Protocol Buffers - Google's data interchange format (grpc依赖) https://developers.google.com/protocol-buffers/
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.

395 lines
14 KiB

16 years ago
#
# This Makefile builds the upb library as well as associated tests, tools, and
# language extensions.
#
# It does not use autoconf/automake/libtool in order to stay lightweight and
# avoid the need for running ./configure.
#
16 years ago
# Summary of compiler flags you may want to use:
#
# * -DNDEBUG: makes binary smaller and faster by removing sanity checks.
# * -O3: optimize for maximum speed
# * -fomit-frame-pointer: makes code smaller and faster by freeing up a reg.
#
# Threading:
# * -DUPB_USE_PTHREADS: configures upb to use pthreads r/w lock.
# * -DUPB_THREAD_UNSAFE: remove all thread-safety.
# * -pthread: required on GCC to enable pthreads (but what does it do?)
#
# Other:
# * -DUPB_UNALIGNED_READS_OK: makes code smaller, but not standard compliant
16 years ago
.PHONY: all lib clean tests test benchmarks benchmark descriptorgen
.PHONY: clean_leave_profile
# Default rule: just build libupb.
all: lib
# User-specified CFLAGS.
USER_CFLAGS=$(strip $(shell test -f perf-cppflags && cat perf-cppflags))
# If the user doesn't specify an -O setting, we default to -O3, except
# for upb_def which gets -Os.
ifeq (, $(findstring -O, $(USER_CFLAGS)))
USER_CFLAGS += -O3
DEF_OPT = -Os
endif
# Basic compiler/flag setup.
CC=gcc
CXX=g++
CFLAGS=-std=gnu99
INCLUDE=-Isrc -Itests -I.
upb_stream: all callbacks registered ahead-of-time. This is a significant change to the upb_stream protocol, and should hopefully be the last significant change. All callbacks are now registered ahead-of-time instead of having delegated callbacks registered at runtime, which makes it much easier to aggressively optimize ahead-of-time (like with a JIT). Other impacts of this change: - You no longer need to have loaded descriptor.proto as a upb_def to load other descriptors! This means the special-case code we used for bootstrapping is no longer necessary, and we no longer need to link the descriptor for descriptor.proto into upb. - A client can now register any upb_value as what will be delivered to their value callback, not just a upb_fielddef*. This should allow for other clients to get more bang out of the streaming decoder. This change unfortunately causes a bit of a performance regression -- I think largely due to highly suboptimal code that GCC generates when structs are returned by value. See: http://blog.reverberate.org/2011/03/19/when-a-compilers-slow-code-actually-bites-you/ On the other hand, once we have a JIT this should no longer matter. Performance numbers: plain.parsestream_googlemessage1.upb_table: 374 -> 396 (5.88) plain.parsestream_googlemessage2.upb_table: 616 -> 449 (-27.11) plain.parsetostruct_googlemessage1.upb_table_byref: 268 -> 269 (0.37) plain.parsetostruct_googlemessage1.upb_table_byval: 215 -> 204 (-5.12) plain.parsetostruct_googlemessage2.upb_table_byref: 307 -> 281 (-8.47) plain.parsetostruct_googlemessage2.upb_table_byval: 297 -> 272 (-8.42) omitfp.parsestream_googlemessage1.upb_table: 423 -> 410 (-3.07) omitfp.parsestream_googlemessage2.upb_table: 679 -> 483 (-28.87) omitfp.parsetostruct_googlemessage1.upb_table_byref: 287 -> 282 (-1.74) omitfp.parsetostruct_googlemessage1.upb_table_byval: 226 -> 219 (-3.10) omitfp.parsetostruct_googlemessage2.upb_table_byref: 315 -> 298 (-5.40) omitfp.parsetostruct_googlemessage2.upb_table_byval: 297 -> 287 (-3.37)
14 years ago
CPPFLAGS=$(INCLUDE) -Wall -Wextra $(USER_CFLAGS)
LDLIBS=-lpthread src/libupb.a
# Build with "make Q=" to see all commands that are being executed.
Q=@
ifeq ($(Q), @)
E=@echo
else
E=@:
endif
# Dependency generating. #######################################################
-include deps
# Unfortuantely we can't easily generate deps for benchmarks or tests because
# of the scheme we use that compiles the same source file multiple times with
# different -D options, which can include different header files.
deps: gen-deps.sh Makefile $(CORE) $(STREAM)
$(Q) CPPFLAGS="$(CPPFLAGS)" ./gen-deps.sh $(CORE) $(STREAM)
$(E) Regenerating dependencies for src/...
