CONCURRENT_STREAMS_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_MAX_CONCURRENT_STREAMS_DEPS)
endif
clean_libend2end_test_max_concurrent_streams:
$(E) "[CLEAN] Cleaning libend2end_test_max_concurrent_streams files"
$(Q) $(RM) $(LIBEND2END_TEST_MAX_CONCURRENT_STREAMS_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_MAX_CONCURRENT_STREAMS_DEPS)
$(Q) $(RM) libs/libend2end_test_max_concurrent_streams.a
$(Q) $(RM) libs/libend2end_test_max_concurrent_streams.so.$(VERSION)
LIBEND2END_TEST_NO_OP_SRC = \
test/core/end2end/tests/no_op.c \
LIBEND2END_TEST_NO_OP_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_NO_OP_SRC))))
LIBEND2END_TEST_NO_OP_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_NO_OP_SRC))))
libs/libend2end_test_no_op.a: $(LIBEND2END_TEST_NO_OP_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_no_op.a $(LIBEND2END_TEST_NO_OP_OBJS)
deps_libend2end_test_no_op: $(LIBEND2END_TEST_NO_OP_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_NO_OP_DEPS)
endif
clean_libend2end_test_no_op:
$(E) "[CLEAN] Cleaning libend2end_test_no_op files"
$(Q) $(RM) $(LIBEND2END_TEST_NO_OP_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_NO_OP_DEPS)
$(Q) $(RM) libs/libend2end_test_no_op.a
$(Q) $(RM) libs/libend2end_test_no_op.so.$(VERSION)
LIBEND2END_TEST_PING_PONG_STREAMING_SRC = \
test/core/end2end/tests/ping_pong_streaming.c \
LIBEND2END_TEST_PING_PONG_STREAMING_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_PING_PONG_STREAMING_SRC))))
LIBEND2END_TEST_PING_PONG_STREAMING_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_PING_PONG_STREAMING_SRC))))
libs/libend2end_test_ping_pong_streaming.a: $(LIBEND2END_TEST_PING_PONG_STREAMING_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_ping_pong_streaming.a $(LIBEND2END_TEST_PING_PONG_STREAMING_OBJS)
deps_libend2end_test_ping_pong_streaming: $(LIBEND2END_TEST_PING_PONG_STREAMING_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_PING_PONG_STREAMING_DEPS)
endif
clean_libend2end_test_ping_pong_streaming:
$(E) "[CLEAN] Cleaning libend2end_test_ping_pong_streaming files"
$(Q) $(RM) $(LIBEND2END_TEST_PING_PONG_STREAMING_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_PING_PONG_STREAMING_DEPS)
$(Q) $(RM) libs/libend2end_test_ping_pong_streaming.a
$(Q) $(RM) libs/libend2end_test_ping_pong_streaming.so.$(VERSION)
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
LIBEND2END_TEST_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_SRC = \
test/core/end2end/tests/request_response_with_binary_metadata_and_payload.c \
LIBEND2END_TEST_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_SRC))))
LIBEND2END_TEST_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_SRC))))
libs/libend2end_test_request_response_with_binary_metadata_and_payload.a: $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_OBJS)
$(E) "[AR] Creating $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(AR) rcs libs/libend2end_test_request_response_with_binary_metadata_and_payload.a $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_OBJS)
deps_libend2end_test_request_response_with_binary_metadata_and_payload: $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_DEPS)
ifneq ($(NO_DEPS),true)
-include $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_DEPS)
endif
clean_libend2end_test_request_response_with_binary_metadata_and_payload:
$(E) "[CLEAN] Cleaning libend2end_test_request_response_with_binary_metadata_and_payload files"
$(Q) $(RM) $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_DEPS)
$(Q) $(RM) libs/libend2end_test_request_response_with_binary_metadata_and_payload.a
$(Q) $(RM) libs/libend2end_test_request_response_with_binary_metadata_and_payload.so.$(VERSION)
Initial import.
10 years ago
LIBEND2END_TEST_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_SRC = \
test/core/end2end/tests/request_response_with_metadata_and_payload.c \
LIBEND2END_TEST_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_SRC))))
LIBEND2END_TEST_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_SRC))))
libs/libend2end_test_request_response_with_metadata_and_payload.a: $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_request_response_with_metadata_and_payload.a $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_OBJS)
deps_libend2end_test_request_response_with_metadata_and_payload: $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_DEPS)
endif
clean_libend2end_test_request_response_with_metadata_and_payload:
$(E) "[CLEAN] Cleaning libend2end_test_request_response_with_metadata_and_payload files"
$(Q) $(RM) $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_DEPS)
$(Q) $(RM) libs/libend2end_test_request_response_with_metadata_and_payload.a
$(Q) $(RM) libs/libend2end_test_request_response_with_metadata_and_payload.so.$(VERSION)
LIBEND2END_TEST_REQUEST_RESPONSE_WITH_PAYLOAD_SRC = \
test/core/end2end/tests/request_response_with_payload.c \
LIBEND2END_TEST_REQUEST_RESPONSE_WITH_PAYLOAD_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_PAYLOAD_SRC))))
LIBEND2END_TEST_REQUEST_RESPONSE_WITH_PAYLOAD_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_PAYLOAD_SRC))))
libs/libend2end_test_request_response_with_payload.a: $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_PAYLOAD_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_request_response_with_payload.a $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_PAYLOAD_OBJS)
deps_libend2end_test_request_response_with_payload: $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_PAYLOAD_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_PAYLOAD_DEPS)
endif
clean_libend2end_test_request_response_with_payload:
$(E) "[CLEAN] Cleaning libend2end_test_request_response_with_payload files"
$(Q) $(RM) $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_PAYLOAD_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_REQUEST_RESPONSE_WITH_PAYLOAD_DEPS)
$(Q) $(RM) libs/libend2end_test_request_response_with_payload.a
$(Q) $(RM) libs/libend2end_test_request_response_with_payload.so.$(VERSION)
LIBEND2END_TEST_SIMPLE_DELAYED_REQUEST_SRC = \
test/core/end2end/tests/simple_delayed_request.c \
LIBEND2END_TEST_SIMPLE_DELAYED_REQUEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_SIMPLE_DELAYED_REQUEST_SRC))))
LIBEND2END_TEST_SIMPLE_DELAYED_REQUEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_SIMPLE_DELAYED_REQUEST_SRC))))
libs/libend2end_test_simple_delayed_request.a: $(LIBEND2END_TEST_SIMPLE_DELAYED_REQUEST_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_simple_delayed_request.a $(LIBEND2END_TEST_SIMPLE_DELAYED_REQUEST_OBJS)
deps_libend2end_test_simple_delayed_request: $(LIBEND2END_TEST_SIMPLE_DELAYED_REQUEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_SIMPLE_DELAYED_REQUEST_DEPS)
endif
clean_libend2end_test_simple_delayed_request:
$(E) "[CLEAN] Cleaning libend2end_test_simple_delayed_request files"
$(Q) $(RM) $(LIBEND2END_TEST_SIMPLE_DELAYED_REQUEST_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_SIMPLE_DELAYED_REQUEST_DEPS)
$(Q) $(RM) libs/libend2end_test_simple_delayed_request.a
$(Q) $(RM) libs/libend2end_test_simple_delayed_request.so.$(VERSION)
LIBEND2END_TEST_SIMPLE_REQUEST_SRC = \
test/core/end2end/tests/simple_request.c \
LIBEND2END_TEST_SIMPLE_REQUEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_SIMPLE_REQUEST_SRC))))
LIBEND2END_TEST_SIMPLE_REQUEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_SIMPLE_REQUEST_SRC))))
libs/libend2end_test_simple_request.a: $(LIBEND2END_TEST_SIMPLE_REQUEST_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_simple_request.a $(LIBEND2END_TEST_SIMPLE_REQUEST_OBJS)
deps_libend2end_test_simple_request: $(LIBEND2END_TEST_SIMPLE_REQUEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_SIMPLE_REQUEST_DEPS)
endif
clean_libend2end_test_simple_request:
$(E) "[CLEAN] Cleaning libend2end_test_simple_request files"
$(Q) $(RM) $(LIBEND2END_TEST_SIMPLE_REQUEST_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_SIMPLE_REQUEST_DEPS)
$(Q) $(RM) libs/libend2end_test_simple_request.a
$(Q) $(RM) libs/libend2end_test_simple_request.so.$(VERSION)
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
LIBEND2END_TEST_THREAD_STRESS_SRC = \
test/core/end2end/tests/thread_stress.c \
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
LIBEND2END_TEST_THREAD_STRESS_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_THREAD_STRESS_SRC))))
LIBEND2END_TEST_THREAD_STRESS_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_THREAD_STRESS_SRC))))
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
libs/libend2end_test_thread_stress.a: $(LIBEND2END_TEST_THREAD_STRESS_OBJS)
Initial import.
10 years ago
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_thread_stress.a $(LIBEND2END_TEST_THREAD_STRESS_OBJS)
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
deps_libend2end_test_thread_stress: $(LIBEND2END_TEST_THREAD_STRESS_DEPS)
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
-include $(LIBEND2END_TEST_THREAD_STRESS_DEPS)
Initial import.
10 years ago
endif
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
clean_libend2end_test_thread_stress:
$(E) "[CLEAN] Cleaning libend2end_test_thread_stress files"
$(Q) $(RM) $(LIBEND2END_TEST_THREAD_STRESS_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_THREAD_STRESS_DEPS)
$(Q) $(RM) libs/libend2end_test_thread_stress.a
$(Q) $(RM) libs/libend2end_test_thread_stress.so.$(VERSION)
Initial import.
10 years ago
LIBEND2END_TEST_WRITES_DONE_HANGS_WITH_PENDING_READ_SRC = \
test/core/end2end/tests/writes_done_hangs_with_pending_read.c \
LIBEND2END_TEST_WRITES_DONE_HANGS_WITH_PENDING_READ_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_TEST_WRITES_DONE_HANGS_WITH_PENDING_READ_SRC))))
LIBEND2END_TEST_WRITES_DONE_HANGS_WITH_PENDING_READ_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_TEST_WRITES_DONE_HANGS_WITH_PENDING_READ_SRC))))
libs/libend2end_test_writes_done_hangs_with_pending_read.a: $(LIBEND2END_TEST_WRITES_DONE_HANGS_WITH_PENDING_READ_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_test_writes_done_hangs_with_pending_read.a $(LIBEND2END_TEST_WRITES_DONE_HANGS_WITH_PENDING_READ_OBJS)
deps_libend2end_test_writes_done_hangs_with_pending_read: $(LIBEND2END_TEST_WRITES_DONE_HANGS_WITH_PENDING_READ_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_TEST_WRITES_DONE_HANGS_WITH_PENDING_READ_DEPS)
endif
clean_libend2end_test_writes_done_hangs_with_pending_read:
$(E) "[CLEAN] Cleaning libend2end_test_writes_done_hangs_with_pending_read files"
$(Q) $(RM) $(LIBEND2END_TEST_WRITES_DONE_HANGS_WITH_PENDING_READ_OBJS)
$(Q) $(RM) $(LIBEND2END_TEST_WRITES_DONE_HANGS_WITH_PENDING_READ_DEPS)
$(Q) $(RM) libs/libend2end_test_writes_done_hangs_with_pending_read.a
$(Q) $(RM) libs/libend2end_test_writes_done_hangs_with_pending_read.so.$(VERSION)
LIBEND2END_CERTS_SRC = \
test/core/end2end/data/ca_cert.c \
test/core/end2end/data/server1_cert.c \
test/core/end2end/data/server1_key.c \
LIBEND2END_CERTS_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBEND2END_CERTS_SRC))))
LIBEND2END_CERTS_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBEND2END_CERTS_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
LIBEND2END_CERTS_OBJS += $(OPENSSL_DEP)
ifeq ($(NO_SECURE),true)
libs/libend2end_certs.a: openssl_dep_error
else
Initial import.
10 years ago
libs/libend2end_certs.a: $(LIBEND2END_CERTS_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libend2end_certs.a $(LIBEND2END_CERTS_OBJS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_libend2end_certs: $(LIBEND2END_CERTS_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBEND2END_CERTS_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_libend2end_certs:
$(E) "[CLEAN] Cleaning libend2end_certs files"
$(Q) $(RM) $(LIBEND2END_CERTS_OBJS)
$(Q) $(RM) $(LIBEND2END_CERTS_DEPS)
$(Q) $(RM) libs/libend2end_certs.a
$(Q) $(RM) libs/libend2end_certs.so.$(VERSION)
LIBGRPC_UNSECURE_SRC = \
src/core/channel/call_op_string.c \
src/core/channel/census_filter.c \
src/core/channel/channel_args.c \
src/core/channel/channel_stack.c \
Split client_channel into client_channel & child_channel. Safely managing disconnection and goaways needs reference counting at the child channel level, which was near impossible to provide with the previous embedding of the child channel in the client channel. This is a (hopefully) no-op refactoring to provide that split. The next CL in this series will actually get disconnection and goaway somewhat right. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81985590
10 years ago
src/core/channel/child_channel.c \
Initial import.
10 years ago
src/core/channel/client_channel.c \
src/core/channel/client_setup.c \
src/core/channel/connected_channel.c \
src/core/channel/http_client_filter.c \
src/core/channel/http_filter.c \
src/core/channel/http_server_filter.c \
src/core/channel/metadata_buffer.c \
src/core/channel/noop_filter.c \
src/core/compression/algorithm.c \
src/core/compression/message_compress.c \
src/core/endpoint/endpoint.c \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
src/core/endpoint/secure_endpoint.c \
src/core/httpcli/format_request.c \
src/core/httpcli/httpcli.c \
src/core/httpcli/httpcli_security_context.c \
src/core/httpcli/parser.c \
src/core/iomgr/endpoint_pair_posix.c \
src/core/iomgr/iomgr_libevent.c \
src/core/iomgr/iomgr_libevent_use_threads.c \
src/core/iomgr/resolve_address_posix.c \
src/core/iomgr/sockaddr_utils.c \
src/core/iomgr/socket_utils_common_posix.c \
src/core/iomgr/socket_utils_linux.c \
src/core/iomgr/socket_utils_posix.c \
src/core/iomgr/tcp_client_posix.c \
src/core/iomgr/tcp_posix.c \
src/core/iomgr/tcp_server_posix.c \
src/core/statistics/census_init.c \
src/core/statistics/census_rpc_stats.c \
src/core/statistics/census_tracing.c \
src/core/statistics/hash_table.c \
Tweaking log calls a bit. -) Introducing gpr_vlog so to spare a few vsprintf later (at least one for now) -) Renaming statistics/log.* to statistics/census_log.* to avoid collisions. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81995756
10 years ago
src/core/statistics/census_log.c \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
src/core/statistics/window_stats.c \
Initial import.
10 years ago
src/core/surface/byte_buffer.c \
src/core/surface/byte_buffer_reader.c \
src/core/surface/call.c \
src/core/surface/channel.c \
src/core/surface/channel_create.c \
src/core/surface/client.c \
src/core/surface/completion_queue.c \
src/core/surface/event_string.c \
src/core/surface/init.c \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
src/core/surface/lame_client.c \
src/core/surface/secure_channel_create.c \
src/core/surface/secure_server_create.c \
Initial import.
10 years ago
src/core/surface/server.c \
src/core/surface/server_chttp2.c \
src/core/surface/server_create.c \
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
src/core/transport/chttp2/alpn.c \
src/core/transport/chttp2/bin_encoder.c \
Initial import.
10 years ago
src/core/transport/chttp2/frame_data.c \
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
src/core/transport/chttp2/frame_goaway.c \
Initial import.
10 years ago
src/core/transport/chttp2/frame_ping.c \
src/core/transport/chttp2/frame_rst_stream.c \
src/core/transport/chttp2/frame_settings.c \
src/core/transport/chttp2/frame_window_update.c \
src/core/transport/chttp2/hpack_parser.c \
src/core/transport/chttp2/hpack_table.c \
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
src/core/transport/chttp2/huffsyms.c \
Initial import.
10 years ago
src/core/transport/chttp2/status_conversion.c \
src/core/transport/chttp2/stream_encoder.c \
src/core/transport/chttp2/stream_map.c \
src/core/transport/chttp2/timeout_encoding.c \
src/core/transport/chttp2_transport.c \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
src/core/transport/chttp2/varint.c \
Initial import.
10 years ago
src/core/transport/metadata.c \
src/core/transport/stream_op.c \
src/core/transport/transport.c \
third_party/cJSON/cJSON.c \
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
PUBLIC_HEADERS_C += \
Initial import.
10 years ago
include/grpc/byte_buffer.h \
include/grpc/byte_buffer_reader.h \
include/grpc/grpc.h \
include/grpc/grpc_security.h \
include/grpc/status.h \
LIBGRPC_UNSECURE_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LIBGRPC_UNSECURE_SRC))))
LIBGRPC_UNSECURE_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LIBGRPC_UNSECURE_SRC))))
libs/libgrpc_unsecure.a: $(LIBGRPC_UNSECURE_OBJS)
$(E) "[AR] Creating $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(AR) rcs libs/libgrpc_unsecure.a $(LIBGRPC_UNSECURE_OBJS)
libs/libgrpc_unsecure.so.$(VERSION): $(LIBGRPC_UNSECURE_OBJS)
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
$(Q) $(LD) $(LDFLAGS) -shared -Wl,-soname,libgrpc_unsecure.so.0 -o libs/libgrpc_unsecure.so.$(VERSION) $(LIBGRPC_UNSECURE_OBJS) $(LDLIBS)
Initial import.
10 years ago
deps_libgrpc_unsecure: $(LIBGRPC_UNSECURE_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LIBGRPC_UNSECURE_DEPS)
endif
clean_libgrpc_unsecure:
$(E) "[CLEAN] Cleaning libgrpc_unsecure files"
$(Q) $(RM) $(LIBGRPC_UNSECURE_OBJS)
$(Q) $(RM) $(LIBGRPC_UNSECURE_DEPS)
$(Q) $(RM) libs/libgrpc_unsecure.a
$(Q) $(RM) libs/libgrpc_unsecure.so.$(VERSION)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
# All of the test targets, and protoc plugins
Initial import.
