* use max_frame_size to control encrypted frame sizes on the sender
* Add comment
* adding logic to set max_frame_size in chttp2 transport and protecting it under a flag
* fix typo
* fix review comments
* set max frame size usage in endpoint_tests
* update endpoint_tests
* adding an interception layer to secure_endpoint_test
* add comments
* reverting some mistaken changes
* Automated change: Fix sanity tests
* try increasing deadline to check if msan passes
* Automated change: Fix sanity tests
Co-authored-by: Vignesh2208 <Vignesh2208@users.noreply.github.com>
* adding a min progress size argument to grpc_endpoint_read
* fix missing argument error
* adding a static_cast
* reverting changes in tcp_posix.cc
* add missing changes to CFStreamEndpointTests.mm
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).
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%)
src/core. exec_ctx is now a thread_local pointer of type ExecCtx instead of
grpc_exec_ctx which is initialized whenever ExecCtx is instantiated. ExecCtx
also keeps track of the previous exec_ctx so that nesting of exec_ctx is
allowed. This means that there is only one exec_ctx being used at any
time. Also, grpc_exec_ctx_finish is called in the destructor of the
object, and the previous exec_ctx is restored to avoid breaking current
functionality. The code still explicitly calls grpc_exec_ctx_finish
because removing all such instances causes the code to break.
- make closures know where they should be executed (eg, on a workqueue,
or a combiner, or on an exec_ctx)
- this allows removal of a large number of trampoline functions that
were appearing whenever we used combiners, and should allow for a much
easier interface to combiner locks