$(ALLSRC): perf-cppflags
# Source files. ###############################################################
# Every source file used in upb should appear here.
upb_stream: all callbacks registered ahead-of-time. This is a significant change to the upb_stream protocol, and should hopefully be the last significant change. All callbacks are now registered ahead-of-time instead of having delegated callbacks registered at runtime, which makes it much easier to aggressively optimize ahead-of-time (like with a JIT). Other impacts of this change: - You no longer need to have loaded descriptor.proto as a upb_def to load other descriptors! This means the special-case code we used for bootstrapping is no longer necessary, and we no longer need to link the descriptor for descriptor.proto into upb. - A client can now register any upb_value as what will be delivered to their value callback, not just a upb_fielddef*. This should allow for other clients to get more bang out of the streaming decoder. This change unfortunately causes a bit of a performance regression -- I think largely due to highly suboptimal code that GCC generates when structs are returned by value. See: http://blog.reverberate.org/2011/03/19/when-a-compilers-slow-code-actually-bites-you/ On the other hand, once we have a JIT this should no longer matter. Performance numbers: plain.parsestream_googlemessage1.upb_table: 374 -> 396 (5.88) plain.parsestream_googlemessage2.upb_table: 616 -> 449 (-27.11) plain.parsetostruct_googlemessage1.upb_table_byref: 268 -> 269 (0.37) plain.parsetostruct_googlemessage1.upb_table_byval: 215 -> 204 (-5.12) plain.parsetostruct_googlemessage2.upb_table_byref: 307 -> 281 (-8.47) plain.parsetostruct_googlemessage2.upb_table_byval: 297 -> 272 (-8.42) omitfp.parsestream_googlemessage1.upb_table: 423 -> 410 (-3.07) omitfp.parsestream_googlemessage2.upb_table: 679 -> 483 (-28.87) omitfp.parsetostruct_googlemessage1.upb_table_byref: 287 -> 282 (-1.74) omitfp.parsetostruct_googlemessage1.upb_table_byval: 226 -> 219 (-3.10) omitfp.parsetostruct_googlemessage2.upb_table_byref: 315 -> 298 (-5.40) omitfp.parsetostruct_googlemessage2.upb_table_byval: 297 -> 287 (-3.37)
14 years ago
# The core library.
CORE= \
src/upb.c \
upb_stream: all callbacks registered ahead-of-time. This is a significant change to the upb_stream protocol, and should hopefully be the last significant change. All callbacks are now registered ahead-of-time instead of having delegated callbacks registered at runtime, which makes it much easier to aggressively optimize ahead-of-time (like with a JIT). Other impacts of this change: - You no longer need to have loaded descriptor.proto as a upb_def to load other descriptors! This means the special-case code we used for bootstrapping is no longer necessary, and we no longer need to link the descriptor for descriptor.proto into upb. - A client can now register any upb_value as what will be delivered to their value callback, not just a upb_fielddef*. This should allow for other clients to get more bang out of the streaming decoder. This change unfortunately causes a bit of a performance regression -- I think largely due to highly suboptimal code that GCC generates when structs are returned by value. See: http://blog.reverberate.org/2011/03/19/when-a-compilers-slow-code-actually-bites-you/ On the other hand, once we have a JIT this should no longer matter. Performance numbers: plain.parsestream_googlemessage1.upb_table: 374 -> 396 (5.88) plain.parsestream_googlemessage2.upb_table: 616 -> 449 (-27.11) plain.parsetostruct_googlemessage1.upb_table_byref: 268 -> 269 (0.37) plain.parsetostruct_googlemessage1.upb_table_byval: 215 -> 204 (-5.12) plain.parsetostruct_googlemessage2.upb_table_byref: 307 -> 281 (-8.47) plain.parsetostruct_googlemessage2.upb_table_byval: 297 -> 272 (-8.42) omitfp.parsestream_googlemessage1.upb_table: 423 -> 410 (-3.07) omitfp.parsestream_googlemessage2.upb_table: 679 -> 483 (-28.87) omitfp.parsetostruct_googlemessage1.upb_table_byref: 287 -> 282 (-1.74) omitfp.parsetostruct_googlemessage1.upb_table_byval: 226 -> 219 (-3.10) omitfp.parsetostruct_googlemessage2.upb_table_byref: 315 -> 298 (-5.40) omitfp.parsetostruct_googlemessage2.upb_table_byval: 297 -> 287 (-3.37)