10 years ago
GEN_HPACK_TABLES_SRC = \
src/core/transport/chttp2/gen_hpack_tables.c \
GEN_HPACK_TABLES_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GEN_HPACK_TABLES_SRC))))
GEN_HPACK_TABLES_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GEN_HPACK_TABLES_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gen_hpack_tables: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/gen_hpack_tables: $(GEN_HPACK_TABLES_OBJS) libs/libgrpc_test_util.a libs/libgpr.a libs/libgrpc.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GEN_HPACK_TABLES_OBJS) -Llibs -lgrpc_test_util -lgpr -lgrpc $(LDLIBS) $(LDLIBS_SECURE) -o bins/gen_hpack_tables
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gen_hpack_tables: $(GEN_HPACK_TABLES_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GEN_HPACK_TABLES_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gen_hpack_tables:
$(E) "[CLEAN] Cleaning gen_hpack_tables files"
$(Q) $(RM) $(GEN_HPACK_TABLES_OBJS)
$(Q) $(RM) $(GEN_HPACK_TABLES_DEPS)
$(Q) $(RM) bins/gen_hpack_tables
First pass at making the compilers open-sourcable. Change on 2014/12/10 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81829127
10 years ago
CPP_PLUGIN_SRC = \
src/compiler/cpp_plugin.cpp \
src/compiler/cpp_generator.cpp \
CPP_PLUGIN_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CPP_PLUGIN_SRC))))
CPP_PLUGIN_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CPP_PLUGIN_SRC))))
bins/cpp_plugin: $(CPP_PLUGIN_OBJS)
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
$(E) "[HOSTLD] Linking $@"
First pass at making the compilers open-sourcable. Change on 2014/12/10 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81829127
10 years ago
$(Q) mkdir -p `dirname $@`
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
$(Q) $(HOST_LDXX) $(HOST_LDFLAGS) $(CPP_PLUGIN_OBJS) -Llibs $(HOST_LDLIBSXX) $(HOST_LDLIBS) $(HOST_LDLIBS_PROTOC) -o bins/cpp_plugin
First pass at making the compilers open-sourcable. Change on 2014/12/10 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81829127
10 years ago
deps_cpp_plugin: $(CPP_PLUGIN_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
First pass at making the compilers open-sourcable. Change on 2014/12/10 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81829127
10 years ago
-include $(CPP_PLUGIN_DEPS)
endif
clean_cpp_plugin:
$(E) "[CLEAN] Cleaning cpp_plugin files"
$(Q) $(RM) $(CPP_PLUGIN_OBJS)
$(Q) $(RM) $(CPP_PLUGIN_DEPS)
$(Q) $(RM) bins/cpp_plugin
RUBY_PLUGIN_SRC = \
src/compiler/ruby_plugin.cpp \
src/compiler/ruby_generator.cpp \
RUBY_PLUGIN_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(RUBY_PLUGIN_SRC))))
RUBY_PLUGIN_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(RUBY_PLUGIN_SRC))))
bins/ruby_plugin: $(RUBY_PLUGIN_OBJS)
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
$(E) "[HOSTLD] Linking $@"
First pass at making the compilers open-sourcable. Change on 2014/12/10 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81829127
10 years ago
$(Q) mkdir -p `dirname $@`
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
$(Q) $(HOST_LDXX) $(HOST_LDFLAGS) $(RUBY_PLUGIN_OBJS) -Llibs $(HOST_LDLIBSXX) $(HOST_LDLIBS) $(HOST_LDLIBS_PROTOC) -o bins/ruby_plugin
First pass at making the compilers open-sourcable. Change on 2014/12/10 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81829127
10 years ago
deps_ruby_plugin: $(RUBY_PLUGIN_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_DEPS),true)
First pass at making the compilers open-sourcable. Change on 2014/12/10 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81829127
10 years ago
-include $(RUBY_PLUGIN_DEPS)
endif
clean_ruby_plugin:
$(E) "[CLEAN] Cleaning ruby_plugin files"
$(Q) $(RM) $(RUBY_PLUGIN_OBJS)
$(Q) $(RM) $(RUBY_PLUGIN_DEPS)
$(Q) $(RM) bins/ruby_plugin
Initial import.
10 years ago
GRPC_BYTE_BUFFER_READER_TEST_SRC = \
test/core/surface/byte_buffer_reader_test.c \
GRPC_BYTE_BUFFER_READER_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GRPC_BYTE_BUFFER_READER_TEST_SRC))))
GRPC_BYTE_BUFFER_READER_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GRPC_BYTE_BUFFER_READER_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/grpc_byte_buffer_reader_test: openssl_dep_error
else
Initial import.
10 years ago
bins/grpc_byte_buffer_reader_test: $(GRPC_BYTE_BUFFER_READER_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GRPC_BYTE_BUFFER_READER_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/grpc_byte_buffer_reader_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_grpc_byte_buffer_reader_test: $(GRPC_BYTE_BUFFER_READER_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GRPC_BYTE_BUFFER_READER_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_grpc_byte_buffer_reader_test:
$(E) "[CLEAN] Cleaning grpc_byte_buffer_reader_test files"
$(Q) $(RM) $(GRPC_BYTE_BUFFER_READER_TEST_OBJS)
$(Q) $(RM) $(GRPC_BYTE_BUFFER_READER_TEST_DEPS)
$(Q) $(RM) bins/grpc_byte_buffer_reader_test
GPR_CANCELLABLE_TEST_SRC = \
test/core/support/cancellable_test.c \
GPR_CANCELLABLE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_CANCELLABLE_TEST_SRC))))
GPR_CANCELLABLE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_CANCELLABLE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gpr_cancellable_test: openssl_dep_error
else
Add mako templates to generate MS Visual Studio projects. Current state for Windows platform: -- gpr and grpc_test_util can be compiled -- gpr_log_test and gpr_cmdline_test pass Change on 2014/12/08 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81615574
10 years ago
bins/gpr_cancellable_test: $(GPR_CANCELLABLE_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Add mako templates to generate MS Visual Studio projects. Current state for Windows platform: -- gpr and grpc_test_util can be compiled -- gpr_log_test and gpr_cmdline_test pass Change on 2014/12/08 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81615574
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GPR_CANCELLABLE_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_cancellable_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gpr_cancellable_test: $(GPR_CANCELLABLE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GPR_CANCELLABLE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gpr_cancellable_test:
$(E) "[CLEAN] Cleaning gpr_cancellable_test files"
$(Q) $(RM) $(GPR_CANCELLABLE_TEST_OBJS)
$(Q) $(RM) $(GPR_CANCELLABLE_TEST_DEPS)
$(Q) $(RM) bins/gpr_cancellable_test
GPR_LOG_TEST_SRC = \
test/core/support/log_test.c \
GPR_LOG_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_LOG_TEST_SRC))))
GPR_LOG_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_LOG_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gpr_log_test: openssl_dep_error
else
Add mako templates to generate MS Visual Studio projects. Current state for Windows platform: -- gpr and grpc_test_util can be compiled -- gpr_log_test and gpr_cmdline_test pass Change on 2014/12/08 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81615574
10 years ago
bins/gpr_log_test: $(GPR_LOG_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Add mako templates to generate MS Visual Studio projects. Current state for Windows platform: -- gpr and grpc_test_util can be compiled -- gpr_log_test and gpr_cmdline_test pass Change on 2014/12/08 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81615574
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GPR_LOG_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_log_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gpr_log_test: $(GPR_LOG_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GPR_LOG_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gpr_log_test:
$(E) "[CLEAN] Cleaning gpr_log_test files"
$(Q) $(RM) $(GPR_LOG_TEST_OBJS)
$(Q) $(RM) $(GPR_LOG_TEST_DEPS)
$(Q) $(RM) bins/gpr_log_test
Add a test for useful.h. Additionally add rotl, rotr which will be needed to implement some hashing functions shortly. Change on 2014/12/15 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82149559
10 years ago
GPR_USEFUL_TEST_SRC = \
test/core/support/useful_test.c \
GPR_USEFUL_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_USEFUL_TEST_SRC))))
GPR_USEFUL_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_USEFUL_TEST_SRC))))
ifeq ($(NO_SECURE),true)
bins/gpr_useful_test: openssl_dep_error
else
bins/gpr_useful_test: $(GPR_USEFUL_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(GPR_USEFUL_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_useful_test
endif
deps_gpr_useful_test: $(GPR_USEFUL_TEST_DEPS)
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
-include $(GPR_USEFUL_TEST_DEPS)
endif
endif
clean_gpr_useful_test:
$(E) "[CLEAN] Cleaning gpr_useful_test files"
$(Q) $(RM) $(GPR_USEFUL_TEST_OBJS)
$(Q) $(RM) $(GPR_USEFUL_TEST_DEPS)
$(Q) $(RM) bins/gpr_useful_test
Initial import.
10 years ago
GPR_CMDLINE_TEST_SRC = \
test/core/support/cmdline_test.c \
GPR_CMDLINE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_CMDLINE_TEST_SRC))))
GPR_CMDLINE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_CMDLINE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gpr_cmdline_test: openssl_dep_error
else
Add mako templates to generate MS Visual Studio projects. Current state for Windows platform: -- gpr and grpc_test_util can be compiled -- gpr_log_test and gpr_cmdline_test pass Change on 2014/12/08 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81615574
10 years ago
bins/gpr_cmdline_test: $(GPR_CMDLINE_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Add mako templates to generate MS Visual Studio projects. Current state for Windows platform: -- gpr and grpc_test_util can be compiled -- gpr_log_test and gpr_cmdline_test pass Change on 2014/12/08 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81615574
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GPR_CMDLINE_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_cmdline_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gpr_cmdline_test: $(GPR_CMDLINE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GPR_CMDLINE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gpr_cmdline_test:
$(E) "[CLEAN] Cleaning gpr_cmdline_test files"
$(Q) $(RM) $(GPR_CMDLINE_TEST_OBJS)
$(Q) $(RM) $(GPR_CMDLINE_TEST_DEPS)
$(Q) $(RM) bins/gpr_cmdline_test
GPR_HISTOGRAM_TEST_SRC = \
test/core/support/histogram_test.c \
GPR_HISTOGRAM_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_HISTOGRAM_TEST_SRC))))
GPR_HISTOGRAM_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_HISTOGRAM_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gpr_histogram_test: openssl_dep_error
else
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
bins/gpr_histogram_test: $(GPR_HISTOGRAM_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GPR_HISTOGRAM_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_histogram_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gpr_histogram_test: $(GPR_HISTOGRAM_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GPR_HISTOGRAM_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gpr_histogram_test:
$(E) "[CLEAN] Cleaning gpr_histogram_test files"
$(Q) $(RM) $(GPR_HISTOGRAM_TEST_OBJS)
$(Q) $(RM) $(GPR_HISTOGRAM_TEST_DEPS)
$(Q) $(RM) bins/gpr_histogram_test
GPR_HOST_PORT_TEST_SRC = \
test/core/support/host_port_test.c \
GPR_HOST_PORT_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_HOST_PORT_TEST_SRC))))
GPR_HOST_PORT_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_HOST_PORT_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gpr_host_port_test: openssl_dep_error
else
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
bins/gpr_host_port_test: $(GPR_HOST_PORT_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GPR_HOST_PORT_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_host_port_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gpr_host_port_test: $(GPR_HOST_PORT_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GPR_HOST_PORT_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gpr_host_port_test:
$(E) "[CLEAN] Cleaning gpr_host_port_test files"
$(Q) $(RM) $(GPR_HOST_PORT_TEST_OBJS)
$(Q) $(RM) $(GPR_HOST_PORT_TEST_DEPS)
$(Q) $(RM) bins/gpr_host_port_test
GPR_SLICE_BUFFER_TEST_SRC = \
test/core/support/slice_buffer_test.c \
GPR_SLICE_BUFFER_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_SLICE_BUFFER_TEST_SRC))))
GPR_SLICE_BUFFER_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_SLICE_BUFFER_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gpr_slice_buffer_test: openssl_dep_error
else
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
bins/gpr_slice_buffer_test: $(GPR_SLICE_BUFFER_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GPR_SLICE_BUFFER_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_slice_buffer_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gpr_slice_buffer_test: $(GPR_SLICE_BUFFER_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GPR_SLICE_BUFFER_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gpr_slice_buffer_test:
$(E) "[CLEAN] Cleaning gpr_slice_buffer_test files"
$(Q) $(RM) $(GPR_SLICE_BUFFER_TEST_OBJS)
$(Q) $(RM) $(GPR_SLICE_BUFFER_TEST_DEPS)
$(Q) $(RM) bins/gpr_slice_buffer_test
GPR_SLICE_TEST_SRC = \
test/core/support/slice_test.c \
GPR_SLICE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_SLICE_TEST_SRC))))
GPR_SLICE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_SLICE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gpr_slice_test: openssl_dep_error
else
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
bins/gpr_slice_test: $(GPR_SLICE_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GPR_SLICE_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_slice_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gpr_slice_test: $(GPR_SLICE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GPR_SLICE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gpr_slice_test:
$(E) "[CLEAN] Cleaning gpr_slice_test files"
$(Q) $(RM) $(GPR_SLICE_TEST_OBJS)
$(Q) $(RM) $(GPR_SLICE_TEST_DEPS)
$(Q) $(RM) bins/gpr_slice_test
GPR_STRING_TEST_SRC = \
test/core/support/string_test.c \
GPR_STRING_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_STRING_TEST_SRC))))
GPR_STRING_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_STRING_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gpr_string_test: openssl_dep_error
else
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
bins/gpr_string_test: $(GPR_STRING_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GPR_STRING_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_string_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gpr_string_test: $(GPR_STRING_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GPR_STRING_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gpr_string_test:
$(E) "[CLEAN] Cleaning gpr_string_test files"
$(Q) $(RM) $(GPR_STRING_TEST_OBJS)
$(Q) $(RM) $(GPR_STRING_TEST_DEPS)
$(Q) $(RM) bins/gpr_string_test
GPR_SYNC_TEST_SRC = \
test/core/support/sync_test.c \
GPR_SYNC_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_SYNC_TEST_SRC))))
GPR_SYNC_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_SYNC_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gpr_sync_test: openssl_dep_error
else
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
bins/gpr_sync_test: $(GPR_SYNC_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GPR_SYNC_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_sync_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gpr_sync_test: $(GPR_SYNC_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GPR_SYNC_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gpr_sync_test:
$(E) "[CLEAN] Cleaning gpr_sync_test files"
$(Q) $(RM) $(GPR_SYNC_TEST_OBJS)
$(Q) $(RM) $(GPR_SYNC_TEST_DEPS)
$(Q) $(RM) bins/gpr_sync_test
GPR_THD_TEST_SRC = \
test/core/support/thd_test.c \
GPR_THD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_THD_TEST_SRC))))
GPR_THD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_THD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gpr_thd_test: openssl_dep_error
else
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
bins/gpr_thd_test: $(GPR_THD_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GPR_THD_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_thd_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gpr_thd_test: $(GPR_THD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GPR_THD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gpr_thd_test:
$(E) "[CLEAN] Cleaning gpr_thd_test files"
$(Q) $(RM) $(GPR_THD_TEST_OBJS)
$(Q) $(RM) $(GPR_THD_TEST_DEPS)
$(Q) $(RM) bins/gpr_thd_test
GPR_TIME_TEST_SRC = \
test/core/support/time_test.c \
GPR_TIME_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GPR_TIME_TEST_SRC))))
GPR_TIME_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GPR_TIME_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/gpr_time_test: openssl_dep_error
else
bins/gpr_time_test: $(GPR_TIME_TEST_OBJS) libs/libgrpc_test_util.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Add batch file that builds and executes all GPR tests. Removed bad format test to fix gpr_string_test. Also took care of some of the MS compiler warnings when compiling gpr_test_util. Change on 2014/12/11 by jtattermusch <jtattermusch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81900639
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GPR_TIME_TEST_OBJS) -Llibs -lgrpc_test_util -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/gpr_time_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_gpr_time_test: $(GPR_TIME_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GPR_TIME_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_gpr_time_test:
$(E) "[CLEAN] Cleaning gpr_time_test files"
$(Q) $(RM) $(GPR_TIME_TEST_OBJS)
$(Q) $(RM) $(GPR_TIME_TEST_DEPS)
$(Q) $(RM) bins/gpr_time_test
MURMUR_HASH_TEST_SRC = \
test/core/support/murmur_hash_test.c \
MURMUR_HASH_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(MURMUR_HASH_TEST_SRC))))
MURMUR_HASH_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(MURMUR_HASH_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/murmur_hash_test: openssl_dep_error
else
Initial import.
10 years ago
bins/murmur_hash_test: $(MURMUR_HASH_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(MURMUR_HASH_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/murmur_hash_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_murmur_hash_test: $(MURMUR_HASH_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(MURMUR_HASH_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_murmur_hash_test:
$(E) "[CLEAN] Cleaning murmur_hash_test files"
$(Q) $(RM) $(MURMUR_HASH_TEST_OBJS)
$(Q) $(RM) $(MURMUR_HASH_TEST_DEPS)
$(Q) $(RM) bins/murmur_hash_test
GRPC_STREAM_OP_TEST_SRC = \
test/core/transport/stream_op_test.c \
GRPC_STREAM_OP_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GRPC_STREAM_OP_TEST_SRC))))
GRPC_STREAM_OP_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GRPC_STREAM_OP_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/grpc_stream_op_test: openssl_dep_error
else
Initial import.
10 years ago
bins/grpc_stream_op_test: $(GRPC_STREAM_OP_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GRPC_STREAM_OP_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/grpc_stream_op_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_grpc_stream_op_test: $(GRPC_STREAM_OP_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GRPC_STREAM_OP_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_grpc_stream_op_test:
$(E) "[CLEAN] Cleaning grpc_stream_op_test files"
$(Q) $(RM) $(GRPC_STREAM_OP_TEST_OBJS)
$(Q) $(RM) $(GRPC_STREAM_OP_TEST_DEPS)
$(Q) $(RM) bins/grpc_stream_op_test
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
ALPN_TEST_SRC = \
test/core/transport/chttp2/alpn_test.c \
ALPN_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(ALPN_TEST_SRC))))
ALPN_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(ALPN_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/alpn_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/alpn_test: $(ALPN_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(ALPN_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/alpn_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_alpn_test: $(ALPN_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(ALPN_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_alpn_test:
$(E) "[CLEAN] Cleaning alpn_test files"
$(Q) $(RM) $(ALPN_TEST_OBJS)
$(Q) $(RM) $(ALPN_TEST_DEPS)
$(Q) $(RM) bins/alpn_test
Initial import.
10 years ago
CHTTP2_STREAM_ENCODER_TEST_SRC = \
test/core/transport/chttp2/stream_encoder_test.c \
CHTTP2_STREAM_ENCODER_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_STREAM_ENCODER_TEST_SRC))))
CHTTP2_STREAM_ENCODER_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_STREAM_ENCODER_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_stream_encoder_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_stream_encoder_test: $(CHTTP2_STREAM_ENCODER_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_STREAM_ENCODER_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_stream_encoder_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_stream_encoder_test: $(CHTTP2_STREAM_ENCODER_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_STREAM_ENCODER_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_stream_encoder_test:
$(E) "[CLEAN] Cleaning chttp2_stream_encoder_test files"
$(Q) $(RM) $(CHTTP2_STREAM_ENCODER_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_STREAM_ENCODER_TEST_DEPS)
$(Q) $(RM) bins/chttp2_stream_encoder_test
HPACK_TABLE_TEST_SRC = \
test/core/transport/chttp2/hpack_table_test.c \
HPACK_TABLE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(HPACK_TABLE_TEST_SRC))))
HPACK_TABLE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(HPACK_TABLE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/hpack_table_test: openssl_dep_error
else
Initial import.
