Chiebot-Mirror/grpc - grpc - Gitea: Git with a cup of tea

Commit Graph

Author	SHA1	Message	Date
Craig Tiller	d3e5803cb2	Move HPACK parser to C++ (#26689 ) This is a fairly low effort migration of the current codebase into a C++ class, instead of free standing C code. It builds upon #26657 as a necessary first step. I've tried to minimize any changes to semantics or logic in this change, except where required to get a minimal amount of encapsulation - which is the major aim of this change. A future change in this series will buffer slices until all HPACK headers are in memory for a stream prior to decoding -- it's important to have an encapsulated API to the parser before doing so however (hence this CL). The next change after that will be an almost complete rewrite of the parsing functionality -- since we'll have the total set of header bytes, we'll no longer need to support suspending decoding at arbitrary points. This will allow us to move to a simple recursive descent parser, eliminate a bunch of indirection in this code, and end up in a much more malleable place for when we start doing metadata API changes. (we likely also end up with some good performance wins!)	3 years ago
Mark D. Roth	4c40ee3f78	move parse_address and sockaddr_utils out of iomgr directory (#26077 ) * move parse_address and sockaddr_utils out of iomgr directory * clang-format * fix python * fix for import * fix build file for import	4 years ago
Esun Kim	460898f11c	Migrated from grpc_error_string to grpc_error_std_string (#26068 ) * Added grpc_error_std_string * Migrated to grpc_error_std_string	4 years ago
Esun Kim	ca945a58e9	Introduced grpc_error_handle (#25902 ) - Define grpc_error_handle - Replace grpc_error* with grpc_error_handle	4 years ago
Yash Tibrewal	e9de13e6ad	Adding test to make sure that http2 transport gets cleaned up (#25714 ) * Add test to make sure that transports get destroyed * Reviewer comments	4 years ago
AJ Heller	0e3a02e903	Revert "Remove the `urgent` argument from iomgr tcp read API (#25494 )" (#25592 ) This reverts commit `a3398f9`. Justification: see b/181367644. tl;dr: assuming urgent==false does not hold in all situations.	4 years ago
AJ Heller	a3398f924c	Remove the `urgent` argument from iomgr tcp read API (#25494 ) The urgent argument is a platform-specific flag that leaked into the (ideally) platform-independent HTTP/2 transport layer. In an effort to clean up the cross-platform API surface, it would be helpful if we can remove this argument from the TCP Read api without losing the performance optimization that was introduced along with it (see #18240).	4 years ago
Yash Tibrewal	f4e55b85bc	Replace disrespectful term	4 years ago
AJ Heller	880a854f0e	Convert URI parser to C++ #24701	4 years ago
Esun Kim	e0d8c498a6	C++ cast for others	4 years ago
Esun Kim	2ff84ed8c2	C++ cast for void*	4 years ago
Yash Tibrewal	6095b67d6c	Reduce BdpPingNotSentWithoutReceiveSideActivity flakiness	4 years ago
Mark D. Roth	27e1c1c56d	clang-tidy: enable google-readability-braces-around-statements	4 years ago
Yash Tibrewal	b555343dd3	Revert "Revert "Deprecate GRPC_ARG_HTTP2_MIN_SENT_PING_INTERVAL_WITHOUT_DATA_MS""	4 years ago
Esun Kim	6ae8463500	Fix by modernize-make-unique	4 years ago
Yash Tibrewal	a9fd34a81d	Revert "Deprecate GRPC_ARG_HTTP2_MIN_SENT_PING_INTERVAL_WITHOUT_DATA_MS"	4 years ago
Yash Tibrewal	e0c4475db0	Deprecate GRPC_ARG_HTTP2_MIN_SENT_PING_INTERVAL_WITHOUT_DATA_MS and make it non-effective	4 years ago
Yash Tibrewal	23cb073a93	Initiate HTTP/2 writes on BDP pings too	4 years ago
Alexander Polcyn	46d4a27cb6	Add a test which reproduces the original crash that is fixed by https://github.com/grpc/grpc/pull/23984	4 years ago
Yash Tibrewal	4a03ce5a14	Remove grpc_channel_ping from surface API	4 years ago
Yash Tibrewal	09904691e0	Do not send BDP pings if there is no receive side activity.	4 years ago
Yash Tibrewal	7019c2ca4b	Long timeout for too_many_pings_test	4 years ago
Mark D. Roth	bd73844208	Move parse_address library from client_channel to iomgr.	4 years ago
Yash Tibrewal	c62d2b483c	Test renaming	4 years ago
Yash Tibrewal	1618d21796	Add roundrobin test and reviewer comments	4 years ago
Yash Tibrewal	cf0b46c4d6	Reviewer comments	4 years ago
Yash Tibrewal	7100ee3bfb	Add keepalive throttling	4 years ago
Yash Tibrewal	2c5291ff71	Plumb absl::Status through connectivity state notifiers	4 years ago
Mark D. Roth	9404f66f70	Replace most uses of gpr_asprintf() with absl calls.	4 years ago
Yash Tibrewal	f8b7fef753	Fix connect deadline issue in settings_timeout_test	4 years ago
Yash Tibrewal	234d64859f	Add some additional comments	5 years ago
Yash Tibrewal	722f1e937a	Reviewer comments	5 years ago
Yash Tibrewal	f41694ad64	Add the missing reset for ping clocks to avoid mistakenly sending GOAWAY frames due to 'too_many_pings'	5 years ago