14 years ago
src/upb_stream.c \
src/upb_table.c \
src/upb_string.c \
src/upb_def.c \
upb_stream: all callbacks registered ahead-of-time. This is a significant change to the upb_stream protocol, and should hopefully be the last significant change. All callbacks are now registered ahead-of-time instead of having delegated callbacks registered at runtime, which makes it much easier to aggressively optimize ahead-of-time (like with a JIT). Other impacts of this change: - You no longer need to have loaded descriptor.proto as a upb_def to load other descriptors! This means the special-case code we used for bootstrapping is no longer necessary, and we no longer need to link the descriptor for descriptor.proto into upb. - A client can now register any upb_value as what will be delivered to their value callback, not just a upb_fielddef*. This should allow for other clients to get more bang out of the streaming decoder. This change unfortunately causes a bit of a performance regression -- I think largely due to highly suboptimal code that GCC generates when structs are returned by value. See: http://blog.reverberate.org/2011/03/19/when-a-compilers-slow-code-actually-bites-you/ On the other hand, once we have a JIT this should no longer matter. Performance numbers: plain.parsestream_googlemessage1.upb_table: 374 -> 396 (5.88) plain.parsestream_googlemessage2.upb_table: 616 -> 449 (-27.11) plain.parsetostruct_googlemessage1.upb_table_byref: 268 -> 269 (0.37) plain.parsetostruct_googlemessage1.upb_table_byval: 215 -> 204 (-5.12) plain.parsetostruct_googlemessage2.upb_table_byref: 307 -> 281 (-8.47) plain.parsetostruct_googlemessage2.upb_table_byval: 297 -> 272 (-8.42) omitfp.parsestream_googlemessage1.upb_table: 423 -> 410 (-3.07) omitfp.parsestream_googlemessage2.upb_table: 679 -> 483 (-28.87) omitfp.parsetostruct_googlemessage1.upb_table_byref: 287 -> 282 (-1.74) omitfp.parsetostruct_googlemessage1.upb_table_byval: 226 -> 219 (-3.10) omitfp.parsetostruct_googlemessage2.upb_table_byref: 315 -> 298 (-5.40) omitfp.parsetostruct_googlemessage2.upb_table_byval: 297 -> 287 (-3.37)
14 years ago
src/upb_msg.c \
upb_stream: all callbacks registered ahead-of-time. This is a significant change to the upb_stream protocol, and should hopefully be the last significant change. All callbacks are now registered ahead-of-time instead of having delegated callbacks registered at runtime, which makes it much easier to aggressively optimize ahead-of-time (like with a JIT). Other impacts of this change: - You no longer need to have loaded descriptor.proto as a upb_def to load other descriptors! This means the special-case code we used for bootstrapping is no longer necessary, and we no longer need to link the descriptor for descriptor.proto into upb. - A client can now register any upb_value as what will be delivered to their value callback, not just a upb_fielddef*. This should allow for other clients to get more bang out of the streaming decoder. This change unfortunately causes a bit of a performance regression -- I think largely due to highly suboptimal code that GCC generates when structs are returned by value. See: http://blog.reverberate.org/2011/03/19/when-a-compilers-slow-code-actually-bites-you/ On the other hand, once we have a JIT this should no longer matter. Performance numbers: plain.parsestream_googlemessage1.upb_table: 374 -> 396 (5.88) plain.parsestream_googlemessage2.upb_table: 616 -> 449 (-27.11) plain.parsetostruct_googlemessage1.upb_table_byref: 268 -> 269 (0.37) plain.parsetostruct_googlemessage1.upb_table_byval: 215 -> 204 (-5.12) plain.parsetostruct_googlemessage2.upb_table_byref: 307 -> 281 (-8.47) plain.parsetostruct_googlemessage2.upb_table_byval: 297 -> 272 (-8.42) omitfp.parsestream_googlemessage1.upb_table: 423 -> 410 (-3.07) omitfp.parsestream_googlemessage2.upb_table: 679 -> 483 (-28.87) omitfp.parsetostruct_googlemessage1.upb_table_byref: 287 -> 282 (-1.74) omitfp.parsetostruct_googlemessage1.upb_table_byval: 226 -> 219 (-3.10) omitfp.parsetostruct_googlemessage2.upb_table_byref: 315 -> 298 (-5.40) omitfp.parsetostruct_googlemessage2.upb_table_byval: 297 -> 287 (-3.37)