10 years ago
bins/hpack_table_test: $(HPACK_TABLE_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(HPACK_TABLE_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/hpack_table_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_hpack_table_test: $(HPACK_TABLE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(HPACK_TABLE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_hpack_table_test:
$(E) "[CLEAN] Cleaning hpack_table_test files"
$(Q) $(RM) $(HPACK_TABLE_TEST_OBJS)
$(Q) $(RM) $(HPACK_TABLE_TEST_DEPS)
$(Q) $(RM) bins/hpack_table_test
CHTTP2_STREAM_MAP_TEST_SRC = \
test/core/transport/chttp2/stream_map_test.c \
CHTTP2_STREAM_MAP_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_STREAM_MAP_TEST_SRC))))
CHTTP2_STREAM_MAP_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_STREAM_MAP_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_stream_map_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_stream_map_test: $(CHTTP2_STREAM_MAP_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_STREAM_MAP_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_stream_map_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_stream_map_test: $(CHTTP2_STREAM_MAP_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_STREAM_MAP_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_stream_map_test:
$(E) "[CLEAN] Cleaning chttp2_stream_map_test files"
$(Q) $(RM) $(CHTTP2_STREAM_MAP_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_STREAM_MAP_TEST_DEPS)
$(Q) $(RM) bins/chttp2_stream_map_test
HPACK_PARSER_TEST_SRC = \
test/core/transport/chttp2/hpack_parser_test.c \
HPACK_PARSER_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(HPACK_PARSER_TEST_SRC))))
HPACK_PARSER_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(HPACK_PARSER_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/hpack_parser_test: openssl_dep_error
else
Initial import.
10 years ago
bins/hpack_parser_test: $(HPACK_PARSER_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(HPACK_PARSER_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/hpack_parser_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_hpack_parser_test: $(HPACK_PARSER_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(HPACK_PARSER_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_hpack_parser_test:
$(E) "[CLEAN] Cleaning hpack_parser_test files"
$(Q) $(RM) $(HPACK_PARSER_TEST_OBJS)
$(Q) $(RM) $(HPACK_PARSER_TEST_DEPS)
$(Q) $(RM) bins/hpack_parser_test
TRANSPORT_METADATA_TEST_SRC = \
test/core/transport/metadata_test.c \
TRANSPORT_METADATA_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(TRANSPORT_METADATA_TEST_SRC))))
TRANSPORT_METADATA_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(TRANSPORT_METADATA_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/transport_metadata_test: openssl_dep_error
else
Initial import.
10 years ago
bins/transport_metadata_test: $(TRANSPORT_METADATA_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(TRANSPORT_METADATA_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/transport_metadata_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_transport_metadata_test: $(TRANSPORT_METADATA_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(TRANSPORT_METADATA_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_transport_metadata_test:
$(E) "[CLEAN] Cleaning transport_metadata_test files"
$(Q) $(RM) $(TRANSPORT_METADATA_TEST_OBJS)
$(Q) $(RM) $(TRANSPORT_METADATA_TEST_DEPS)
$(Q) $(RM) bins/transport_metadata_test
CHTTP2_STATUS_CONVERSION_TEST_SRC = \
test/core/transport/chttp2/status_conversion_test.c \
CHTTP2_STATUS_CONVERSION_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_STATUS_CONVERSION_TEST_SRC))))
CHTTP2_STATUS_CONVERSION_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_STATUS_CONVERSION_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_status_conversion_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_status_conversion_test: $(CHTTP2_STATUS_CONVERSION_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_STATUS_CONVERSION_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_status_conversion_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_status_conversion_test: $(CHTTP2_STATUS_CONVERSION_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_STATUS_CONVERSION_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_status_conversion_test:
$(E) "[CLEAN] Cleaning chttp2_status_conversion_test files"
$(Q) $(RM) $(CHTTP2_STATUS_CONVERSION_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_STATUS_CONVERSION_TEST_DEPS)
$(Q) $(RM) bins/chttp2_status_conversion_test
CHTTP2_TRANSPORT_END2END_TEST_SRC = \
test/core/transport/chttp2_transport_end2end_test.c \
CHTTP2_TRANSPORT_END2END_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_TRANSPORT_END2END_TEST_SRC))))
CHTTP2_TRANSPORT_END2END_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_TRANSPORT_END2END_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_transport_end2end_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_transport_end2end_test: $(CHTTP2_TRANSPORT_END2END_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_TRANSPORT_END2END_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_transport_end2end_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_transport_end2end_test: $(CHTTP2_TRANSPORT_END2END_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_TRANSPORT_END2END_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_transport_end2end_test:
$(E) "[CLEAN] Cleaning chttp2_transport_end2end_test files"
$(Q) $(RM) $(CHTTP2_TRANSPORT_END2END_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_TRANSPORT_END2END_TEST_DEPS)
$(Q) $(RM) bins/chttp2_transport_end2end_test
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
TCP_POSIX_TEST_SRC = \
test/core/iomgr/tcp_posix_test.c \
Initial import.
10 years ago
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
TCP_POSIX_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(TCP_POSIX_TEST_SRC))))
TCP_POSIX_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(TCP_POSIX_TEST_SRC))))
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/tcp_posix_test: openssl_dep_error
else
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
bins/tcp_posix_test: $(TCP_POSIX_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
$(Q) $(LD) $(LDFLAGS) $(TCP_POSIX_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/tcp_posix_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
deps_tcp_posix_test: $(TCP_POSIX_TEST_DEPS)
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
-include $(TCP_POSIX_TEST_DEPS)
Initial import.
10 years ago
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
clean_tcp_posix_test:
$(E) "[CLEAN] Cleaning tcp_posix_test files"
$(Q) $(RM) $(TCP_POSIX_TEST_OBJS)
$(Q) $(RM) $(TCP_POSIX_TEST_DEPS)
$(Q) $(RM) bins/tcp_posix_test
Initial import.
10 years ago
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
DUALSTACK_SOCKET_TEST_SRC = \
test/core/end2end/dualstack_socket_test.c \
DUALSTACK_SOCKET_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(DUALSTACK_SOCKET_TEST_SRC))))
DUALSTACK_SOCKET_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(DUALSTACK_SOCKET_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/dualstack_socket_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/dualstack_socket_test: $(DUALSTACK_SOCKET_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(DUALSTACK_SOCKET_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/dualstack_socket_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_dualstack_socket_test: $(DUALSTACK_SOCKET_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(DUALSTACK_SOCKET_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_dualstack_socket_test:
$(E) "[CLEAN] Cleaning dualstack_socket_test files"
$(Q) $(RM) $(DUALSTACK_SOCKET_TEST_OBJS)
$(Q) $(RM) $(DUALSTACK_SOCKET_TEST_DEPS)
$(Q) $(RM) bins/dualstack_socket_test
NO_SERVER_TEST_SRC = \
test/core/end2end/no_server_test.c \
NO_SERVER_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(NO_SERVER_TEST_SRC))))
NO_SERVER_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(NO_SERVER_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/no_server_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/no_server_test: $(NO_SERVER_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(NO_SERVER_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/no_server_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_no_server_test: $(NO_SERVER_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(NO_SERVER_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_no_server_test:
$(E) "[CLEAN] Cleaning no_server_test files"
$(Q) $(RM) $(NO_SERVER_TEST_OBJS)
$(Q) $(RM) $(NO_SERVER_TEST_DEPS)
$(Q) $(RM) bins/no_server_test
Initial import.
10 years ago
RESOLVE_ADDRESS_TEST_SRC = \
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
test/core/iomgr/resolve_address_test.c \
Initial import.
10 years ago
RESOLVE_ADDRESS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(RESOLVE_ADDRESS_TEST_SRC))))
RESOLVE_ADDRESS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(RESOLVE_ADDRESS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/resolve_address_test: openssl_dep_error
else
Initial import.
10 years ago
bins/resolve_address_test: $(RESOLVE_ADDRESS_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(RESOLVE_ADDRESS_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/resolve_address_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_resolve_address_test: $(RESOLVE_ADDRESS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(RESOLVE_ADDRESS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_resolve_address_test:
$(E) "[CLEAN] Cleaning resolve_address_test files"
$(Q) $(RM) $(RESOLVE_ADDRESS_TEST_OBJS)
$(Q) $(RM) $(RESOLVE_ADDRESS_TEST_DEPS)
$(Q) $(RM) bins/resolve_address_test
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
SOCKADDR_UTILS_TEST_SRC = \
test/core/iomgr/sockaddr_utils_test.c \
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
SOCKADDR_UTILS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(SOCKADDR_UTILS_TEST_SRC))))
SOCKADDR_UTILS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(SOCKADDR_UTILS_TEST_SRC))))
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/sockaddr_utils_test: openssl_dep_error
else
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
bins/sockaddr_utils_test: $(SOCKADDR_UTILS_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
$(Q) $(LD) $(LDFLAGS) $(SOCKADDR_UTILS_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/sockaddr_utils_test
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
deps_sockaddr_utils_test: $(SOCKADDR_UTILS_TEST_DEPS)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
-include $(SOCKADDR_UTILS_TEST_DEPS)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
clean_sockaddr_utils_test:
$(E) "[CLEAN] Cleaning sockaddr_utils_test files"
$(Q) $(RM) $(SOCKADDR_UTILS_TEST_OBJS)
$(Q) $(RM) $(SOCKADDR_UTILS_TEST_DEPS)
$(Q) $(RM) bins/sockaddr_utils_test
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
TCP_SERVER_POSIX_TEST_SRC = \
test/core/iomgr/tcp_server_posix_test.c \
Initial import.
10 years ago
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
TCP_SERVER_POSIX_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(TCP_SERVER_POSIX_TEST_SRC))))
TCP_SERVER_POSIX_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(TCP_SERVER_POSIX_TEST_SRC))))
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/tcp_server_posix_test: openssl_dep_error
else
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
bins/tcp_server_posix_test: $(TCP_SERVER_POSIX_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
$(Q) $(LD) $(LDFLAGS) $(TCP_SERVER_POSIX_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/tcp_server_posix_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
deps_tcp_server_posix_test: $(TCP_SERVER_POSIX_TEST_DEPS)
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
-include $(TCP_SERVER_POSIX_TEST_DEPS)
Initial import.
10 years ago
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
clean_tcp_server_posix_test:
$(E) "[CLEAN] Cleaning tcp_server_posix_test files"
$(Q) $(RM) $(TCP_SERVER_POSIX_TEST_OBJS)
$(Q) $(RM) $(TCP_SERVER_POSIX_TEST_DEPS)
$(Q) $(RM) bins/tcp_server_posix_test
Initial import.
10 years ago
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
TCP_CLIENT_POSIX_TEST_SRC = \
test/core/iomgr/tcp_client_posix_test.c \
Initial import.
10 years ago
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
TCP_CLIENT_POSIX_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(TCP_CLIENT_POSIX_TEST_SRC))))
TCP_CLIENT_POSIX_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(TCP_CLIENT_POSIX_TEST_SRC))))
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/tcp_client_posix_test: openssl_dep_error
else
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
bins/tcp_client_posix_test: $(TCP_CLIENT_POSIX_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
$(Q) $(LD) $(LDFLAGS) $(TCP_CLIENT_POSIX_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/tcp_client_posix_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
deps_tcp_client_posix_test: $(TCP_CLIENT_POSIX_TEST_DEPS)
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
-include $(TCP_CLIENT_POSIX_TEST_DEPS)
Initial import.
10 years ago
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
Introducing iomgr. Move eventmanager and platform dependent endpoint functionality into a single library called 'iomgr'. This is primarily to prepare for a Windows port - where posix socket semantics lead to poor quality code. Mostly this is a code movement CL, with some small changes to help prepare the way for porting: - em style fd objects can only be held internally in iomgr, and own their memory - added grpc_iomgr_create_endpoint_pair() to accomodate the common pattern of creating a tcp endpoint from the output of socketpair - this will help keep our tests portable - separated em alarm interface into a separate file, as this part of event manager is needed higher up the stack - made the eventmanager bits a true singleton, simplifying API's across the stack as there's no longer a reason to carry a pointer there. Initial design document is here: https://docs.google.com/document/d/1VmafcHvvrP5kwtQkz84R5yXF7u7fW-9Pn0bkSUQHDt8/edit?disco=AAAAARNByxg Change on 2014/12/09 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81716456
10 years ago
clean_tcp_client_posix_test:
$(E) "[CLEAN] Cleaning tcp_client_posix_test files"
$(Q) $(RM) $(TCP_CLIENT_POSIX_TEST_OBJS)
$(Q) $(RM) $(TCP_CLIENT_POSIX_TEST_DEPS)
$(Q) $(RM) bins/tcp_client_posix_test
Initial import.
10 years ago
GRPC_CHANNEL_STACK_TEST_SRC = \
test/core/channel/channel_stack_test.c \
GRPC_CHANNEL_STACK_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GRPC_CHANNEL_STACK_TEST_SRC))))
GRPC_CHANNEL_STACK_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GRPC_CHANNEL_STACK_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/grpc_channel_stack_test: openssl_dep_error
else
Initial import.
10 years ago
bins/grpc_channel_stack_test: $(GRPC_CHANNEL_STACK_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GRPC_CHANNEL_STACK_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/grpc_channel_stack_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_grpc_channel_stack_test: $(GRPC_CHANNEL_STACK_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GRPC_CHANNEL_STACK_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_grpc_channel_stack_test:
$(E) "[CLEAN] Cleaning grpc_channel_stack_test files"
$(Q) $(RM) $(GRPC_CHANNEL_STACK_TEST_OBJS)
$(Q) $(RM) $(GRPC_CHANNEL_STACK_TEST_DEPS)
$(Q) $(RM) bins/grpc_channel_stack_test
METADATA_BUFFER_TEST_SRC = \
test/core/channel/metadata_buffer_test.c \
METADATA_BUFFER_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(METADATA_BUFFER_TEST_SRC))))
METADATA_BUFFER_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(METADATA_BUFFER_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/metadata_buffer_test: openssl_dep_error
else
Initial import.
10 years ago
bins/metadata_buffer_test: $(METADATA_BUFFER_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(METADATA_BUFFER_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/metadata_buffer_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_metadata_buffer_test: $(METADATA_BUFFER_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(METADATA_BUFFER_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_metadata_buffer_test:
$(E) "[CLEAN] Cleaning metadata_buffer_test files"
$(Q) $(RM) $(METADATA_BUFFER_TEST_OBJS)
$(Q) $(RM) $(METADATA_BUFFER_TEST_DEPS)
$(Q) $(RM) bins/metadata_buffer_test
GRPC_COMPLETION_QUEUE_TEST_SRC = \
test/core/surface/completion_queue_test.c \
GRPC_COMPLETION_QUEUE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GRPC_COMPLETION_QUEUE_TEST_SRC))))
GRPC_COMPLETION_QUEUE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GRPC_COMPLETION_QUEUE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/grpc_completion_queue_test: openssl_dep_error
else
Initial import.
10 years ago
bins/grpc_completion_queue_test: $(GRPC_COMPLETION_QUEUE_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GRPC_COMPLETION_QUEUE_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/grpc_completion_queue_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_grpc_completion_queue_test: $(GRPC_COMPLETION_QUEUE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GRPC_COMPLETION_QUEUE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_grpc_completion_queue_test:
$(E) "[CLEAN] Cleaning grpc_completion_queue_test files"
$(Q) $(RM) $(GRPC_COMPLETION_QUEUE_TEST_OBJS)
$(Q) $(RM) $(GRPC_COMPLETION_QUEUE_TEST_DEPS)
$(Q) $(RM) bins/grpc_completion_queue_test
GRPC_COMPLETION_QUEUE_BENCHMARK_SRC = \
test/core/surface/completion_queue_benchmark.c \
GRPC_COMPLETION_QUEUE_BENCHMARK_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GRPC_COMPLETION_QUEUE_BENCHMARK_SRC))))
GRPC_COMPLETION_QUEUE_BENCHMARK_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GRPC_COMPLETION_QUEUE_BENCHMARK_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/grpc_completion_queue_benchmark: openssl_dep_error
else
Initial import.
10 years ago
bins/grpc_completion_queue_benchmark: $(GRPC_COMPLETION_QUEUE_BENCHMARK_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GRPC_COMPLETION_QUEUE_BENCHMARK_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/grpc_completion_queue_benchmark
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_grpc_completion_queue_benchmark: $(GRPC_COMPLETION_QUEUE_BENCHMARK_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GRPC_COMPLETION_QUEUE_BENCHMARK_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_grpc_completion_queue_benchmark:
$(E) "[CLEAN] Cleaning grpc_completion_queue_benchmark files"
$(Q) $(RM) $(GRPC_COMPLETION_QUEUE_BENCHMARK_OBJS)
$(Q) $(RM) $(GRPC_COMPLETION_QUEUE_BENCHMARK_DEPS)
$(Q) $(RM) bins/grpc_completion_queue_benchmark
CENSUS_WINDOW_STATS_TEST_SRC = \
test/core/statistics/window_stats_test.c \
CENSUS_WINDOW_STATS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CENSUS_WINDOW_STATS_TEST_SRC))))
CENSUS_WINDOW_STATS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CENSUS_WINDOW_STATS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/census_window_stats_test: openssl_dep_error
else
Initial import.
10 years ago
bins/census_window_stats_test: $(CENSUS_WINDOW_STATS_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CENSUS_WINDOW_STATS_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/census_window_stats_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_census_window_stats_test: $(CENSUS_WINDOW_STATS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CENSUS_WINDOW_STATS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_census_window_stats_test:
$(E) "[CLEAN] Cleaning census_window_stats_test files"
$(Q) $(RM) $(CENSUS_WINDOW_STATS_TEST_OBJS)
$(Q) $(RM) $(CENSUS_WINDOW_STATS_TEST_DEPS)
$(Q) $(RM) bins/census_window_stats_test
CENSUS_STATISTICS_QUICK_TEST_SRC = \
test/core/statistics/quick_test.c \
CENSUS_STATISTICS_QUICK_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CENSUS_STATISTICS_QUICK_TEST_SRC))))
CENSUS_STATISTICS_QUICK_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CENSUS_STATISTICS_QUICK_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/census_statistics_quick_test: openssl_dep_error
else
Initial import.