Brian O'Connor	d38f7cfe5e	Fixing bug with END_STREAM if header has continuations - The HEADER frame should get the END_STREAM flag per the HTTP/2 spec; the old code put END_STREAM on the last CONTINUATION frame. - Removing deprecated parameter, is_last_in_stream, from finish_frame(); it has been superseded by the is_end_of_stream member of the framer_state struct. - Adding some gRPC frame validation tests to hpack_encoder_test.cc, and explicting checking that the END_STREAM flag is not on the CONTINUATION frame. fixes #21436	5 years ago
Donna Dionne	46439bcb0f	Fixing the settings_timeout_test flake by reducing pollset_work deadline Intermittently (when we run the test 100 times) we see that at times we are not notified about successful socket connection even within 1 second. The pollset_work deadline is set to 1 second, so when our connection timeout is also set to 1 second, we only had 1 poll. And if that poll didn't let us know about the socket, we didn't get another chance. Reducing the pollset_work deadline to 100ms, now the test runs faster and pass consistently: result for 500 runs: https://source.cloud.google.com/results/invocations/dd1b38f4-0f28-470b-a78b-b485445e44f2/targets/%2F%2Ftest%2Fcore%2Ftransport%2Fchttp2:settings_timeout_test/tests	5 years ago
Jan Tattermusch	cbb0ef2717	try marking some offenders as flaky	5 years ago
yang-g	5bc57b64fe	InitGoogleTest before grpc_init	5 years ago
Esun Kim	74d2fb842e	Timeout for shutdown!	5 years ago
Esun Kim	9ae81e1d3f	Make test code use TestEnvironment	5 years ago
Lidi Zheng	c34bef4ca4	Format all Bazel files with buildifier	5 years ago
Vijay Pai	2477b966e2	Remove unused-parameter warnings, round 2 (17 of 19)	5 years ago
Vijay Pai	17c8c4c4f2	Remove unused parameter warning (13 of 20)	5 years ago
Jan Tattermusch	8e79705d29	disable fuzzers on windows	5 years ago
Jan Tattermusch	6ebc1c616f	set uses_polling=False for some bazel tests	5 years ago
Arjun Roy	b46e3668d3	s/branch/tail_call/ for CH2 on_hdr(). on_hdr() checks if a void-return function pointer is null before jumping to it. If it is null, it returns an error; else it executes that function and returns success. This change converts the void-returning function to one that returns a grpc_error* and thus saves a branch in on_hdr() (since we're branching once by following the function pointer anyways, we're effectively coalescing these two branches).	5 years ago
Hope Casey-Allen	59564ebd96	Fix warnings to unblock gcc8 support	5 years ago
Arjun Roy	b2cda1e185	Reduced ops for grpc_chttp2_stream_map_find(). Several asserts in grpc_chttp2_stream_map_find() can be converted to debug asserts. This PR also templatizes the internal find() method to have it be strict in the delete case (which saves some branches).	5 years ago
Yash Tibrewal	2d498d2b6c	Add a few tests for ContextList and BufferList. Also, check that timestamps is non-null before modifying it.	5 years ago
Arjun Roy	b1d73a01f1	Removed duplicate static table from hpack table. Removed an or instruction for every usage of static grpc metadata. Inlined hpack table lookups for static metadata. This leads to faster hpack parser creation: BM_HpackParserInitDestroy 5.32µs ± 1% 0.06µs ± 1% -98.91% (p=0.000 n=18+19) And slightly faster parsing: BM_HpackParserParseHeader<RepresentativeClientInitialMetadata, OnInitialHeader> 456ns ± 1% 435ns ± 1% -4.74% (p=0.000 n=18+19) BM_HpackParserParseHeader<MoreRepresentativeClientInitialMetadata, OnInitialHeader> 1.06µs ± 2% 1.04µs ± 2% -1.82% (p=0.000 n=19+20) It also yields a slight (0.5 - 1.0 microsecond) reduction in CPU time for fullstack unary pingpong: BM_UnaryPingPong<TCP, NoOpMutator, NoOpMutator>/0/512 [polls/iter:3.0001 ] 23.9µs ± 2% 23.0µs ± 1% -3.63% (p=0.002 n=6+6) BM_UnaryPingPong<TCP, NoOpMutator, NoOpMutator>/0/32768 [polls/iter:3.00015 ] 35.1µs ± 1% 34.2µs ± 1% -2.57% (p=0.036 n=5+3) BM_UnaryPingPong<MinTCP, NoOpMutator, NoOpMutator>/8/0 [polls/iter:3.00011 ] 21.7µs ± 3% 21.2µs ± 2% -2.44% (p=0.017 n=6+5)	5 years ago
Soheil Hassas Yeganeh	18b19105f2	Implement TCP_INQ for gRPC in Linux TCP_INQ is a socket option we added to Linux to report pending bytes on the socket as a control message. Using TCP_INQ we can accurately decide whether to continue read or not. Add an urgent parameter, when we do not want to wait for EPOLLIN. This commit improves the latency of 1 RPC unary (minimal benchmark) significantly: Before: l_50: 61.3584984733 l_90: 94.8328711277 l_99: 126.211351174 l_999: 158.722406029 After: l_50: 51.3546011488 (-16%) l_90: 72.3420731581 (-23%) l_99: 103.280218974 (-18%) l_999: 130.905689996 (-17%)	6 years ago

1 2 3 4 5 ...

266 Commits (48ce79f7e5db829da87d7cee2c9b0ddeb05d2af1)