14 years ago
# Common encoders/decoders -- you're almost certain to want these.
STREAM= \
src/upb_decoder.c \
src/upb_stdio.c \
src/upb_textprinter.c \
src/upb_strstream.c \
src/upb_glue.c \
# Parts of core that are yet to be converted.
upb_stream: all callbacks registered ahead-of-time. This is a significant change to the upb_stream protocol, and should hopefully be the last significant change. All callbacks are now registered ahead-of-time instead of having delegated callbacks registered at runtime, which makes it much easier to aggressively optimize ahead-of-time (like with a JIT). Other impacts of this change: - You no longer need to have loaded descriptor.proto as a upb_def to load other descriptors! This means the special-case code we used for bootstrapping is no longer necessary, and we no longer need to link the descriptor for descriptor.proto into upb. - A client can now register any upb_value as what will be delivered to their value callback, not just a upb_fielddef*. This should allow for other clients to get more bang out of the streaming decoder. This change unfortunately causes a bit of a performance regression -- I think largely due to highly suboptimal code that GCC generates when structs are returned by value. See: http://blog.reverberate.org/2011/03/19/when-a-compilers-slow-code-actually-bites-you/ On the other hand, once we have a JIT this should no longer matter. Performance numbers: plain.parsestream_googlemessage1.upb_table: 374 -> 396 (5.88) plain.parsestream_googlemessage2.upb_table: 616 -> 449 (-27.11) plain.parsetostruct_googlemessage1.upb_table_byref: 268 -> 269 (0.37) plain.parsetostruct_googlemessage1.upb_table_byval: 215 -> 204 (-5.12) plain.parsetostruct_googlemessage2.upb_table_byref: 307 -> 281 (-8.47) plain.parsetostruct_googlemessage2.upb_table_byval: 297 -> 272 (-8.42) omitfp.parsestream_googlemessage1.upb_table: 423 -> 410 (-3.07) omitfp.parsestream_googlemessage2.upb_table: 679 -> 483 (-28.87) omitfp.parsetostruct_googlemessage1.upb_table_byref: 287 -> 282 (-1.74) omitfp.parsetostruct_googlemessage1.upb_table_byval: 226 -> 219 (-3.10) omitfp.parsetostruct_googlemessage2.upb_table_byref: 315 -> 298 (-5.40) omitfp.parsetostruct_googlemessage2.upb_table_byval: 297 -> 287 (-3.37)
14 years ago
OTHERSRC=src/upb_encoder.c
BENCHMARKS_SRC= \
benchmarks/main.c \
benchmarks/parsestream.upb_table.c \
benchmarks/parsetostruct.upb_table.c
TESTS_SRC= \
tests/test_decoder.c \
tests/test_def.c \
tests/test_string.c \
tests/tests.c \
tests/tests_varint.c \
tests/test_vs_proto2.cc
upb_stream: all callbacks registered ahead-of-time. This is a significant change to the upb_stream protocol, and should hopefully be the last significant change. All callbacks are now registered ahead-of-time instead of having delegated callbacks registered at runtime, which makes it much easier to aggressively optimize ahead-of-time (like with a JIT). Other impacts of this change: - You no longer need to have loaded descriptor.proto as a upb_def to load other descriptors! This means the special-case code we used for bootstrapping is no longer necessary, and we no longer need to link the descriptor for descriptor.proto into upb. - A client can now register any upb_value as what will be delivered to their value callback, not just a upb_fielddef*. This should allow for other clients to get more bang out of the streaming decoder. This change unfortunately causes a bit of a performance regression -- I think largely due to highly suboptimal code that GCC generates when structs are returned by value. See: http://blog.reverberate.org/2011/03/19/when-a-compilers-slow-code-actually-bites-you/ On the other hand, once we have a JIT this should no longer matter. Performance numbers: plain.parsestream_googlemessage1.upb_table: 374 -> 396 (5.88) plain.parsestream_googlemessage2.upb_table: 616 -> 449 (-27.11) plain.parsetostruct_googlemessage1.upb_table_byref: 268 -> 269 (0.37) plain.parsetostruct_googlemessage1.upb_table_byval: 215 -> 204 (-5.12) plain.parsetostruct_googlemessage2.upb_table_byref: 307 -> 281 (-8.47) plain.parsetostruct_googlemessage2.upb_table_byval: 297 -> 272 (-8.42) omitfp.parsestream_googlemessage1.upb_table: 423 -> 410 (-3.07) omitfp.parsestream_googlemessage2.upb_table: 679 -> 483 (-28.87) omitfp.parsetostruct_googlemessage1.upb_table_byref: 287 -> 282 (-1.74) omitfp.parsetostruct_googlemessage1.upb_table_byval: 226 -> 219 (-3.10) omitfp.parsetostruct_googlemessage2.upb_table_byref: 315 -> 298 (-5.40) omitfp.parsetostruct_googlemessage2.upb_table_byval: 297 -> 287 (-3.37)
14 years ago
#tests/test_stream.c \
ALLSRC=$(CORE) $(STREAM) $(BENCHMARKS_SRC) $(TESTS_SRC)
# Rules. #######################################################################
clean_leave_profile:
rm -rf $(LIBUPB) $(LIBUPB_PIC)
rm -rf $(call rwildcard,,*.o) $(call rwildcard,,*.lo) $(call rwildcard,,*.dSYM)
rm -rf src/upb_decoder_x86.h
rm -rf benchmark/google_messages.proto.pb benchmark/google_messages.pb.* benchmarks/b.* benchmarks/*.pb*
rm -rf src/jit_debug_elf_file.o
rm -rf src/jit_debug_elf_file.h
rm -rf $(TESTS) tests/t.*
rm -rf src/descriptor.pb
rm -rf src/upbc deps
rm -rf lang_ext/lua/upb.so
cd lang_ext/python && python setup.py clean --all
clean: clean_leave_profile
rm -rf $(call rwildcard,,*.gcno) $(call rwildcard,,*.gcda)