10 years ago
bins/census_statistics_quick_test: $(CENSUS_STATISTICS_QUICK_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CENSUS_STATISTICS_QUICK_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/census_statistics_quick_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_census_statistics_quick_test: $(CENSUS_STATISTICS_QUICK_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CENSUS_STATISTICS_QUICK_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_census_statistics_quick_test:
$(E) "[CLEAN] Cleaning census_statistics_quick_test files"
$(Q) $(RM) $(CENSUS_STATISTICS_QUICK_TEST_OBJS)
$(Q) $(RM) $(CENSUS_STATISTICS_QUICK_TEST_DEPS)
$(Q) $(RM) bins/census_statistics_quick_test
Fix a bug in base-log, and add test. The tests compute usable space in the log, but do so using subtraction on unsigned values and did not correctly account for number of blocks used per-core. This could lead underflow, and an incorrect space calculation. In addition, testing showed that with current defaults, tests would fail on a system with > ~64 cores. This can be alleviated through use of a larger log. An explicit check in the log initialization has been added for this case, and the log size increased if necessary. Change on 2014/12/15 by aveitch <aveitch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82154432
10 years ago
CENSUS_STATISTICS_SMALL_LOG_TEST_SRC = \
test/core/statistics/small_log_test.c \
CENSUS_STATISTICS_SMALL_LOG_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CENSUS_STATISTICS_SMALL_LOG_TEST_SRC))))
CENSUS_STATISTICS_SMALL_LOG_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CENSUS_STATISTICS_SMALL_LOG_TEST_SRC))))
ifeq ($(NO_SECURE),true)
bins/census_statistics_small_log_test: openssl_dep_error
else
bins/census_statistics_small_log_test: $(CENSUS_STATISTICS_SMALL_LOG_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(CENSUS_STATISTICS_SMALL_LOG_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/census_statistics_small_log_test
endif
deps_census_statistics_small_log_test: $(CENSUS_STATISTICS_SMALL_LOG_TEST_DEPS)
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
-include $(CENSUS_STATISTICS_SMALL_LOG_TEST_DEPS)
endif
endif
clean_census_statistics_small_log_test:
$(E) "[CLEAN] Cleaning census_statistics_small_log_test files"
$(Q) $(RM) $(CENSUS_STATISTICS_SMALL_LOG_TEST_OBJS)
$(Q) $(RM) $(CENSUS_STATISTICS_SMALL_LOG_TEST_DEPS)
$(Q) $(RM) bins/census_statistics_small_log_test
Initial import.
10 years ago
CENSUS_STATISTICS_PERFORMANCE_TEST_SRC = \
test/core/statistics/performance_test.c \
CENSUS_STATISTICS_PERFORMANCE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CENSUS_STATISTICS_PERFORMANCE_TEST_SRC))))
CENSUS_STATISTICS_PERFORMANCE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CENSUS_STATISTICS_PERFORMANCE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/census_statistics_performance_test: openssl_dep_error
else
Initial import.
10 years ago
bins/census_statistics_performance_test: $(CENSUS_STATISTICS_PERFORMANCE_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CENSUS_STATISTICS_PERFORMANCE_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/census_statistics_performance_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_census_statistics_performance_test: $(CENSUS_STATISTICS_PERFORMANCE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CENSUS_STATISTICS_PERFORMANCE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_census_statistics_performance_test:
$(E) "[CLEAN] Cleaning census_statistics_performance_test files"
$(Q) $(RM) $(CENSUS_STATISTICS_PERFORMANCE_TEST_OBJS)
$(Q) $(RM) $(CENSUS_STATISTICS_PERFORMANCE_TEST_DEPS)
$(Q) $(RM) bins/census_statistics_performance_test
CENSUS_STATISTICS_MULTIPLE_WRITERS_TEST_SRC = \
test/core/statistics/multiple_writers_test.c \
CENSUS_STATISTICS_MULTIPLE_WRITERS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CENSUS_STATISTICS_MULTIPLE_WRITERS_TEST_SRC))))
CENSUS_STATISTICS_MULTIPLE_WRITERS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CENSUS_STATISTICS_MULTIPLE_WRITERS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/census_statistics_multiple_writers_test: openssl_dep_error
else
Initial import.
10 years ago
bins/census_statistics_multiple_writers_test: $(CENSUS_STATISTICS_MULTIPLE_WRITERS_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CENSUS_STATISTICS_MULTIPLE_WRITERS_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/census_statistics_multiple_writers_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_census_statistics_multiple_writers_test: $(CENSUS_STATISTICS_MULTIPLE_WRITERS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CENSUS_STATISTICS_MULTIPLE_WRITERS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_census_statistics_multiple_writers_test:
$(E) "[CLEAN] Cleaning census_statistics_multiple_writers_test files"
$(Q) $(RM) $(CENSUS_STATISTICS_MULTIPLE_WRITERS_TEST_OBJS)
$(Q) $(RM) $(CENSUS_STATISTICS_MULTIPLE_WRITERS_TEST_DEPS)
$(Q) $(RM) bins/census_statistics_multiple_writers_test
CENSUS_STATISTICS_MULTIPLE_WRITERS_CIRCULAR_BUFFER_TEST_SRC = \
test/core/statistics/multiple_writers_circular_buffer_test.c \
CENSUS_STATISTICS_MULTIPLE_WRITERS_CIRCULAR_BUFFER_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CENSUS_STATISTICS_MULTIPLE_WRITERS_CIRCULAR_BUFFER_TEST_SRC))))
CENSUS_STATISTICS_MULTIPLE_WRITERS_CIRCULAR_BUFFER_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CENSUS_STATISTICS_MULTIPLE_WRITERS_CIRCULAR_BUFFER_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/census_statistics_multiple_writers_circular_buffer_test: openssl_dep_error
else
Initial import.
10 years ago
bins/census_statistics_multiple_writers_circular_buffer_test: $(CENSUS_STATISTICS_MULTIPLE_WRITERS_CIRCULAR_BUFFER_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CENSUS_STATISTICS_MULTIPLE_WRITERS_CIRCULAR_BUFFER_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/census_statistics_multiple_writers_circular_buffer_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_census_statistics_multiple_writers_circular_buffer_test: $(CENSUS_STATISTICS_MULTIPLE_WRITERS_CIRCULAR_BUFFER_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CENSUS_STATISTICS_MULTIPLE_WRITERS_CIRCULAR_BUFFER_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_census_statistics_multiple_writers_circular_buffer_test:
$(E) "[CLEAN] Cleaning census_statistics_multiple_writers_circular_buffer_test files"
$(Q) $(RM) $(CENSUS_STATISTICS_MULTIPLE_WRITERS_CIRCULAR_BUFFER_TEST_OBJS)
$(Q) $(RM) $(CENSUS_STATISTICS_MULTIPLE_WRITERS_CIRCULAR_BUFFER_TEST_DEPS)
$(Q) $(RM) bins/census_statistics_multiple_writers_circular_buffer_test
CENSUS_STUB_TEST_SRC = \
test/core/statistics/census_stub_test.c \
CENSUS_STUB_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CENSUS_STUB_TEST_SRC))))
CENSUS_STUB_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CENSUS_STUB_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/census_stub_test: openssl_dep_error
else
Initial import.
10 years ago
bins/census_stub_test: $(CENSUS_STUB_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CENSUS_STUB_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/census_stub_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_census_stub_test: $(CENSUS_STUB_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CENSUS_STUB_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_census_stub_test:
$(E) "[CLEAN] Cleaning census_stub_test files"
$(Q) $(RM) $(CENSUS_STUB_TEST_OBJS)
$(Q) $(RM) $(CENSUS_STUB_TEST_DEPS)
$(Q) $(RM) bins/census_stub_test
CENSUS_HASH_TABLE_TEST_SRC = \
test/core/statistics/hash_table_test.c \
CENSUS_HASH_TABLE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CENSUS_HASH_TABLE_TEST_SRC))))
CENSUS_HASH_TABLE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CENSUS_HASH_TABLE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/census_hash_table_test: openssl_dep_error
else
Initial import.
10 years ago
bins/census_hash_table_test: $(CENSUS_HASH_TABLE_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CENSUS_HASH_TABLE_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/census_hash_table_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_census_hash_table_test: $(CENSUS_HASH_TABLE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CENSUS_HASH_TABLE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_census_hash_table_test:
$(E) "[CLEAN] Cleaning census_hash_table_test files"
$(Q) $(RM) $(CENSUS_HASH_TABLE_TEST_OBJS)
$(Q) $(RM) $(CENSUS_HASH_TABLE_TEST_DEPS)
$(Q) $(RM) bins/census_hash_table_test
FLING_SERVER_SRC = \
test/core/fling/server.c \
FLING_SERVER_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(FLING_SERVER_SRC))))
FLING_SERVER_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(FLING_SERVER_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/fling_server: openssl_dep_error
else
Initial import.
10 years ago
bins/fling_server: $(FLING_SERVER_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(FLING_SERVER_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/fling_server
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_fling_server: $(FLING_SERVER_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(FLING_SERVER_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_fling_server:
$(E) "[CLEAN] Cleaning fling_server files"
$(Q) $(RM) $(FLING_SERVER_OBJS)
$(Q) $(RM) $(FLING_SERVER_DEPS)
$(Q) $(RM) bins/fling_server
FLING_CLIENT_SRC = \
test/core/fling/client.c \
FLING_CLIENT_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(FLING_CLIENT_SRC))))
FLING_CLIENT_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(FLING_CLIENT_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/fling_client: openssl_dep_error
else
Initial import.
10 years ago
bins/fling_client: $(FLING_CLIENT_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(FLING_CLIENT_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/fling_client
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_fling_client: $(FLING_CLIENT_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(FLING_CLIENT_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_fling_client:
$(E) "[CLEAN] Cleaning fling_client files"
$(Q) $(RM) $(FLING_CLIENT_OBJS)
$(Q) $(RM) $(FLING_CLIENT_DEPS)
$(Q) $(RM) bins/fling_client
FLING_TEST_SRC = \
test/core/fling/fling_test.c \
FLING_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(FLING_TEST_SRC))))
FLING_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(FLING_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/fling_test: openssl_dep_error
else
Initial import.
10 years ago
bins/fling_test: $(FLING_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(FLING_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/fling_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_fling_test: $(FLING_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(FLING_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_fling_test:
$(E) "[CLEAN] Cleaning fling_test files"
$(Q) $(RM) $(FLING_TEST_OBJS)
$(Q) $(RM) $(FLING_TEST_DEPS)
$(Q) $(RM) bins/fling_test
ECHO_SERVER_SRC = \
test/core/echo/server.c \
ECHO_SERVER_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(ECHO_SERVER_SRC))))
ECHO_SERVER_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(ECHO_SERVER_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/echo_server: openssl_dep_error
else
Initial import.
10 years ago
bins/echo_server: $(ECHO_SERVER_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(ECHO_SERVER_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/echo_server
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_echo_server: $(ECHO_SERVER_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(ECHO_SERVER_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_echo_server:
$(E) "[CLEAN] Cleaning echo_server files"
$(Q) $(RM) $(ECHO_SERVER_OBJS)
$(Q) $(RM) $(ECHO_SERVER_DEPS)
$(Q) $(RM) bins/echo_server
ECHO_CLIENT_SRC = \
test/core/echo/client.c \
ECHO_CLIENT_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(ECHO_CLIENT_SRC))))
ECHO_CLIENT_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(ECHO_CLIENT_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/echo_client: openssl_dep_error
else
Initial import.
10 years ago
bins/echo_client: $(ECHO_CLIENT_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(ECHO_CLIENT_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/echo_client
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_echo_client: $(ECHO_CLIENT_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(ECHO_CLIENT_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_echo_client:
$(E) "[CLEAN] Cleaning echo_client files"
$(Q) $(RM) $(ECHO_CLIENT_OBJS)
$(Q) $(RM) $(ECHO_CLIENT_DEPS)
$(Q) $(RM) bins/echo_client
ECHO_TEST_SRC = \
test/core/echo/echo_test.c \
ECHO_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(ECHO_TEST_SRC))))
ECHO_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(ECHO_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/echo_test: openssl_dep_error
else
Initial import.
10 years ago
bins/echo_test: $(ECHO_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(ECHO_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/echo_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_echo_test: $(ECHO_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(ECHO_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_echo_test:
$(E) "[CLEAN] Cleaning echo_test files"
$(Q) $(RM) $(ECHO_TEST_OBJS)
$(Q) $(RM) $(ECHO_TEST_DEPS)
$(Q) $(RM) bins/echo_test
LOW_LEVEL_PING_PONG_BENCHMARK_SRC = \
test/core/network_benchmarks/low_level_ping_pong.c \
LOW_LEVEL_PING_PONG_BENCHMARK_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LOW_LEVEL_PING_PONG_BENCHMARK_SRC))))
LOW_LEVEL_PING_PONG_BENCHMARK_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LOW_LEVEL_PING_PONG_BENCHMARK_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/low_level_ping_pong_benchmark: openssl_dep_error
else
Initial import.
10 years ago
bins/low_level_ping_pong_benchmark: $(LOW_LEVEL_PING_PONG_BENCHMARK_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(LOW_LEVEL_PING_PONG_BENCHMARK_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/low_level_ping_pong_benchmark
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_low_level_ping_pong_benchmark: $(LOW_LEVEL_PING_PONG_BENCHMARK_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LOW_LEVEL_PING_PONG_BENCHMARK_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_low_level_ping_pong_benchmark:
$(E) "[CLEAN] Cleaning low_level_ping_pong_benchmark files"
$(Q) $(RM) $(LOW_LEVEL_PING_PONG_BENCHMARK_OBJS)
$(Q) $(RM) $(LOW_LEVEL_PING_PONG_BENCHMARK_DEPS)
$(Q) $(RM) bins/low_level_ping_pong_benchmark
MESSAGE_COMPRESS_TEST_SRC = \
test/core/compression/message_compress_test.c \
MESSAGE_COMPRESS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(MESSAGE_COMPRESS_TEST_SRC))))
MESSAGE_COMPRESS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(MESSAGE_COMPRESS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/message_compress_test: openssl_dep_error
else
Initial import.
10 years ago
bins/message_compress_test: $(MESSAGE_COMPRESS_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(MESSAGE_COMPRESS_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/message_compress_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_message_compress_test: $(MESSAGE_COMPRESS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(MESSAGE_COMPRESS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_message_compress_test:
$(E) "[CLEAN] Cleaning message_compress_test files"
$(Q) $(RM) $(MESSAGE_COMPRESS_TEST_OBJS)
$(Q) $(RM) $(MESSAGE_COMPRESS_TEST_DEPS)
$(Q) $(RM) bins/message_compress_test
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
BIN_ENCODER_TEST_SRC = \
test/core/transport/chttp2/bin_encoder_test.c \
BIN_ENCODER_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(BIN_ENCODER_TEST_SRC))))
BIN_ENCODER_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(BIN_ENCODER_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/bin_encoder_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/bin_encoder_test: $(BIN_ENCODER_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(BIN_ENCODER_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/bin_encoder_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_bin_encoder_test: $(BIN_ENCODER_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(BIN_ENCODER_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_bin_encoder_test:
$(E) "[CLEAN] Cleaning bin_encoder_test files"
$(Q) $(RM) $(BIN_ENCODER_TEST_OBJS)
$(Q) $(RM) $(BIN_ENCODER_TEST_DEPS)
$(Q) $(RM) bins/bin_encoder_test
Initial import.
10 years ago
SECURE_ENDPOINT_TEST_SRC = \
test/core/endpoint/secure_endpoint_test.c \
SECURE_ENDPOINT_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(SECURE_ENDPOINT_TEST_SRC))))
SECURE_ENDPOINT_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(SECURE_ENDPOINT_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/secure_endpoint_test: openssl_dep_error
else
Initial import.
10 years ago
bins/secure_endpoint_test: $(SECURE_ENDPOINT_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(SECURE_ENDPOINT_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/secure_endpoint_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_secure_endpoint_test: $(SECURE_ENDPOINT_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(SECURE_ENDPOINT_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_secure_endpoint_test:
$(E) "[CLEAN] Cleaning secure_endpoint_test files"
$(Q) $(RM) $(SECURE_ENDPOINT_TEST_OBJS)
$(Q) $(RM) $(SECURE_ENDPOINT_TEST_DEPS)
$(Q) $(RM) bins/secure_endpoint_test
HTTPCLI_FORMAT_REQUEST_TEST_SRC = \
test/core/httpcli/format_request_test.c \
HTTPCLI_FORMAT_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(HTTPCLI_FORMAT_REQUEST_TEST_SRC))))
HTTPCLI_FORMAT_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(HTTPCLI_FORMAT_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/httpcli_format_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/httpcli_format_request_test: $(HTTPCLI_FORMAT_REQUEST_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(HTTPCLI_FORMAT_REQUEST_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/httpcli_format_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_httpcli_format_request_test: $(HTTPCLI_FORMAT_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(HTTPCLI_FORMAT_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_httpcli_format_request_test:
$(E) "[CLEAN] Cleaning httpcli_format_request_test files"
$(Q) $(RM) $(HTTPCLI_FORMAT_REQUEST_TEST_OBJS)
$(Q) $(RM) $(HTTPCLI_FORMAT_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/httpcli_format_request_test
HTTPCLI_PARSER_TEST_SRC = \
test/core/httpcli/parser_test.c \
HTTPCLI_PARSER_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(HTTPCLI_PARSER_TEST_SRC))))
HTTPCLI_PARSER_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(HTTPCLI_PARSER_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/httpcli_parser_test: openssl_dep_error
else
Initial import.
10 years ago
bins/httpcli_parser_test: $(HTTPCLI_PARSER_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(HTTPCLI_PARSER_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/httpcli_parser_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_httpcli_parser_test: $(HTTPCLI_PARSER_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(HTTPCLI_PARSER_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_httpcli_parser_test:
$(E) "[CLEAN] Cleaning httpcli_parser_test files"
$(Q) $(RM) $(HTTPCLI_PARSER_TEST_OBJS)
$(Q) $(RM) $(HTTPCLI_PARSER_TEST_DEPS)
$(Q) $(RM) bins/httpcli_parser_test
HTTPCLI_TEST_SRC = \
test/core/httpcli/httpcli_test.c \
HTTPCLI_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(HTTPCLI_TEST_SRC))))
HTTPCLI_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(HTTPCLI_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/httpcli_test: openssl_dep_error
else
Initial import.
10 years ago
bins/httpcli_test: $(HTTPCLI_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(HTTPCLI_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/httpcli_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_httpcli_test: $(HTTPCLI_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(HTTPCLI_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_httpcli_test:
$(E) "[CLEAN] Cleaning httpcli_test files"
$(Q) $(RM) $(HTTPCLI_TEST_OBJS)
$(Q) $(RM) $(HTTPCLI_TEST_DEPS)
$(Q) $(RM) bins/httpcli_test
GRPC_CREDENTIALS_TEST_SRC = \
test/core/security/credentials_test.c \
GRPC_CREDENTIALS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GRPC_CREDENTIALS_TEST_SRC))))
GRPC_CREDENTIALS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GRPC_CREDENTIALS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/grpc_credentials_test: openssl_dep_error
else
Initial import.