# Core library (libupb.a).
SRC=$(CORE) $(STREAM)
LIBUPB=src/libupb.a
LIBUPB_PIC=src/libupb_pic.a
lib: $(LIBUPB)
OBJ=$(patsubst %.c,%.o,$(SRC))
PICOBJ=$(patsubst %.c,%.lo,$(SRC))
ifneq (, $(findstring DUPB_USE_JIT_X64, $(USER_CFLAGS)))
src/upb_decoder.o: src/upb_decoder_x86.h
ifeq (, $(findstring DNDEBUG, $(USER_CFLAGS)))
$(error "JIT only works with -DNDEBUG enabled!")
endif
endif
$(LIBUPB): $(OBJ)
$(E) AR $(LIBUPB)
$(Q) ar rcs $(LIBUPB) $(OBJ)
$(LIBUPB_PIC): $(PICOBJ)
$(E) AR $(LIBUPB_PIC)
$(Q) ar rcs $(LIBUPB_PIC) $(PICOBJ)
%.o : %.c
$(E) CC $<
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) -c -o $@ $<
%.lo : %.c
$(E) 'CC -fPIC' $<
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) -c -o $@ $< -fPIC
# Override the optimization level for upb_def.o, because it is not in the
# critical path but gets very large when -O3 is used.
src/upb_def.o: src/upb_def.c
$(E) CC $<
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) $(DEF_OPT) -c -o $@ $<
src/upb_def.lo: src/upb_def.c
$(E) 'CC -fPIC' $<
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) $(DEF_OPT) -c -o $@ $< -fPIC
src/upb_decoder_x86.h: src/jit_debug_elf_file.h
src/upb_decoder_x86.h: src/upb_decoder_x86.dasc
$(E) DYNASM $<
$(Q) lua dynasm/dynasm.lua src/upb_decoder_x86.dasc > src/upb_decoder_x86.h
src/jit_debug_elf_file.o: src/jit_debug_elf_file.s
$(E) GAS $<
$(Q) gcc -c src/jit_debug_elf_file.s -o src/jit_debug_elf_file.o
src/jit_debug_elf_file.h: src/jit_debug_elf_file.o
$(E) XXD $<
$(Q) xxd -i src/jit_debug_elf_file.o > src/jit_debug_elf_file.h
# Function to expand a wildcard pattern recursively.
rwildcard=$(strip $(foreach d,$(wildcard $1*),$(call rwildcard,$d/,$2)$(filter $(subst *,%,$2),$d)))
# Regenerating the auto-generated files in src/.
src/descriptor.pb: src/descriptor.proto
@# TODO: replace with upbc
protoc src/descriptor.proto -osrc/descriptor.pb
descriptorgen: src/descriptor.pb src/upbc
@# Regenerate descriptor_const.h
./src/upbc -o src/descriptor src/descriptor.pb
src/upbc: src/upbc.c $(LIBUPB)
# Language extensions.
python: $(LIBUPB_PIC)
cd lang_ext/python && python setup.py build
# Tests. #######################################################################
tests/test.proto.pb: tests/test.proto
@# TODO: replace with upbc
protoc tests/test.proto -otests/test.proto.pb
SIMPLE_TESTS= \
tests/test_string \
tests/test_def \
tests/test_varint \
tests/tests
# tests/test_decoder \
upb_stream: all callbacks registered ahead-of-time. This is a significant change to the upb_stream protocol, and should hopefully be the last significant change. All callbacks are now registered ahead-of-time instead of having delegated callbacks registered at runtime, which makes it much easier to aggressively optimize ahead-of-time (like with a JIT). Other impacts of this change: - You no longer need to have loaded descriptor.proto as a upb_def to load other descriptors! This means the special-case code we used for bootstrapping is no longer necessary, and we no longer need to link the descriptor for descriptor.proto into upb. - A client can now register any upb_value as what will be delivered to their value callback, not just a upb_fielddef*. This should allow for other clients to get more bang out of the streaming decoder. This change unfortunately causes a bit of a performance regression -- I think largely due to highly suboptimal code that GCC generates when structs are returned by value. See: http://blog.reverberate.org/2011/03/19/when-a-compilers-slow-code-actually-bites-you/ On the other hand, once we have a JIT this should no longer matter. Performance numbers: plain.parsestream_googlemessage1.upb_table: 374 -> 396 (5.88) plain.parsestream_googlemessage2.upb_table: 616 -> 449 (-27.11) plain.parsetostruct_googlemessage1.upb_table_byref: 268 -> 269 (0.37) plain.parsetostruct_googlemessage1.upb_table_byval: 215 -> 204 (-5.12) plain.parsetostruct_googlemessage2.upb_table_byref: 307 -> 281 (-8.47) plain.parsetostruct_googlemessage2.upb_table_byval: 297 -> 272 (-8.42) omitfp.parsestream_googlemessage1.upb_table: 423 -> 410 (-3.07) omitfp.parsestream_googlemessage2.upb_table: 679 -> 483 (-28.87) omitfp.parsetostruct_googlemessage1.upb_table_byref: 287 -> 282 (-1.74) omitfp.parsetostruct_googlemessage1.upb_table_byval: 226 -> 219 (-3.10) omitfp.parsetostruct_googlemessage2.upb_table_byref: 315 -> 298 (-5.40) omitfp.parsetostruct_googlemessage2.upb_table_byval: 297 -> 287 (-3.37)
14 years ago
tests/test_stream \
SIMPLE_CXX_TESTS= \
tests/test_table
VARIADIC_TESTS= \
tests/t.test_vs_proto2.googlemessage1 \
tests/t.test_vs_proto2.googlemessage2 \
TESTS=$(SIMPLE_TESTS) $(SIMPLE_CXX_TESTS) $(VARIADIC_TESTS)
tests: $(TESTS)
$(TESTS): $(LIBUPB)
tests/tests: tests/test.proto.pb
$(SIMPLE_TESTS): % : %.c
$(E) CC $<
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) -o $@ $< $(LIBUPB)
VALGRIND=valgrind --leak-check=full --error-exitcode=1
test: tests
@echo Running all tests under valgrind.
@set -e # Abort on error.
@for test in $(TESTS); do \
if [ -x ./$$test ] ; then \
echo !!! $(VALGRIND) ./$$test; \
$(VALGRIND) ./$$test || exit 1; \
fi \
done; \
echo "All tests passed!"
tests/t.test_vs_proto2.googlemessage1 \
tests/t.test_vs_proto2.googlemessage2: \
tests/test_vs_proto2.cc $(LIBUPB) benchmarks/google_messages.proto.pb \
benchmarks/google_messages.pb.cc
$(E) CXX $< '(benchmarks::SpeedMessage1)'
$(Q) $(CXX) $(CXXFLAGS) $(CPPFLAGS) -o tests/t.test_vs_proto2.googlemessage1 $< \
-DMESSAGE_NAME=\"benchmarks.SpeedMessage1\" \
-DMESSAGE_DESCRIPTOR_FILE=\"../benchmarks/google_messages.proto.pb\" \
-DMESSAGE_FILE=\"../benchmarks/google_message1.dat\" \
-DMESSAGE_CIDENT="benchmarks::SpeedMessage1" \
-DMESSAGE_HFILE=\"../benchmarks/google_messages.pb.h\" \
benchmarks/google_messages.pb.cc -lprotobuf -lpthread $(LIBUPB)
$(E) CXX $< '(benchmarks::SpeedMessage2)'
$(Q) $(CXX) $(CXXFLAGS) $(CPPFLAGS) -o tests/t.test_vs_proto2.googlemessage2 $< \
-DMESSAGE_NAME=\"benchmarks.SpeedMessage2\" \
-DMESSAGE_DESCRIPTOR_FILE=\"../benchmarks/google_messages.proto.pb\" \
-DMESSAGE_FILE=\"../benchmarks/google_message2.dat\" \
-DMESSAGE_CIDENT="benchmarks::SpeedMessage2" \
-DMESSAGE_HFILE=\"../benchmarks/google_messages.pb.h\" \
benchmarks/google_messages.pb.cc -lprotobuf -lpthread $(LIBUPB)
tests/test_table: tests/test_table.cc
@# Includes <hash_set> which is a deprecated header.