10 years ago
bins/grpc_credentials_test: $(GRPC_CREDENTIALS_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(GRPC_CREDENTIALS_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/grpc_credentials_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_grpc_credentials_test: $(GRPC_CREDENTIALS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(GRPC_CREDENTIALS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_grpc_credentials_test:
$(E) "[CLEAN] Cleaning grpc_credentials_test files"
$(Q) $(RM) $(GRPC_CREDENTIALS_TEST_OBJS)
$(Q) $(RM) $(GRPC_CREDENTIALS_TEST_DEPS)
$(Q) $(RM) bins/grpc_credentials_test
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
GRPC_BASE64_TEST_SRC = \
test/core/security/base64_test.c \
GRPC_BASE64_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GRPC_BASE64_TEST_SRC))))
GRPC_BASE64_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GRPC_BASE64_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/grpc_base64_test: openssl_dep_error
else
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
bins/grpc_base64_test: $(GRPC_BASE64_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(GRPC_BASE64_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/grpc_base64_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
deps_grpc_base64_test: $(GRPC_BASE64_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
-include $(GRPC_BASE64_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
clean_grpc_base64_test:
$(E) "[CLEAN] Cleaning grpc_base64_test files"
$(Q) $(RM) $(GRPC_BASE64_TEST_OBJS)
$(Q) $(RM) $(GRPC_BASE64_TEST_DEPS)
$(Q) $(RM) bins/grpc_base64_test
GRPC_JSON_TOKEN_TEST_SRC = \
test/core/security/json_token_test.c \
GRPC_JSON_TOKEN_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(GRPC_JSON_TOKEN_TEST_SRC))))
GRPC_JSON_TOKEN_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(GRPC_JSON_TOKEN_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/grpc_json_token_test: openssl_dep_error
else
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
bins/grpc_json_token_test: $(GRPC_JSON_TOKEN_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(GRPC_JSON_TOKEN_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/grpc_json_token_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
deps_grpc_json_token_test: $(GRPC_JSON_TOKEN_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
-include $(GRPC_JSON_TOKEN_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Adding base utils for JWT service account workflow. OpenSSL base64 decoding is a disaster and does not support url_safe which we need for the JWT. Change on 2014/12/12 by jboeuf <jboeuf@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020601
10 years ago
clean_grpc_json_token_test:
$(E) "[CLEAN] Cleaning grpc_json_token_test files"
$(Q) $(RM) $(GRPC_JSON_TOKEN_TEST_OBJS)
$(Q) $(RM) $(GRPC_JSON_TOKEN_TEST_DEPS)
$(Q) $(RM) bins/grpc_json_token_test
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
TIMEOUT_ENCODING_TEST_SRC = \
test/core/transport/chttp2/timeout_encoding_test.c \
TIMEOUT_ENCODING_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(TIMEOUT_ENCODING_TEST_SRC))))
TIMEOUT_ENCODING_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(TIMEOUT_ENCODING_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/timeout_encoding_test: openssl_dep_error
else
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
bins/timeout_encoding_test: $(TIMEOUT_ENCODING_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(TIMEOUT_ENCODING_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/timeout_encoding_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
deps_timeout_encoding_test: $(TIMEOUT_ENCODING_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
-include $(TIMEOUT_ENCODING_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
clean_timeout_encoding_test:
$(E) "[CLEAN] Cleaning timeout_encoding_test files"
$(Q) $(RM) $(TIMEOUT_ENCODING_TEST_OBJS)
$(Q) $(RM) $(TIMEOUT_ENCODING_TEST_DEPS)
$(Q) $(RM) bins/timeout_encoding_test
FD_POSIX_TEST_SRC = \
test/core/iomgr/fd_posix_test.c \
FD_POSIX_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(FD_POSIX_TEST_SRC))))
FD_POSIX_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(FD_POSIX_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/fd_posix_test: openssl_dep_error
else
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
bins/fd_posix_test: $(FD_POSIX_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(FD_POSIX_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/fd_posix_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
deps_fd_posix_test: $(FD_POSIX_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
-include $(FD_POSIX_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
clean_fd_posix_test:
$(E) "[CLEAN] Cleaning fd_posix_test files"
$(Q) $(RM) $(FD_POSIX_TEST_OBJS)
$(Q) $(RM) $(FD_POSIX_TEST_DEPS)
$(Q) $(RM) bins/fd_posix_test
Initial import.
10 years ago
FLING_STREAM_TEST_SRC = \
test/core/fling/fling_stream_test.c \
FLING_STREAM_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(FLING_STREAM_TEST_SRC))))
FLING_STREAM_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(FLING_STREAM_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/fling_stream_test: openssl_dep_error
else
Initial import.
10 years ago
bins/fling_stream_test: $(FLING_STREAM_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(FLING_STREAM_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/fling_stream_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_fling_stream_test: $(FLING_STREAM_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(FLING_STREAM_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_fling_stream_test:
$(E) "[CLEAN] Cleaning fling_stream_test files"
$(Q) $(RM) $(FLING_STREAM_TEST_OBJS)
$(Q) $(RM) $(FLING_STREAM_TEST_DEPS)
$(Q) $(RM) bins/fling_stream_test
LAME_CLIENT_TEST_SRC = \
test/core/surface/lame_client_test.c \
LAME_CLIENT_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(LAME_CLIENT_TEST_SRC))))
LAME_CLIENT_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(LAME_CLIENT_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/lame_client_test: openssl_dep_error
else
Initial import.
10 years ago
bins/lame_client_test: $(LAME_CLIENT_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(LAME_CLIENT_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/lame_client_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_lame_client_test: $(LAME_CLIENT_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(LAME_CLIENT_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_lame_client_test:
$(E) "[CLEAN] Cleaning lame_client_test files"
$(Q) $(RM) $(LAME_CLIENT_TEST_OBJS)
$(Q) $(RM) $(LAME_CLIENT_TEST_DEPS)
$(Q) $(RM) bins/lame_client_test
THREAD_POOL_TEST_SRC = \
test/cpp/server/thread_pool_test.cc \
THREAD_POOL_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(THREAD_POOL_TEST_SRC))))
THREAD_POOL_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(THREAD_POOL_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/thread_pool_test: openssl_dep_error
else
Initial import.
10 years ago
bins/thread_pool_test: $(THREAD_POOL_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc++.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LDXX) $(LDFLAGS) $(THREAD_POOL_TEST_OBJS) $(GTEST_LIB) -Llibs -lgrpc_test_util -lgrpc++ -lgrpc -lgpr $(LDLIBSXX) $(LDLIBS) $(LDLIBS_SECURE) -o bins/thread_pool_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_thread_pool_test: $(THREAD_POOL_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(THREAD_POOL_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_thread_pool_test:
$(E) "[CLEAN] Cleaning thread_pool_test files"
$(Q) $(RM) $(THREAD_POOL_TEST_OBJS)
$(Q) $(RM) $(THREAD_POOL_TEST_DEPS)
$(Q) $(RM) bins/thread_pool_test
STATUS_TEST_SRC = \
test/cpp/util/status_test.cc \
STATUS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(STATUS_TEST_SRC))))
STATUS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(STATUS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/status_test: openssl_dep_error
else
Initial import.
10 years ago
bins/status_test: $(STATUS_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc++.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LDXX) $(LDFLAGS) $(STATUS_TEST_OBJS) $(GTEST_LIB) -Llibs -lgrpc_test_util -lgrpc++ -lgrpc -lgpr $(LDLIBSXX) $(LDLIBS) $(LDLIBS_SECURE) -o bins/status_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_status_test: $(STATUS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(STATUS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_status_test:
$(E) "[CLEAN] Cleaning status_test files"
$(Q) $(RM) $(STATUS_TEST_OBJS)
$(Q) $(RM) $(STATUS_TEST_DEPS)
$(Q) $(RM) bins/status_test
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
SYNC_CLIENT_ASYNC_SERVER_TEST_SRC = \
test/cpp/end2end/sync_client_async_server_test.cc \
SYNC_CLIENT_ASYNC_SERVER_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(SYNC_CLIENT_ASYNC_SERVER_TEST_SRC))))
SYNC_CLIENT_ASYNC_SERVER_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(SYNC_CLIENT_ASYNC_SERVER_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/sync_client_async_server_test: openssl_dep_error
else
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
bins/sync_client_async_server_test: $(SYNC_CLIENT_ASYNC_SERVER_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc++.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LDXX) $(LDFLAGS) $(SYNC_CLIENT_ASYNC_SERVER_TEST_OBJS) $(GTEST_LIB) -Llibs -lgrpc_test_util -lgrpc++ -lgrpc -lgpr $(LDLIBSXX) $(LDLIBS) $(LDLIBS_SECURE) -o bins/sync_client_async_server_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
deps_sync_client_async_server_test: $(SYNC_CLIENT_ASYNC_SERVER_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
-include $(SYNC_CLIENT_ASYNC_SERVER_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
clean_sync_client_async_server_test:
$(E) "[CLEAN] Cleaning sync_client_async_server_test files"
$(Q) $(RM) $(SYNC_CLIENT_ASYNC_SERVER_TEST_OBJS)
$(Q) $(RM) $(SYNC_CLIENT_ASYNC_SERVER_TEST_DEPS)
$(Q) $(RM) bins/sync_client_async_server_test
QPS_CLIENT_SRC = \
test/cpp/qps/qps_client.cc \
QPS_CLIENT_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(QPS_CLIENT_SRC))))
QPS_CLIENT_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(QPS_CLIENT_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/qps_client: openssl_dep_error
else
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
bins/qps_client: $(QPS_CLIENT_OBJS) libs/libgrpc_test_util.a libs/libgrpc++.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LDXX) $(LDFLAGS) $(QPS_CLIENT_OBJS) $(GTEST_LIB) -Llibs -lgrpc_test_util -lgrpc++ -lgrpc -lgpr $(LDLIBSXX) $(LDLIBS) $(LDLIBS_SECURE) -o bins/qps_client
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
deps_qps_client: $(QPS_CLIENT_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
-include $(QPS_CLIENT_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
clean_qps_client:
$(E) "[CLEAN] Cleaning qps_client files"
$(Q) $(RM) $(QPS_CLIENT_OBJS)
$(Q) $(RM) $(QPS_CLIENT_DEPS)
$(Q) $(RM) bins/qps_client
QPS_SERVER_SRC = \
test/cpp/end2end/qps_server.cc \
QPS_SERVER_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(QPS_SERVER_SRC))))
QPS_SERVER_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(QPS_SERVER_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/qps_server: openssl_dep_error
else
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
bins/qps_server: $(QPS_SERVER_OBJS) libs/libgrpc_test_util.a libs/libgrpc++.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LDXX) $(LDFLAGS) $(QPS_SERVER_OBJS) $(GTEST_LIB) -Llibs -lgrpc_test_util -lgrpc++ -lgrpc -lgpr $(LDLIBSXX) $(LDLIBS) $(LDLIBS_SECURE) -o bins/qps_server
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
deps_qps_server: $(QPS_SERVER_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
-include $(QPS_SERVER_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
clean_qps_server:
$(E) "[CLEAN] Cleaning qps_server files"
$(Q) $(RM) $(QPS_SERVER_OBJS)
$(Q) $(RM) $(QPS_SERVER_DEPS)
$(Q) $(RM) bins/qps_server
INTEROP_SERVER_SRC = \
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
gens/test/cpp/interop/empty.pb.cc \
gens/test/cpp/interop/messages.pb.cc \
gens/test/cpp/interop/test.pb.cc \
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
test/cpp/interop/server.cc \
INTEROP_SERVER_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(INTEROP_SERVER_SRC))))
INTEROP_SERVER_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(INTEROP_SERVER_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/interop_server: openssl_dep_error
else
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
bins/interop_server: $(INTEROP_SERVER_OBJS) libs/libgrpc++_test_util.a libs/libgrpc_test_util.a libs/libgrpc++.a libs/libgrpc.a libs/libgpr.a
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
$(Q) $(LDXX) $(LDFLAGS) $(INTEROP_SERVER_OBJS) $(GTEST_LIB) -Llibs -lgrpc++_test_util -lgrpc_test_util -lgrpc++ -lgrpc -lgpr $(LDLIBSXX) $(LDLIBS) $(LDLIBS_SECURE) -o bins/interop_server
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
deps_interop_server: $(INTEROP_SERVER_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
-include $(INTEROP_SERVER_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
clean_interop_server:
$(E) "[CLEAN] Cleaning interop_server files"
$(Q) $(RM) $(INTEROP_SERVER_OBJS)
$(Q) $(RM) $(INTEROP_SERVER_DEPS)
$(Q) $(RM) bins/interop_server
INTEROP_CLIENT_SRC = \
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
gens/test/cpp/interop/empty.pb.cc \
gens/test/cpp/interop/messages.pb.cc \
gens/test/cpp/interop/test.pb.cc \
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
test/cpp/interop/client.cc \
INTEROP_CLIENT_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(INTEROP_CLIENT_SRC))))
INTEROP_CLIENT_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(INTEROP_CLIENT_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/interop_client: openssl_dep_error
else
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
bins/interop_client: $(INTEROP_CLIENT_OBJS) libs/libgrpc++_test_util.a libs/libgrpc_test_util.a libs/libgrpc++.a libs/libgrpc.a libs/libgpr.a
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
Getting the generated Makefile to support .proto files. -) Compiling the proto plugin using the HOSTCC compiler. -) Set up proper rules to invoke the plugin from protoc. -) Few various renaming hacks to fully get out of []. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82000361
10 years ago
$(Q) $(LDXX) $(LDFLAGS) $(INTEROP_CLIENT_OBJS) $(GTEST_LIB) -Llibs -lgrpc++_test_util -lgrpc_test_util -lgrpc++ -lgrpc -lgpr $(LDLIBSXX) $(LDLIBS) $(LDLIBS_SECURE) -o bins/interop_client
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
deps_interop_client: $(INTEROP_CLIENT_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
-include $(INTEROP_CLIENT_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
clean_interop_client:
$(E) "[CLEAN] Cleaning interop_client files"
$(Q) $(RM) $(INTEROP_CLIENT_OBJS)
$(Q) $(RM) $(INTEROP_CLIENT_DEPS)
$(Q) $(RM) bins/interop_client
END2END_TEST_SRC = \
test/cpp/end2end/end2end_test.cc \
END2END_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(END2END_TEST_SRC))))
END2END_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(END2END_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/end2end_test: openssl_dep_error
else
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
bins/end2end_test: $(END2END_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc++.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LDXX) $(LDFLAGS) $(END2END_TEST_OBJS) $(GTEST_LIB) -Llibs -lgrpc_test_util -lgrpc++ -lgrpc -lgpr $(LDLIBSXX) $(LDLIBS) $(LDLIBS_SECURE) -o bins/end2end_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
deps_end2end_test: $(END2END_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
-include $(END2END_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
clean_end2end_test:
$(E) "[CLEAN] Cleaning end2end_test files"
$(Q) $(RM) $(END2END_TEST_OBJS)
$(Q) $(RM) $(END2END_TEST_DEPS)
$(Q) $(RM) bins/end2end_test
ALARM_TEST_SRC = \
test/core/iomgr/alarm_test.c \
ALARM_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(ALARM_TEST_SRC))))
ALARM_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(ALARM_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/alarm_test: openssl_dep_error
else
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
bins/alarm_test: $(ALARM_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(ALARM_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/alarm_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
deps_alarm_test: $(ALARM_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
-include $(ALARM_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
clean_alarm_test:
$(E) "[CLEAN] Cleaning alarm_test files"
$(Q) $(RM) $(ALARM_TEST_OBJS)
$(Q) $(RM) $(ALARM_TEST_DEPS)
$(Q) $(RM) bins/alarm_test
TIME_TEST_SRC = \
test/core/support/time_test.c \
TIME_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(TIME_TEST_SRC))))
TIME_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(TIME_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/time_test: openssl_dep_error
else
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
bins/time_test: $(TIME_TEST_OBJS) libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(TIME_TEST_OBJS) -Llibs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/time_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
deps_time_test: $(TIME_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
-include $(TIME_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Reinstate build file validity checks. Also, make those checks pass. Change on 2014/12/10 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81792364
10 years ago
clean_time_test:
$(E) "[CLEAN] Cleaning time_test files"
$(Q) $(RM) $(TIME_TEST_OBJS)
$(Q) $(RM) $(TIME_TEST_DEPS)
$(Q) $(RM) bins/time_test
Initial import.