$(E) CXX $<
$(Q) $(CXX) $(CXXFLAGS) $(CPPFLAGS) -Wno-deprecated -o $@ $< $(LIBUPB)
tests/tests: src/libupb.a
# Benchmarks. ##################################################################
# Benchmarks
UPB_BENCHMARKS=benchmarks/b.parsestream_googlemessage1.upb_table \
benchmarks/b.parsestream_googlemessage2.upb_table \
benchmarks/b.parsetostruct_googlemessage1.upb_table_byref \
benchmarks/b.parsetostruct_googlemessage2.upb_table_byref
BENCHMARKS=$(UPB_BENCHMARKS) \
benchmarks/b.parsetostruct_googlemessage1.proto2_table \
benchmarks/b.parsetostruct_googlemessage2.proto2_table \
benchmarks/b.parsetostruct_googlemessage1.proto2_compiled \
benchmarks/b.parsetostruct_googlemessage2.proto2_compiled
upb_benchmarks: $(UPB_BENCHMARKS)
benchmarks: $(BENCHMARKS)
benchmark:
@rm -f benchmarks/results
@rm -rf benchmarks/*.dSYM
@for test in benchmarks/b.* ; do ./$$test ; done
benchmarks/google_messages.proto.pb: benchmarks/google_messages.proto
@# TODO: replace with upbc.
protoc benchmarks/google_messages.proto -obenchmarks/google_messages.proto.pb
benchmarks/google_messages.pb.cc: benchmarks/google_messages.proto
protoc benchmarks/google_messages.proto --cpp_out=.
benchmarks/b.parsetostruct_googlemessage1.upb_table_byval \
benchmarks/b.parsetostruct_googlemessage1.upb_table_byref \
benchmarks/b.parsetostruct_googlemessage2.upb_table_byval \
benchmarks/b.parsetostruct_googlemessage2.upb_table_byref: \
benchmarks/parsetostruct.upb_table.c $(LIBUPB) benchmarks/google_messages.proto.pb
$(E) 'CC benchmarks/parsetostruct.upb_table.c (benchmarks.SpeedMessage1, byval)'
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) -o benchmarks/b.parsetostruct_googlemessage1.upb_table_byval $< \
-DMESSAGE_NAME=\"benchmarks.SpeedMessage1\" \
-DMESSAGE_DESCRIPTOR_FILE=\"google_messages.proto.pb\" \
-DMESSAGE_FILE=\"google_message1.dat\" \
-DBYREF=false $(LIBUPB)
$(E) 'CC benchmarks/parsetostruct.upb_table.c (benchmarks.SpeedMessage1, byref)'
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) -o benchmarks/b.parsetostruct_googlemessage1.upb_table_byref $< \
-DMESSAGE_NAME=\"benchmarks.SpeedMessage1\" \
-DMESSAGE_DESCRIPTOR_FILE=\"google_messages.proto.pb\" \
-DMESSAGE_FILE=\"google_message1.dat\" \
-DBYREF=true $(LIBUPB)
$(E) 'CC benchmarks/parsetostruct.upb_table.c (benchmarks.SpeedMessage2, byval)'
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) -o benchmarks/b.parsetostruct_googlemessage2.upb_table_byval $< \
-DMESSAGE_NAME=\"benchmarks.SpeedMessage2\" \
-DMESSAGE_DESCRIPTOR_FILE=\"google_messages.proto.pb\" \
-DMESSAGE_FILE=\"google_message2.dat\" \
-DBYREF=false $(LIBUPB)
$(E) 'CC benchmarks/parsetostruct.upb_table.c (benchmarks.SpeedMessage2, byref)'
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) -o benchmarks/b.parsetostruct_googlemessage2.upb_table_byref $< \
-DMESSAGE_NAME=\"benchmarks.SpeedMessage2\" \
-DMESSAGE_DESCRIPTOR_FILE=\"google_messages.proto.pb\" \
-DMESSAGE_FILE=\"google_message2.dat\" \
-DBYREF=true $(LIBUPB)
benchmarks/b.parsestream_googlemessage1.upb_table \
benchmarks/b.parsestream_googlemessage2.upb_table: \
benchmarks/parsestream.upb_table.c $(LIBUPB) benchmarks/google_messages.proto.pb