10 years ago
CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_TEST_SRC = \
CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_TEST_SRC))))
CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_cancel_after_accept_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_cancel_after_accept_test: $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_cancel_after_accept.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_cancel_after_accept -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_cancel_after_accept_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_cancel_after_accept_test: $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_cancel_after_accept_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_cancel_after_accept_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_cancel_after_accept_test
CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC = \
CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_cancel_after_accept_and_writes_closed_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_cancel_after_accept_and_writes_closed_test: $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_cancel_after_accept_and_writes_closed.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_cancel_after_accept_and_writes_closed -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_cancel_after_accept_and_writes_closed_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_cancel_after_accept_and_writes_closed_test: $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_cancel_after_accept_and_writes_closed_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_cancel_after_accept_and_writes_closed_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_cancel_after_accept_and_writes_closed_test
CHTTP2_FAKE_SECURITY_CANCEL_AFTER_INVOKE_TEST_SRC = \
CHTTP2_FAKE_SECURITY_CANCEL_AFTER_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_INVOKE_TEST_SRC))))
CHTTP2_FAKE_SECURITY_CANCEL_AFTER_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_cancel_after_invoke_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_cancel_after_invoke_test: $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_cancel_after_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_cancel_after_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_cancel_after_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_cancel_after_invoke_test: $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_cancel_after_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_cancel_after_invoke_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_CANCEL_AFTER_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_cancel_after_invoke_test
CHTTP2_FAKE_SECURITY_CANCEL_BEFORE_INVOKE_TEST_SRC = \
CHTTP2_FAKE_SECURITY_CANCEL_BEFORE_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_CANCEL_BEFORE_INVOKE_TEST_SRC))))
CHTTP2_FAKE_SECURITY_CANCEL_BEFORE_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_CANCEL_BEFORE_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_cancel_before_invoke_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_cancel_before_invoke_test: $(CHTTP2_FAKE_SECURITY_CANCEL_BEFORE_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_cancel_before_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_CANCEL_BEFORE_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_cancel_before_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_cancel_before_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_cancel_before_invoke_test: $(CHTTP2_FAKE_SECURITY_CANCEL_BEFORE_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_CANCEL_BEFORE_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_cancel_before_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_cancel_before_invoke_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_CANCEL_BEFORE_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_CANCEL_BEFORE_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_cancel_before_invoke_test
CHTTP2_FAKE_SECURITY_CANCEL_IN_A_VACUUM_TEST_SRC = \
CHTTP2_FAKE_SECURITY_CANCEL_IN_A_VACUUM_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_CANCEL_IN_A_VACUUM_TEST_SRC))))
CHTTP2_FAKE_SECURITY_CANCEL_IN_A_VACUUM_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_CANCEL_IN_A_VACUUM_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_cancel_in_a_vacuum_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_cancel_in_a_vacuum_test: $(CHTTP2_FAKE_SECURITY_CANCEL_IN_A_VACUUM_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_cancel_in_a_vacuum.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_CANCEL_IN_A_VACUUM_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_cancel_in_a_vacuum -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_cancel_in_a_vacuum_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_cancel_in_a_vacuum_test: $(CHTTP2_FAKE_SECURITY_CANCEL_IN_A_VACUUM_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_CANCEL_IN_A_VACUUM_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_cancel_in_a_vacuum_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_cancel_in_a_vacuum_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_CANCEL_IN_A_VACUUM_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_CANCEL_IN_A_VACUUM_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_cancel_in_a_vacuum_test
CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC = \
CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_early_server_shutdown_finishes_inflight_calls -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_early_server_shutdown_finishes_inflight_calls_test
CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC = \
CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_early_server_shutdown_finishes_tags_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_early_server_shutdown_finishes_tags_test: $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_early_server_shutdown_finishes_tags.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_early_server_shutdown_finishes_tags -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_early_server_shutdown_finishes_tags_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_early_server_shutdown_finishes_tags_test: $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_early_server_shutdown_finishes_tags_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_early_server_shutdown_finishes_tags_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_early_server_shutdown_finishes_tags_test
CHTTP2_FAKE_SECURITY_INVOKE_LARGE_REQUEST_TEST_SRC = \
CHTTP2_FAKE_SECURITY_INVOKE_LARGE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_INVOKE_LARGE_REQUEST_TEST_SRC))))
CHTTP2_FAKE_SECURITY_INVOKE_LARGE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_INVOKE_LARGE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_invoke_large_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_invoke_large_request_test: $(CHTTP2_FAKE_SECURITY_INVOKE_LARGE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_invoke_large_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_INVOKE_LARGE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_invoke_large_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_invoke_large_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_invoke_large_request_test: $(CHTTP2_FAKE_SECURITY_INVOKE_LARGE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_INVOKE_LARGE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_invoke_large_request_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_invoke_large_request_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_INVOKE_LARGE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_INVOKE_LARGE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_invoke_large_request_test
CHTTP2_FAKE_SECURITY_MAX_CONCURRENT_STREAMS_TEST_SRC = \
CHTTP2_FAKE_SECURITY_MAX_CONCURRENT_STREAMS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_MAX_CONCURRENT_STREAMS_TEST_SRC))))
CHTTP2_FAKE_SECURITY_MAX_CONCURRENT_STREAMS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_MAX_CONCURRENT_STREAMS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_max_concurrent_streams_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_max_concurrent_streams_test: $(CHTTP2_FAKE_SECURITY_MAX_CONCURRENT_STREAMS_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_max_concurrent_streams.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_MAX_CONCURRENT_STREAMS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_max_concurrent_streams -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_max_concurrent_streams_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_max_concurrent_streams_test: $(CHTTP2_FAKE_SECURITY_MAX_CONCURRENT_STREAMS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_MAX_CONCURRENT_STREAMS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_max_concurrent_streams_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_max_concurrent_streams_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_MAX_CONCURRENT_STREAMS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_MAX_CONCURRENT_STREAMS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_max_concurrent_streams_test
CHTTP2_FAKE_SECURITY_NO_OP_TEST_SRC = \
CHTTP2_FAKE_SECURITY_NO_OP_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_NO_OP_TEST_SRC))))
CHTTP2_FAKE_SECURITY_NO_OP_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_NO_OP_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_no_op_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_no_op_test: $(CHTTP2_FAKE_SECURITY_NO_OP_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_no_op.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_NO_OP_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_no_op -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_no_op_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_no_op_test: $(CHTTP2_FAKE_SECURITY_NO_OP_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_NO_OP_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_no_op_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_no_op_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_NO_OP_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_NO_OP_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_no_op_test
CHTTP2_FAKE_SECURITY_PING_PONG_STREAMING_TEST_SRC = \
CHTTP2_FAKE_SECURITY_PING_PONG_STREAMING_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_PING_PONG_STREAMING_TEST_SRC))))
CHTTP2_FAKE_SECURITY_PING_PONG_STREAMING_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_PING_PONG_STREAMING_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_ping_pong_streaming_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_ping_pong_streaming_test: $(CHTTP2_FAKE_SECURITY_PING_PONG_STREAMING_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_ping_pong_streaming.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_PING_PONG_STREAMING_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_ping_pong_streaming -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_ping_pong_streaming_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_ping_pong_streaming_test: $(CHTTP2_FAKE_SECURITY_PING_PONG_STREAMING_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_PING_PONG_STREAMING_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_ping_pong_streaming_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_ping_pong_streaming_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_PING_PONG_STREAMING_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_PING_PONG_STREAMING_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_ping_pong_streaming_test
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_request_response_with_binary_metadata_and_payload_test: openssl_dep_error
else
bins/chttp2_fake_security_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_request_response_with_binary_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_request_response_with_binary_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_request_response_with_binary_metadata_and_payload_test
endif
deps_chttp2_fake_security_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
-include $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
endif
clean_chttp2_fake_security_request_response_with_binary_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_request_response_with_binary_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_request_response_with_binary_metadata_and_payload_test
Initial import.
10 years ago
CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_request_response_with_metadata_and_payload_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_request_response_with_metadata_and_payload_test: $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_request_response_with_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_request_response_with_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_request_response_with_metadata_and_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_request_response_with_metadata_and_payload_test: $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_request_response_with_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_request_response_with_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_request_response_with_metadata_and_payload_test
CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC = \
CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_request_response_with_payload_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_request_response_with_payload_test: $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_request_response_with_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_request_response_with_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_request_response_with_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_request_response_with_payload_test: $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_request_response_with_payload_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_request_response_with_payload_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_request_response_with_payload_test
CHTTP2_FAKE_SECURITY_SIMPLE_DELAYED_REQUEST_TEST_SRC = \
CHTTP2_FAKE_SECURITY_SIMPLE_DELAYED_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
CHTTP2_FAKE_SECURITY_SIMPLE_DELAYED_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_simple_delayed_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_simple_delayed_request_test: $(CHTTP2_FAKE_SECURITY_SIMPLE_DELAYED_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_simple_delayed_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_SIMPLE_DELAYED_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_simple_delayed_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_simple_delayed_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_simple_delayed_request_test: $(CHTTP2_FAKE_SECURITY_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_simple_delayed_request_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_simple_delayed_request_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_SIMPLE_DELAYED_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_simple_delayed_request_test
CHTTP2_FAKE_SECURITY_SIMPLE_REQUEST_TEST_SRC = \
CHTTP2_FAKE_SECURITY_SIMPLE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_SIMPLE_REQUEST_TEST_SRC))))
CHTTP2_FAKE_SECURITY_SIMPLE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_SIMPLE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_simple_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_simple_request_test: $(CHTTP2_FAKE_SECURITY_SIMPLE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_simple_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_SIMPLE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_simple_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_simple_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_simple_request_test: $(CHTTP2_FAKE_SECURITY_SIMPLE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_SIMPLE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_simple_request_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_simple_request_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_SIMPLE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_SIMPLE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_simple_request_test
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_FAKE_SECURITY_THREAD_STRESS_TEST_SRC = \
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_FAKE_SECURITY_THREAD_STRESS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_THREAD_STRESS_TEST_SRC))))
CHTTP2_FAKE_SECURITY_THREAD_STRESS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_THREAD_STRESS_TEST_SRC))))
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_thread_stress_test: openssl_dep_error
else
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
bins/chttp2_fake_security_thread_stress_test: $(CHTTP2_FAKE_SECURITY_THREAD_STRESS_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_thread_stress.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_THREAD_STRESS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_thread_stress -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_thread_stress_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
deps_chttp2_fake_security_thread_stress_test: $(CHTTP2_FAKE_SECURITY_THREAD_STRESS_TEST_DEPS)
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
-include $(CHTTP2_FAKE_SECURITY_THREAD_STRESS_TEST_DEPS)
Initial import.
10 years ago
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
clean_chttp2_fake_security_thread_stress_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_thread_stress_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_THREAD_STRESS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_THREAD_STRESS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_thread_stress_test
Initial import.
10 years ago
CHTTP2_FAKE_SECURITY_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC = \
CHTTP2_FAKE_SECURITY_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FAKE_SECURITY_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
CHTTP2_FAKE_SECURITY_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FAKE_SECURITY_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fake_security_writes_done_hangs_with_pending_read_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fake_security_writes_done_hangs_with_pending_read_test: $(CHTTP2_FAKE_SECURITY_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) libs/libend2end_fixture_chttp2_fake_security.a libs/libend2end_test_writes_done_hangs_with_pending_read.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FAKE_SECURITY_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fake_security -lend2end_test_writes_done_hangs_with_pending_read -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fake_security_writes_done_hangs_with_pending_read_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fake_security_writes_done_hangs_with_pending_read_test: $(CHTTP2_FAKE_SECURITY_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FAKE_SECURITY_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fake_security_writes_done_hangs_with_pending_read_test:
$(E) "[CLEAN] Cleaning chttp2_fake_security_writes_done_hangs_with_pending_read_test files"
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FAKE_SECURITY_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fake_security_writes_done_hangs_with_pending_read_test
CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_SRC = \
CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_SRC))))
CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_cancel_after_accept_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_cancel_after_accept_test: $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_cancel_after_accept.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_cancel_after_accept -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_cancel_after_accept_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_cancel_after_accept_test: $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_cancel_after_accept_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_cancel_after_accept_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_cancel_after_accept_test
CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC = \
CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_cancel_after_accept_and_writes_closed_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_cancel_after_accept_and_writes_closed_test: $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_cancel_after_accept_and_writes_closed.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_cancel_after_accept_and_writes_closed -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_cancel_after_accept_and_writes_closed_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_cancel_after_accept_and_writes_closed_test: $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_cancel_after_accept_and_writes_closed_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_cancel_after_accept_and_writes_closed_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_cancel_after_accept_and_writes_closed_test
CHTTP2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_SRC = \
CHTTP2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_SRC))))
CHTTP2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_cancel_after_invoke_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_cancel_after_invoke_test: $(CHTTP2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_cancel_after_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_cancel_after_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_cancel_after_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_cancel_after_invoke_test: $(CHTTP2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_cancel_after_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_cancel_after_invoke_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_cancel_after_invoke_test
CHTTP2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_SRC = \
CHTTP2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_SRC))))
CHTTP2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_cancel_before_invoke_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_cancel_before_invoke_test: $(CHTTP2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_cancel_before_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_cancel_before_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_cancel_before_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_cancel_before_invoke_test: $(CHTTP2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_cancel_before_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_cancel_before_invoke_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_cancel_before_invoke_test
CHTTP2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_SRC = \
CHTTP2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_SRC))))
CHTTP2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_cancel_in_a_vacuum_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_cancel_in_a_vacuum_test: $(CHTTP2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_cancel_in_a_vacuum.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_cancel_in_a_vacuum -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_cancel_in_a_vacuum_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_cancel_in_a_vacuum_test: $(CHTTP2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_cancel_in_a_vacuum_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_cancel_in_a_vacuum_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_cancel_in_a_vacuum_test
CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC = \
CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_early_server_shutdown_finishes_inflight_calls -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_early_server_shutdown_finishes_inflight_calls_test
CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC = \
CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_early_server_shutdown_finishes_tags_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_early_server_shutdown_finishes_tags_test: $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_early_server_shutdown_finishes_tags.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_early_server_shutdown_finishes_tags -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_early_server_shutdown_finishes_tags_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_early_server_shutdown_finishes_tags_test: $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_early_server_shutdown_finishes_tags_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_early_server_shutdown_finishes_tags_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_early_server_shutdown_finishes_tags_test
CHTTP2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_SRC = \
CHTTP2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_SRC))))
CHTTP2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_invoke_large_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_invoke_large_request_test: $(CHTTP2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_invoke_large_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_invoke_large_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_invoke_large_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_invoke_large_request_test: $(CHTTP2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_invoke_large_request_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_invoke_large_request_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_invoke_large_request_test
CHTTP2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_SRC = \
CHTTP2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_SRC))))
CHTTP2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_max_concurrent_streams_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_max_concurrent_streams_test: $(CHTTP2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_max_concurrent_streams.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_max_concurrent_streams -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_max_concurrent_streams_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_max_concurrent_streams_test: $(CHTTP2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_max_concurrent_streams_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_max_concurrent_streams_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_max_concurrent_streams_test
CHTTP2_FULLSTACK_NO_OP_TEST_SRC = \
CHTTP2_FULLSTACK_NO_OP_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_NO_OP_TEST_SRC))))
CHTTP2_FULLSTACK_NO_OP_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_NO_OP_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_no_op_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_no_op_test: $(CHTTP2_FULLSTACK_NO_OP_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_no_op.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_NO_OP_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_no_op -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_no_op_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_no_op_test: $(CHTTP2_FULLSTACK_NO_OP_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_NO_OP_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_no_op_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_no_op_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_NO_OP_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_NO_OP_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_no_op_test
CHTTP2_FULLSTACK_PING_PONG_STREAMING_TEST_SRC = \
CHTTP2_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_PING_PONG_STREAMING_TEST_SRC))))
CHTTP2_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_PING_PONG_STREAMING_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_ping_pong_streaming_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_ping_pong_streaming_test: $(CHTTP2_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_ping_pong_streaming.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_ping_pong_streaming -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_ping_pong_streaming_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_ping_pong_streaming_test: $(CHTTP2_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_ping_pong_streaming_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_ping_pong_streaming_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_ping_pong_streaming_test
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_request_response_with_binary_metadata_and_payload_test: openssl_dep_error
else
bins/chttp2_fullstack_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_request_response_with_binary_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_request_response_with_binary_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_request_response_with_binary_metadata_and_payload_test
endif
deps_chttp2_fullstack_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
-include $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
endif
clean_chttp2_fullstack_request_response_with_binary_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_request_response_with_binary_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_request_response_with_binary_metadata_and_payload_test
Initial import.
10 years ago
CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_request_response_with_metadata_and_payload_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_request_response_with_metadata_and_payload_test: $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_request_response_with_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_request_response_with_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_request_response_with_metadata_and_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_request_response_with_metadata_and_payload_test: $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_request_response_with_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_request_response_with_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_request_response_with_metadata_and_payload_test
CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC = \
CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_request_response_with_payload_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_request_response_with_payload_test: $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_request_response_with_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_request_response_with_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_request_response_with_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_request_response_with_payload_test: $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_request_response_with_payload_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_request_response_with_payload_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_request_response_with_payload_test
CHTTP2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_SRC = \
CHTTP2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
CHTTP2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_simple_delayed_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_simple_delayed_request_test: $(CHTTP2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_simple_delayed_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_simple_delayed_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_simple_delayed_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_simple_delayed_request_test: $(CHTTP2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_simple_delayed_request_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_simple_delayed_request_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_simple_delayed_request_test
CHTTP2_FULLSTACK_SIMPLE_REQUEST_TEST_SRC = \
CHTTP2_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_SIMPLE_REQUEST_TEST_SRC))))
CHTTP2_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_SIMPLE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_simple_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_simple_request_test: $(CHTTP2_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_simple_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_simple_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_simple_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_simple_request_test: $(CHTTP2_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_simple_request_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_simple_request_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_simple_request_test
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_FULLSTACK_THREAD_STRESS_TEST_SRC = \
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_FULLSTACK_THREAD_STRESS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_THREAD_STRESS_TEST_SRC))))
CHTTP2_FULLSTACK_THREAD_STRESS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_THREAD_STRESS_TEST_SRC))))
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_thread_stress_test: openssl_dep_error
else
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
bins/chttp2_fullstack_thread_stress_test: $(CHTTP2_FULLSTACK_THREAD_STRESS_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_thread_stress.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_THREAD_STRESS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_thread_stress -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_thread_stress_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
deps_chttp2_fullstack_thread_stress_test: $(CHTTP2_FULLSTACK_THREAD_STRESS_TEST_DEPS)
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
-include $(CHTTP2_FULLSTACK_THREAD_STRESS_TEST_DEPS)
Initial import.
10 years ago
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
clean_chttp2_fullstack_thread_stress_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_thread_stress_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_THREAD_STRESS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_THREAD_STRESS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_thread_stress_test
Initial import.