$(E) 'CC benchmarks/parsestream.upb_table.c (benchmarks.SpeedMessage1)'
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) -o benchmarks/b.parsestream_googlemessage1.upb_table $< \
-DMESSAGE_NAME=\"benchmarks.SpeedMessage1\" \
-DMESSAGE_DESCRIPTOR_FILE=\"google_messages.proto.pb\" \
-DMESSAGE_FILE=\"google_message1.dat\" \
$(LIBUPB)
$(E) 'CC benchmarks/parsestream.upb_table.c (benchmarks.SpeedMessage2)'
$(Q) $(CC) $(CFLAGS) $(CPPFLAGS) -o benchmarks/b.parsestream_googlemessage2.upb_table $< \
-DMESSAGE_NAME=\"benchmarks.SpeedMessage2\" \
-DMESSAGE_DESCRIPTOR_FILE=\"google_messages.proto.pb\" \
-DMESSAGE_FILE=\"google_message2.dat\" \
$(LIBUPB)
benchmarks/b.parsetostruct_googlemessage1.proto2_table \
benchmarks/b.parsetostruct_googlemessage2.proto2_table: \
benchmarks/parsetostruct.proto2_table.cc benchmarks/google_messages.pb.cc
$(E) 'CXX benchmarks/parsetostruct.proto2_table.cc (benchmarks.SpeedMessage1)'
$(Q) $(CXX) $(CXXFLAGS) $(CPPFLAGS) -o benchmarks/b.parsetostruct_googlemessage1.proto2_table $< \
-DMESSAGE_CIDENT="benchmarks::SpeedMessage1" \
-DMESSAGE_FILE=\"google_message1.dat\" \
-DMESSAGE_HFILE=\"google_messages.pb.h\" \
benchmarks/google_messages.pb.cc -lprotobuf -lpthread
$(E) 'CXX benchmarks/parsetostruct.proto2_table.cc (benchmarks.SpeedMessage2)'
$(Q) $(CXX) $(CXXFLAGS) $(CPPFLAGS) -o benchmarks/b.parsetostruct_googlemessage2.proto2_table $< \
-DMESSAGE_CIDENT="benchmarks::SpeedMessage2" \
-DMESSAGE_FILE=\"google_message2.dat\" \
-DMESSAGE_HFILE=\"google_messages.pb.h\" \
benchmarks/google_messages.pb.cc -lprotobuf -lpthread
benchmarks/b.parsetostruct_googlemessage1.proto2_compiled \
benchmarks/b.parsetostruct_googlemessage2.proto2_compiled: \
benchmarks/parsetostruct.proto2_compiled.cc \
benchmarks/parsetostruct.proto2_table.cc benchmarks/google_messages.pb.cc
$(E) 'CXX benchmarks/parsetostruct.proto2_compiled.cc (benchmarks.SpeedMessage1)'
$(Q) $(CXX) $(CXXFLAGS) $(CPPFLAGS) -o benchmarks/b.parsetostruct_googlemessage1.proto2_compiled $< \
-DMESSAGE_CIDENT="benchmarks::SpeedMessage1" \
-DMESSAGE_FILE=\"google_message1.dat\" \
-DMESSAGE_HFILE=\"google_messages.pb.h\" \
benchmarks/google_messages.pb.cc -lprotobuf -lpthread
$(E) 'CXX benchmarks/parsetostruct.proto2_compiled.cc (benchmarks.SpeedMessage2)'
$(Q) $(CXX) $(CXXFLAGS) $(CPPFLAGS) -o benchmarks/b.parsetostruct_googlemessage2.proto2_compiled $< \
-DMESSAGE_CIDENT="benchmarks::SpeedMessage2" \
-DMESSAGE_FILE=\"google_message2.dat\" \
-DMESSAGE_HFILE=\"google_messages.pb.h\" \
benchmarks/google_messages.pb.cc -lprotobuf -lpthread
# Lua extension ##################################################################
LUA_CPPFLAGS = $(strip $(shell pkg-config --silence-errors --cflags lua || pkg-config --silence-errors --cflags lua5.1 || echo '-I/usr/local/include'))
ifeq ($(shell uname), Darwin)
LUA_LDFLAGS = -undefined dynamic_lookup
else
LUA_LDFLAGS =
endif
LUAEXT=lang_ext/lua/upb.so
lua: $(LUAEXT)
lang_ext/lua/upb.so: lang_ext/lua/upb.c $(LIBUPB_PIC)
@echo CC lang_ext/lua/upb.c
@$(CC) $(CFLAGS) $(CPPFLAGS) $(LUA_CPPFLAGS) -fpic -shared -o $@ $< src/libupb_pic.a $(LUA_LDFLAGS)