10 years ago
CHTTP2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC = \
CHTTP2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
CHTTP2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_fullstack_writes_done_hangs_with_pending_read_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_fullstack_writes_done_hangs_with_pending_read_test: $(CHTTP2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) libs/libend2end_fixture_chttp2_fullstack.a libs/libend2end_test_writes_done_hangs_with_pending_read.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_fullstack -lend2end_test_writes_done_hangs_with_pending_read -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_fullstack_writes_done_hangs_with_pending_read_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_fullstack_writes_done_hangs_with_pending_read_test: $(CHTTP2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_fullstack_writes_done_hangs_with_pending_read_test:
$(E) "[CLEAN] Cleaning chttp2_fullstack_writes_done_hangs_with_pending_read_test files"
$(Q) $(RM) $(CHTTP2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
$(Q) $(RM) bins/chttp2_fullstack_writes_done_hangs_with_pending_read_test
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_cancel_after_accept_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_cancel_after_accept_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_cancel_after_accept.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_cancel_after_accept -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_cancel_after_accept_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_cancel_after_accept_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_cancel_after_accept_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_cancel_after_accept_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_cancel_after_accept_test
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_cancel_after_accept_and_writes_closed.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_cancel_after_accept_and_writes_closed -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_cancel_after_accept_and_writes_closed_test
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_cancel_after_invoke_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_cancel_after_invoke_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_cancel_after_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_cancel_after_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_cancel_after_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_cancel_after_invoke_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_cancel_after_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_cancel_after_invoke_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_cancel_after_invoke_test
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_cancel_before_invoke_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_cancel_before_invoke_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_cancel_before_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_cancel_before_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_cancel_before_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_cancel_before_invoke_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_cancel_before_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_cancel_before_invoke_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_cancel_before_invoke_test
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_cancel_in_a_vacuum.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_cancel_in_a_vacuum -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_cancel_in_a_vacuum_test
CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_early_server_shutdown_finishes_inflight_calls -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_inflight_calls_test
CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_tags_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_tags_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_early_server_shutdown_finishes_tags.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_early_server_shutdown_finishes_tags -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_tags_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_tags_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_tags_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_tags_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_early_server_shutdown_finishes_tags_test
CHTTP2_SIMPLE_SSL_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_invoke_large_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_invoke_large_request_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_invoke_large_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_invoke_large_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_invoke_large_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_invoke_large_request_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_invoke_large_request_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_invoke_large_request_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_invoke_large_request_test
CHTTP2_SIMPLE_SSL_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_max_concurrent_streams_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_max_concurrent_streams_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_max_concurrent_streams.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_max_concurrent_streams -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_max_concurrent_streams_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_max_concurrent_streams_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_max_concurrent_streams_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_max_concurrent_streams_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_max_concurrent_streams_test
CHTTP2_SIMPLE_SSL_FULLSTACK_NO_OP_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_NO_OP_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_NO_OP_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_NO_OP_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_NO_OP_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_no_op_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_no_op_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_NO_OP_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_no_op.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_NO_OP_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_no_op -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_no_op_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_no_op_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_NO_OP_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_NO_OP_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_no_op_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_no_op_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_NO_OP_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_NO_OP_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_no_op_test
CHTTP2_SIMPLE_SSL_FULLSTACK_PING_PONG_STREAMING_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_PING_PONG_STREAMING_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_PING_PONG_STREAMING_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_ping_pong_streaming_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_ping_pong_streaming_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_ping_pong_streaming.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_ping_pong_streaming -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_ping_pong_streaming_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_ping_pong_streaming_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_ping_pong_streaming_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_ping_pong_streaming_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_ping_pong_streaming_test
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_request_response_with_binary_metadata_and_payload_test: openssl_dep_error
else
bins/chttp2_simple_ssl_fullstack_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_request_response_with_binary_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_request_response_with_binary_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_request_response_with_binary_metadata_and_payload_test
endif
deps_chttp2_simple_ssl_fullstack_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
endif
clean_chttp2_simple_ssl_fullstack_request_response_with_binary_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_request_response_with_binary_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_request_response_with_binary_metadata_and_payload_test
Initial import.
10 years ago
CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_request_response_with_metadata_and_payload_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_request_response_with_metadata_and_payload_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_request_response_with_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_request_response_with_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_request_response_with_metadata_and_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_request_response_with_metadata_and_payload_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_request_response_with_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_request_response_with_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_request_response_with_metadata_and_payload_test
CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_request_response_with_payload_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_request_response_with_payload_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_request_response_with_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_request_response_with_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_request_response_with_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_request_response_with_payload_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_request_response_with_payload_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_request_response_with_payload_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_request_response_with_payload_test
CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_simple_delayed_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_simple_delayed_request_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_simple_delayed_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_simple_delayed_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_simple_delayed_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_simple_delayed_request_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_simple_delayed_request_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_simple_delayed_request_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_simple_delayed_request_test
CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_REQUEST_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_REQUEST_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_simple_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_simple_request_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_simple_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_simple_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_simple_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_simple_request_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_simple_request_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_simple_request_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_simple_request_test
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_SIMPLE_SSL_FULLSTACK_THREAD_STRESS_TEST_SRC = \
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_SIMPLE_SSL_FULLSTACK_THREAD_STRESS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_THREAD_STRESS_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_THREAD_STRESS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_THREAD_STRESS_TEST_SRC))))
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_thread_stress_test: openssl_dep_error
else
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
bins/chttp2_simple_ssl_fullstack_thread_stress_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_THREAD_STRESS_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_thread_stress.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_THREAD_STRESS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_thread_stress -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_thread_stress_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
deps_chttp2_simple_ssl_fullstack_thread_stress_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_THREAD_STRESS_TEST_DEPS)
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_THREAD_STRESS_TEST_DEPS)
Initial import.
10 years ago
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
clean_chttp2_simple_ssl_fullstack_thread_stress_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_thread_stress_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_THREAD_STRESS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_THREAD_STRESS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_thread_stress_test
Initial import.
10 years ago
CHTTP2_SIMPLE_SSL_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC = \
CHTTP2_SIMPLE_SSL_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
CHTTP2_SIMPLE_SSL_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_fullstack_writes_done_hangs_with_pending_read_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_fullstack_writes_done_hangs_with_pending_read_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_fullstack.a libs/libend2end_test_writes_done_hangs_with_pending_read.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_fullstack -lend2end_test_writes_done_hangs_with_pending_read -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_fullstack_writes_done_hangs_with_pending_read_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_fullstack_writes_done_hangs_with_pending_read_test: $(CHTTP2_SIMPLE_SSL_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_fullstack_writes_done_hangs_with_pending_read_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_fullstack_writes_done_hangs_with_pending_read_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_fullstack_writes_done_hangs_with_pending_read_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_cancel_after_accept.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_cancel_after_accept -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_and_writes_closed_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_and_writes_closed_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_cancel_after_accept_and_writes_closed.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_cancel_after_accept_and_writes_closed -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_and_writes_closed_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_and_writes_closed_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_and_writes_closed_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_and_writes_closed_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_accept_and_writes_closed_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_invoke_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_invoke_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_cancel_after_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_cancel_after_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_invoke_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_invoke_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_AFTER_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_after_invoke_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_before_invoke_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_before_invoke_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_cancel_before_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_cancel_before_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_before_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_cancel_before_invoke_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_cancel_before_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_cancel_before_invoke_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_BEFORE_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_before_invoke_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_in_a_vacuum_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_in_a_vacuum_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_cancel_in_a_vacuum.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_cancel_in_a_vacuum -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_in_a_vacuum_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_cancel_in_a_vacuum_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_cancel_in_a_vacuum_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_cancel_in_a_vacuum_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_CANCEL_IN_A_VACUUM_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_cancel_in_a_vacuum_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_inflight_calls_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_early_server_shutdown_finishes_inflight_calls -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_inflight_calls_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_inflight_calls_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_inflight_calls_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_inflight_calls_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_tags_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_tags_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_early_server_shutdown_finishes_tags.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_early_server_shutdown_finishes_tags -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_tags_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_tags_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_tags_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_tags_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_early_server_shutdown_finishes_tags_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_invoke_large_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_invoke_large_request_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_invoke_large_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_invoke_large_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_invoke_large_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_invoke_large_request_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_invoke_large_request_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_invoke_large_request_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_INVOKE_LARGE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_invoke_large_request_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_max_concurrent_streams_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_max_concurrent_streams_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_max_concurrent_streams.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_max_concurrent_streams -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_max_concurrent_streams_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_max_concurrent_streams_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_max_concurrent_streams_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_max_concurrent_streams_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_MAX_CONCURRENT_STREAMS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_max_concurrent_streams_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_NO_OP_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_NO_OP_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_NO_OP_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_NO_OP_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_NO_OP_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_no_op_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_no_op_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_NO_OP_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_no_op.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_NO_OP_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_no_op -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_no_op_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_no_op_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_NO_OP_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_NO_OP_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_no_op_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_no_op_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_NO_OP_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_NO_OP_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_no_op_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_PING_PONG_STREAMING_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_PING_PONG_STREAMING_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_PING_PONG_STREAMING_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_ping_pong_streaming_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_ping_pong_streaming_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_ping_pong_streaming.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_ping_pong_streaming -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_ping_pong_streaming_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_ping_pong_streaming_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_ping_pong_streaming_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_ping_pong_streaming_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_PING_PONG_STREAMING_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_PING_PONG_STREAMING_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_ping_pong_streaming_test
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_binary_metadata_and_payload_test: openssl_dep_error
else
bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_request_response_with_binary_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_request_response_with_binary_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_binary_metadata_and_payload_test
endif
deps_chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
endif
clean_chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_binary_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_binary_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_binary_metadata_and_payload_test
Initial import.
10 years ago
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_metadata_and_payload_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_metadata_and_payload_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_request_response_with_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_request_response_with_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_metadata_and_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_metadata_and_payload_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_metadata_and_payload_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_payload_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_payload_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_request_response_with_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_request_response_with_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_payload_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_payload_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_payload_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_request_response_with_payload_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_simple_delayed_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_simple_delayed_request_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_simple_delayed_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_simple_delayed_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_simple_delayed_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_simple_delayed_request_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_simple_delayed_request_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_simple_delayed_request_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_simple_delayed_request_test
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_REQUEST_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_REQUEST_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_simple_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_simple_request_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_simple_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_simple_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_simple_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_simple_request_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_simple_request_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_simple_request_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_SIMPLE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_simple_request_test
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_THREAD_STRESS_TEST_SRC = \
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_THREAD_STRESS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_THREAD_STRESS_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_THREAD_STRESS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_THREAD_STRESS_TEST_SRC))))
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_thread_stress_test: openssl_dep_error
else
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_thread_stress_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_THREAD_STRESS_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_thread_stress.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_THREAD_STRESS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_thread_stress -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_thread_stress_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_thread_stress_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_THREAD_STRESS_TEST_DEPS)
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_THREAD_STRESS_TEST_DEPS)
Initial import.
10 years ago
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_thread_stress_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_thread_stress_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_THREAD_STRESS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_THREAD_STRESS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_thread_stress_test
Initial import.
10 years ago
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC = \
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_simple_ssl_with_oauth2_fullstack_writes_done_hangs_with_pending_read_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_simple_ssl_with_oauth2_fullstack_writes_done_hangs_with_pending_read_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) libs/libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack.a libs/libend2end_test_writes_done_hangs_with_pending_read.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack -lend2end_test_writes_done_hangs_with_pending_read -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_simple_ssl_with_oauth2_fullstack_writes_done_hangs_with_pending_read_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_simple_ssl_with_oauth2_fullstack_writes_done_hangs_with_pending_read_test: $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_simple_ssl_with_oauth2_fullstack_writes_done_hangs_with_pending_read_test:
$(E) "[CLEAN] Cleaning chttp2_simple_ssl_with_oauth2_fullstack_writes_done_hangs_with_pending_read_test files"
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SIMPLE_SSL_WITH_OAUTH2_FULLSTACK_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
$(Q) $(RM) bins/chttp2_simple_ssl_with_oauth2_fullstack_writes_done_hangs_with_pending_read_test
CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_TEST_SRC = \
CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_TEST_SRC))))
CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_cancel_after_accept_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_cancel_after_accept_test: $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_cancel_after_accept.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_cancel_after_accept -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_cancel_after_accept_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_cancel_after_accept_test: $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_cancel_after_accept_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_cancel_after_accept_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_cancel_after_accept_test
CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC = \
CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_cancel_after_accept_and_writes_closed_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_cancel_after_accept_and_writes_closed_test: $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_cancel_after_accept_and_writes_closed.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_cancel_after_accept_and_writes_closed -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_cancel_after_accept_and_writes_closed_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_cancel_after_accept_and_writes_closed_test: $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_cancel_after_accept_and_writes_closed_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_cancel_after_accept_and_writes_closed_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_cancel_after_accept_and_writes_closed_test
CHTTP2_SOCKET_PAIR_CANCEL_AFTER_INVOKE_TEST_SRC = \
CHTTP2_SOCKET_PAIR_CANCEL_AFTER_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_INVOKE_TEST_SRC))))
CHTTP2_SOCKET_PAIR_CANCEL_AFTER_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_cancel_after_invoke_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_cancel_after_invoke_test: $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_cancel_after_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_cancel_after_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_cancel_after_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_cancel_after_invoke_test: $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_cancel_after_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_cancel_after_invoke_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_CANCEL_AFTER_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_cancel_after_invoke_test
CHTTP2_SOCKET_PAIR_CANCEL_BEFORE_INVOKE_TEST_SRC = \
CHTTP2_SOCKET_PAIR_CANCEL_BEFORE_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_CANCEL_BEFORE_INVOKE_TEST_SRC))))
CHTTP2_SOCKET_PAIR_CANCEL_BEFORE_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_CANCEL_BEFORE_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_cancel_before_invoke_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_cancel_before_invoke_test: $(CHTTP2_SOCKET_PAIR_CANCEL_BEFORE_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_cancel_before_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_CANCEL_BEFORE_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_cancel_before_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_cancel_before_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_cancel_before_invoke_test: $(CHTTP2_SOCKET_PAIR_CANCEL_BEFORE_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_CANCEL_BEFORE_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_cancel_before_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_cancel_before_invoke_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_CANCEL_BEFORE_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_CANCEL_BEFORE_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_cancel_before_invoke_test
CHTTP2_SOCKET_PAIR_CANCEL_IN_A_VACUUM_TEST_SRC = \
CHTTP2_SOCKET_PAIR_CANCEL_IN_A_VACUUM_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_CANCEL_IN_A_VACUUM_TEST_SRC))))
CHTTP2_SOCKET_PAIR_CANCEL_IN_A_VACUUM_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_CANCEL_IN_A_VACUUM_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_cancel_in_a_vacuum_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_cancel_in_a_vacuum_test: $(CHTTP2_SOCKET_PAIR_CANCEL_IN_A_VACUUM_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_cancel_in_a_vacuum.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_CANCEL_IN_A_VACUUM_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_cancel_in_a_vacuum -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_cancel_in_a_vacuum_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_cancel_in_a_vacuum_test: $(CHTTP2_SOCKET_PAIR_CANCEL_IN_A_VACUUM_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_CANCEL_IN_A_VACUUM_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_cancel_in_a_vacuum_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_cancel_in_a_vacuum_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_CANCEL_IN_A_VACUUM_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_CANCEL_IN_A_VACUUM_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_cancel_in_a_vacuum_test
CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC = \
CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_early_server_shutdown_finishes_inflight_calls_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_early_server_shutdown_finishes_inflight_calls -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_early_server_shutdown_finishes_inflight_calls_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_early_server_shutdown_finishes_inflight_calls_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_early_server_shutdown_finishes_inflight_calls_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_early_server_shutdown_finishes_inflight_calls_test
CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC = \
CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_early_server_shutdown_finishes_tags_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_early_server_shutdown_finishes_tags_test: $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_early_server_shutdown_finishes_tags.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_early_server_shutdown_finishes_tags -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_early_server_shutdown_finishes_tags_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_early_server_shutdown_finishes_tags_test: $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_early_server_shutdown_finishes_tags_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_early_server_shutdown_finishes_tags_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_early_server_shutdown_finishes_tags_test
CHTTP2_SOCKET_PAIR_INVOKE_LARGE_REQUEST_TEST_SRC = \
CHTTP2_SOCKET_PAIR_INVOKE_LARGE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_INVOKE_LARGE_REQUEST_TEST_SRC))))
CHTTP2_SOCKET_PAIR_INVOKE_LARGE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_INVOKE_LARGE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_invoke_large_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_invoke_large_request_test: $(CHTTP2_SOCKET_PAIR_INVOKE_LARGE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_invoke_large_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_INVOKE_LARGE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_invoke_large_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_invoke_large_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_invoke_large_request_test: $(CHTTP2_SOCKET_PAIR_INVOKE_LARGE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_INVOKE_LARGE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_invoke_large_request_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_invoke_large_request_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_INVOKE_LARGE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_INVOKE_LARGE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_invoke_large_request_test
CHTTP2_SOCKET_PAIR_MAX_CONCURRENT_STREAMS_TEST_SRC = \
CHTTP2_SOCKET_PAIR_MAX_CONCURRENT_STREAMS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_MAX_CONCURRENT_STREAMS_TEST_SRC))))
CHTTP2_SOCKET_PAIR_MAX_CONCURRENT_STREAMS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_MAX_CONCURRENT_STREAMS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_max_concurrent_streams_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_max_concurrent_streams_test: $(CHTTP2_SOCKET_PAIR_MAX_CONCURRENT_STREAMS_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_max_concurrent_streams.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_MAX_CONCURRENT_STREAMS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_max_concurrent_streams -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_max_concurrent_streams_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_max_concurrent_streams_test: $(CHTTP2_SOCKET_PAIR_MAX_CONCURRENT_STREAMS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_MAX_CONCURRENT_STREAMS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_max_concurrent_streams_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_max_concurrent_streams_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_MAX_CONCURRENT_STREAMS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_MAX_CONCURRENT_STREAMS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_max_concurrent_streams_test
CHTTP2_SOCKET_PAIR_NO_OP_TEST_SRC = \
CHTTP2_SOCKET_PAIR_NO_OP_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_NO_OP_TEST_SRC))))
CHTTP2_SOCKET_PAIR_NO_OP_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_NO_OP_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_no_op_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_no_op_test: $(CHTTP2_SOCKET_PAIR_NO_OP_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_no_op.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_NO_OP_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_no_op -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_no_op_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_no_op_test: $(CHTTP2_SOCKET_PAIR_NO_OP_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_NO_OP_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_no_op_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_no_op_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_NO_OP_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_NO_OP_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_no_op_test
CHTTP2_SOCKET_PAIR_PING_PONG_STREAMING_TEST_SRC = \
CHTTP2_SOCKET_PAIR_PING_PONG_STREAMING_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_PING_PONG_STREAMING_TEST_SRC))))
CHTTP2_SOCKET_PAIR_PING_PONG_STREAMING_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_PING_PONG_STREAMING_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_ping_pong_streaming_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_ping_pong_streaming_test: $(CHTTP2_SOCKET_PAIR_PING_PONG_STREAMING_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_ping_pong_streaming.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_PING_PONG_STREAMING_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_ping_pong_streaming -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_ping_pong_streaming_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_ping_pong_streaming_test: $(CHTTP2_SOCKET_PAIR_PING_PONG_STREAMING_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_PING_PONG_STREAMING_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_ping_pong_streaming_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_ping_pong_streaming_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_PING_PONG_STREAMING_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_PING_PONG_STREAMING_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_ping_pong_streaming_test
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_request_response_with_binary_metadata_and_payload_test: openssl_dep_error
else
bins/chttp2_socket_pair_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_request_response_with_binary_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_request_response_with_binary_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_request_response_with_binary_metadata_and_payload_test
endif
deps_chttp2_socket_pair_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
-include $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
endif
clean_chttp2_socket_pair_request_response_with_binary_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_request_response_with_binary_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_request_response_with_binary_metadata_and_payload_test
Initial import.
10 years ago
CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_request_response_with_metadata_and_payload_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_request_response_with_metadata_and_payload_test: $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_request_response_with_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_request_response_with_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_request_response_with_metadata_and_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_request_response_with_metadata_and_payload_test: $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_request_response_with_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_request_response_with_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_request_response_with_metadata_and_payload_test
CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC = \
CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_request_response_with_payload_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_request_response_with_payload_test: $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_request_response_with_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_request_response_with_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_request_response_with_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_request_response_with_payload_test: $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_request_response_with_payload_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_request_response_with_payload_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_request_response_with_payload_test
CHTTP2_SOCKET_PAIR_SIMPLE_DELAYED_REQUEST_TEST_SRC = \
CHTTP2_SOCKET_PAIR_SIMPLE_DELAYED_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
CHTTP2_SOCKET_PAIR_SIMPLE_DELAYED_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_simple_delayed_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_simple_delayed_request_test: $(CHTTP2_SOCKET_PAIR_SIMPLE_DELAYED_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_simple_delayed_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_SIMPLE_DELAYED_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_simple_delayed_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_simple_delayed_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_simple_delayed_request_test: $(CHTTP2_SOCKET_PAIR_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_simple_delayed_request_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_simple_delayed_request_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_SIMPLE_DELAYED_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_simple_delayed_request_test
CHTTP2_SOCKET_PAIR_SIMPLE_REQUEST_TEST_SRC = \
CHTTP2_SOCKET_PAIR_SIMPLE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_SIMPLE_REQUEST_TEST_SRC))))
CHTTP2_SOCKET_PAIR_SIMPLE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_SIMPLE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_simple_request_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_simple_request_test: $(CHTTP2_SOCKET_PAIR_SIMPLE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_simple_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_SIMPLE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_simple_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_simple_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_simple_request_test: $(CHTTP2_SOCKET_PAIR_SIMPLE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_SIMPLE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_simple_request_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_simple_request_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_SIMPLE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_SIMPLE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_simple_request_test
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_SOCKET_PAIR_THREAD_STRESS_TEST_SRC = \
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_SOCKET_PAIR_THREAD_STRESS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_THREAD_STRESS_TEST_SRC))))
CHTTP2_SOCKET_PAIR_THREAD_STRESS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_THREAD_STRESS_TEST_SRC))))
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_thread_stress_test: openssl_dep_error
else
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
bins/chttp2_socket_pair_thread_stress_test: $(CHTTP2_SOCKET_PAIR_THREAD_STRESS_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_thread_stress.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
Initial import.
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_THREAD_STRESS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_thread_stress -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_thread_stress_test
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
deps_chttp2_socket_pair_thread_stress_test: $(CHTTP2_SOCKET_PAIR_THREAD_STRESS_TEST_DEPS)
Initial import.
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
-include $(CHTTP2_SOCKET_PAIR_THREAD_STRESS_TEST_DEPS)
Initial import.
10 years ago
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
clean_chttp2_socket_pair_thread_stress_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_thread_stress_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_THREAD_STRESS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_THREAD_STRESS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_thread_stress_test
Initial import.
10 years ago
CHTTP2_SOCKET_PAIR_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC = \
CHTTP2_SOCKET_PAIR_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
CHTTP2_SOCKET_PAIR_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_writes_done_hangs_with_pending_read_test: openssl_dep_error
else
Initial import.
10 years ago
bins/chttp2_socket_pair_writes_done_hangs_with_pending_read_test: $(CHTTP2_SOCKET_PAIR_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair.a libs/libend2end_test_writes_done_hangs_with_pending_read.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Initial import.
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair -lend2end_test_writes_done_hangs_with_pending_read -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_writes_done_hangs_with_pending_read_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
deps_chttp2_socket_pair_writes_done_hangs_with_pending_read_test: $(CHTTP2_SOCKET_PAIR_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Initial import.
10 years ago
-include $(CHTTP2_SOCKET_PAIR_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Initial import.
10 years ago
clean_chttp2_socket_pair_writes_done_hangs_with_pending_read_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_writes_done_hangs_with_pending_read_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_writes_done_hangs_with_pending_read_test
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_cancel_after_accept.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_cancel_after_accept -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_and_writes_closed_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_and_writes_closed_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_cancel_after_accept_and_writes_closed.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_cancel_after_accept_and_writes_closed -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_and_writes_closed_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_and_writes_closed_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_and_writes_closed_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_and_writes_closed_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_ACCEPT_AND_WRITES_CLOSED_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_accept_and_writes_closed_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_INVOKE_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_INVOKE_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_invoke_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_invoke_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_cancel_after_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_cancel_after_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_cancel_after_invoke_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_cancel_after_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_cancel_after_invoke_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_AFTER_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_cancel_after_invoke_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_BEFORE_INVOKE_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_BEFORE_INVOKE_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_BEFORE_INVOKE_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_BEFORE_INVOKE_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_BEFORE_INVOKE_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_cancel_before_invoke_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_cancel_before_invoke_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_BEFORE_INVOKE_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_cancel_before_invoke.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_BEFORE_INVOKE_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_cancel_before_invoke -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_cancel_before_invoke_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_cancel_before_invoke_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_BEFORE_INVOKE_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_BEFORE_INVOKE_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_cancel_before_invoke_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_cancel_before_invoke_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_BEFORE_INVOKE_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_BEFORE_INVOKE_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_cancel_before_invoke_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_IN_A_VACUUM_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_IN_A_VACUUM_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_IN_A_VACUUM_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_IN_A_VACUUM_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_IN_A_VACUUM_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_cancel_in_a_vacuum_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_cancel_in_a_vacuum_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_IN_A_VACUUM_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_cancel_in_a_vacuum.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_IN_A_VACUUM_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_cancel_in_a_vacuum -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_cancel_in_a_vacuum_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_cancel_in_a_vacuum_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_IN_A_VACUUM_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_IN_A_VACUUM_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_cancel_in_a_vacuum_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_cancel_in_a_vacuum_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_IN_A_VACUUM_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_CANCEL_IN_A_VACUUM_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_cancel_in_a_vacuum_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_inflight_calls_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_early_server_shutdown_finishes_inflight_calls.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_early_server_shutdown_finishes_inflight_calls -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_inflight_calls_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_inflight_calls_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_inflight_calls_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_inflight_calls_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_INFLIGHT_CALLS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_inflight_calls_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_tags_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_tags_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_early_server_shutdown_finishes_tags.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_early_server_shutdown_finishes_tags -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_tags_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_tags_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_tags_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_tags_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_EARLY_SERVER_SHUTDOWN_FINISHES_TAGS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_early_server_shutdown_finishes_tags_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_INVOKE_LARGE_REQUEST_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_INVOKE_LARGE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_INVOKE_LARGE_REQUEST_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_INVOKE_LARGE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_INVOKE_LARGE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_invoke_large_request_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_invoke_large_request_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_INVOKE_LARGE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_invoke_large_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_INVOKE_LARGE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_invoke_large_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_invoke_large_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_invoke_large_request_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_INVOKE_LARGE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_INVOKE_LARGE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_invoke_large_request_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_invoke_large_request_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_INVOKE_LARGE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_INVOKE_LARGE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_invoke_large_request_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_MAX_CONCURRENT_STREAMS_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_MAX_CONCURRENT_STREAMS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_MAX_CONCURRENT_STREAMS_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_MAX_CONCURRENT_STREAMS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_MAX_CONCURRENT_STREAMS_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_max_concurrent_streams_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_max_concurrent_streams_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_MAX_CONCURRENT_STREAMS_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_max_concurrent_streams.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_MAX_CONCURRENT_STREAMS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_max_concurrent_streams -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_max_concurrent_streams_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_max_concurrent_streams_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_MAX_CONCURRENT_STREAMS_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_MAX_CONCURRENT_STREAMS_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_max_concurrent_streams_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_max_concurrent_streams_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_MAX_CONCURRENT_STREAMS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_MAX_CONCURRENT_STREAMS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_max_concurrent_streams_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_NO_OP_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_NO_OP_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_NO_OP_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_NO_OP_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_NO_OP_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_no_op_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_no_op_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_NO_OP_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_no_op.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_NO_OP_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_no_op -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_no_op_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_no_op_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_NO_OP_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_NO_OP_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_no_op_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_no_op_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_NO_OP_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_NO_OP_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_no_op_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_PING_PONG_STREAMING_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_PING_PONG_STREAMING_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_PING_PONG_STREAMING_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_PING_PONG_STREAMING_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_PING_PONG_STREAMING_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_ping_pong_streaming_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_ping_pong_streaming_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_PING_PONG_STREAMING_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_ping_pong_streaming.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_PING_PONG_STREAMING_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_ping_pong_streaming -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_ping_pong_streaming_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_ping_pong_streaming_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_PING_PONG_STREAMING_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_PING_PONG_STREAMING_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_ping_pong_streaming_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_ping_pong_streaming_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_PING_PONG_STREAMING_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_PING_PONG_STREAMING_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_ping_pong_streaming_test
Binary header encoding. Also fixes a rather embarrassing bug in bin_encoder.c. Change on 2014/12/12 by ctiller <ctiller@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82024795
10 years ago
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_SRC))))
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_binary_metadata_and_payload_test: openssl_dep_error
else
bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_request_response_with_binary_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
$(Q) mkdir -p `dirname $@`
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_request_response_with_binary_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_binary_metadata_and_payload_test
endif
deps_chttp2_socket_pair_one_byte_at_a_time_request_response_with_binary_metadata_and_payload_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
endif
clean_chttp2_socket_pair_one_byte_at_a_time_request_response_with_binary_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_request_response_with_binary_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_BINARY_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_binary_metadata_and_payload_test
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_metadata_and_payload_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_metadata_and_payload_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_request_response_with_metadata_and_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_request_response_with_metadata_and_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_metadata_and_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_request_response_with_metadata_and_payload_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_request_response_with_metadata_and_payload_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_request_response_with_metadata_and_payload_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_METADATA_AND_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_metadata_and_payload_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_payload_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_payload_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_request_response_with_payload.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_request_response_with_payload -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_payload_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_request_response_with_payload_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_request_response_with_payload_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_request_response_with_payload_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_REQUEST_RESPONSE_WITH_PAYLOAD_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_request_response_with_payload_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_DELAYED_REQUEST_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_DELAYED_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_DELAYED_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_DELAYED_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_simple_delayed_request_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_simple_delayed_request_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_DELAYED_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_simple_delayed_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_DELAYED_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_simple_delayed_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_simple_delayed_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_simple_delayed_request_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_simple_delayed_request_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_simple_delayed_request_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_DELAYED_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_DELAYED_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_simple_delayed_request_test
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_REQUEST_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_REQUEST_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_REQUEST_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_REQUEST_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_REQUEST_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_simple_request_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_simple_request_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_REQUEST_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_simple_request.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_REQUEST_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_simple_request -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_simple_request_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_simple_request_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_REQUEST_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_REQUEST_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_simple_request_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_simple_request_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_REQUEST_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_SIMPLE_REQUEST_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_simple_request_test
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_THREAD_STRESS_TEST_SRC = \
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_THREAD_STRESS_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_THREAD_STRESS_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_THREAD_STRESS_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_THREAD_STRESS_TEST_SRC))))
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_thread_stress_test: openssl_dep_error
else
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_thread_stress_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_THREAD_STRESS_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_thread_stress.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_THREAD_STRESS_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_thread_stress -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_thread_stress_test
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_thread_stress_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_THREAD_STRESS_TEST_DEPS)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_THREAD_STRESS_TEST_DEPS)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Strip "_test" from the name of thread_stress_test.c. Having a BUILD target named chttp2_fullstack_thread_stress_test_test is deceptive as it is a direct test of the system rather than a test of a test of the system. Change on 2014/12/09 by nathaniel <nathaniel@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81688649
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_thread_stress_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_thread_stress_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_THREAD_STRESS_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_THREAD_STRESS_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_thread_stress_test
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC = \
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS = $(addprefix objs/, $(addsuffix .o, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS = $(addprefix deps/, $(addsuffix .dep, $(basename $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_SRC))))
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifeq ($(NO_SECURE),true)
bins/chttp2_socket_pair_one_byte_at_a_time_writes_done_hangs_with_pending_read_test: openssl_dep_error
else
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
bins/chttp2_socket_pair_one_byte_at_a_time_writes_done_hangs_with_pending_read_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) libs/libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time.a libs/libend2end_test_writes_done_hangs_with_pending_read.a libs/libend2end_certs.a libs/libgrpc_test_util.a libs/libgrpc.a libs/libgpr.a
$(E) "[LD] Linking $@"
C/C++ separation of the install target. -) install, static, shared and strip are now fully dissociated -) make_dirs is gone Change on 2014/12/08 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81631734
10 years ago
$(Q) mkdir -p `dirname $@`
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
$(Q) $(LD) $(LDFLAGS) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS) -Llibs -lend2end_fixture_chttp2_socket_pair_one_byte_at_a_time -lend2end_test_writes_done_hangs_with_pending_read -lend2end_certs -lgrpc_test_util -lgrpc -lgpr $(LDLIBS) $(LDLIBS_SECURE) -o bins/chttp2_socket_pair_one_byte_at_a_time_writes_done_hangs_with_pending_read_test
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
deps_chttp2_socket_pair_one_byte_at_a_time_writes_done_hangs_with_pending_read_test: $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
ifneq ($(NO_SECURE),true)
ifneq ($(NO_DEPS),true)
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
-include $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
endif
Makefile will now build zlib and/or OpenSSL if needed. -) Detecting system and embedded presence of zlib and OpenSSL with ALPN. -) Automatically disabling secure targets if no OpenSSL present --> make all won't work if no OpenSSL is present, forcing the users to select the nonsecure Makefile targets explicitely. -) Removing build instructions for OpenSSL - this isn't really necessary anymore. -) Adding more blurb about OpenSSL and the new Makefile features. Change on 2014/12/12 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82020890
10 years ago
endif
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
clean_chttp2_socket_pair_one_byte_at_a_time_writes_done_hangs_with_pending_read_test:
$(E) "[CLEAN] Cleaning chttp2_socket_pair_one_byte_at_a_time_writes_done_hangs_with_pending_read_test files"
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_OBJS)
$(Q) $(RM) $(CHTTP2_SOCKET_PAIR_ONE_BYTE_AT_A_TIME_WRITES_DONE_HANGS_WITH_PENDING_READ_TEST_DEPS)
$(Q) $(RM) bins/chttp2_socket_pair_one_byte_at_a_time_writes_done_hangs_with_pending_read_test
Initial import.
10 years ago
Fix a bug in base-log, and add test. The tests compute usable space in the log, but do so using subtraction on unsigned values and did not correctly account for number of blocks used per-core. This could lead underflow, and an incorrect space calculation. In addition, testing showed that with current defaults, tests would fail on a system with > ~64 cores. This can be alleviated through use of a larger log. An explicit check in the log initialization has been added for this case, and the log size increased if necessary. Change on 2014/12/15 by aveitch <aveitch@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=82154432
10 years ago
.PHONY: all strip tools dep_error openssl_dep_error openssl_dep_message git_update stop buildtests buildtests_c buildtests_cxx test test_c test_cxx install install_c install_cxx install-headers install-headers_c install-headers_cxx install-shared install-shared_c install-shared_cxx install-static install-static_c install-static_cxx strip strip-shared strip-static strip_c strip-shared_c strip-static_c strip_cxx strip-shared_cxx strip-static_cxx clean deps_libgpr clean_libgpr deps_libgrpc clean_libgrpc deps_libgrpc_test_util clean_libgrpc_test_util deps_libgrpc++ clean_libgrpc++ deps_libgrpc++_test_util clean_libgrpc++_test_util deps_libend2end_fixture_chttp2_fake_security clean_libend2end_fixture_chttp2_fake_security deps_libend2end_fixture_chttp2_fullstack clean_libend2end_fixture_chttp2_fullstack deps_libend2end_fixture_chttp2_simple_ssl_fullstack clean_libend2end_fixture_chttp2_simple_ssl_fullstack deps_libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack clean_libend2end_fixture_chttp2_simple_ssl_with_oauth2_fullstack deps_libend2end_fixture_chttp2_socket_pair clean_libend2end_fixture_chttp2_socket_pair deps_libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time clean_libend2end_fixture_chttp2_socket_pair_one_byte_at_a_time deps_libend2end_test_cancel_after_accept clean_libend2end_test_cancel_after_accept deps_libend2end_test_cancel_after_accept_and_writes_closed clean_libend2end_test_cancel_after_accept_and_writes_closed deps_libend2end_test_cancel_after_invoke clean_libend2end_test_cancel_after_invoke deps_libend2end_test_cancel_before_invoke clean_libend2end_test_cancel_before_invoke deps_libend2end_test_cancel_in_a_vacuum clean_libend2end_test_cancel_in_a_vacuum deps_libend2end_test_early_server_shutdown_finishes_inflight_calls clean_libend2end_test_early_server_shutdown_finishes_inflight_calls deps_libend2end_test_early_server_shutdown_finishes_tags clean_libend2end_test_early_server_shutdown_finishes_tags deps_libend2end_test_invoke_large_request clean_libend2end_test_invoke_large_request deps_libend2end_test_max_concurrent_streams clean_libend2end_test_max_concurrent_streams deps_libend2end_test_no_op clean_libend2end_test_no_op deps_libend2end_test_ping_pong_streaming clean_libend2end_test_ping_pong_streaming deps_libend2end_test_request_response_with_binary_metadata_and_payload clean_libend2end_test_request_response_with_binary_metadata_and_payload deps_libend2end_test_request_response_with_metadata_and_payload clean_libend2end_test_request_response_with_metadata_and_payload deps_libend2end_test_request_response_with_payload clean_libend2end_test_request_response_with_payload deps_libend2end_test_simple_delayed_request clean_libend2end_test_simple_delayed_request deps_libend2end_test_simple_request clean_libend2end_test_simple_request deps_libend2end_test_thread_stress clean_libend2end_test_thread_stress deps_libend2end_test_writes_done_hangs_with_pending_read clean_libend2end_test_writes_done_hangs_with_pending_read deps_libend2end_certs clean_libend2end_certs deps_libgrpc_unsecure clean_libgrpc_unsecure deps_gen_hpack_tables clean_gen_hpack_tables deps_cpp_plugin clean_cpp_plugin deps_ruby_plugin clean_ruby_plugin deps_grpc_byte_buffer_reader_test clean_grpc_byte_buffer_reader_test deps_gpr_cancellable_test clean_gpr_cancellable_test deps_gpr_log_test clean_gpr_log_test deps_gpr_useful_test clean_gpr_useful_test deps_gpr_cmdline_test clean_gpr_cmdline_test deps_gpr_histogram_test clean_gpr_histogram_test deps_gpr_host_port_test clean_gpr_host_port_test deps_gpr_slice_buffer_test clean_gpr_slice_buffer_test deps_gpr_slice_test clean_gpr_slice_test deps_gpr_string_test clean_gpr_string_test deps_gpr_sync_test clean_gpr_sync_test deps_gpr_thd_test clean_gpr_thd_test deps_gpr_time_test clean_gpr_time_test deps_murmur_hash_test clean_murmur_hash_test deps_grpc_stream_op_test clean_grpc_stream_op_test deps_alpn_test clean_alpn_test deps_chttp2_stream_encoder_test clean_chttp2_stream_encoder_test deps_hpack_table_test clean_hpack_table_test deps_chttp2_stream_map_test clean_chttp2_stream_map_test deps_hpa