* Save ENOBUFS errno correctly in tcp_posix for subsequent handling

* remove log statement

* Tryonly running fetch_build_eggs once

* Run pip freeze

* Use the correct pip

* Try explicitly printing stacktrace

* Try to pin/upgrade the wheel package

* Try running only 3.8

* Show the initial state of the egg directory

* Whoops

* More permissive

* What does this button do?

* Try just 3.9

* Try 3.10

* Back to 3.8

* Try pre-installing protobuf

* Clean up

* Clean up better

* update win RBE docker image and toolchain

and switch win RBE builds to using it.

* delete legacy RBE windows toolchain config

* switching to VS2019 makes #28544 obsolete

* replace darwin checks in extconf.rb to exclude TruffleRuby

* inherit RANLIB and STRIP from RbConfig, set LDXX

* enable overriding ranlib command in top-level makefile

* ensure the -no_warning_for_no_symbols flag is only used with Apple's ranlib

* don't embed openssl & zlib on truffleruby

* add RbConfig's cppflag to CPPFLAGS when using TruffleRuby

* this ensure the paths to find the system's OpenSSL are set up correctly with TruffleRuby (includes being able to find an OpenSSL installed via Homebrew etc)

* don't statically link standard libraries on Linux with Truffleruby

* This does not work when compiling to bitcode.

* Prefer SIGTERM to SIGQUIT for graceful shutdown in examples

* Overriding SIGQUIT is suboptimal, for example on JVM where it is very
  useful to dump the thread stacktraces.

* Keep the rb_tr_abi_version symbol for TruffleRuby in grpc_c.so

* Otherwise TruffleRuby cannot verify the ABI version is correct.
* See ...

* Revert "Revert "Reland bytestream removal (#29911)" (#29964)"

This reverts commit e6c6840d.

* initial fc fuzzer

* fixes

* add rq to fc fuzzer

* fleshing things out

* Automated change: Fix sanity tests

* cleanup

* send with payload

* ensure if no reader no flow control tokens are granted

* remove some public methods

* remove bogus benchmarks

* account for pending size

* Automated change: Fix sanity tests

* Automated change: Fix sanity tests

* better logic

* Automated change: Fix sanity tests

* fix

* fixes

* fuzz pending size

* Automated change: Fix sanity tests

* fix

* Automated change: Fix sanity tests

* huh

* increase too short timeout

* review feedback

* review feedback

* fix u32 overflow

* fix

* robustness fixes for channelz_servicer_test

* fix

* Automated change: Fix sanity tests

* fix

* fix

* fix

* Automated change: Fix s...

* Add regular tests for PSM benchmarks CI

* Remove temporary EventEngine initialization function

* Automated change: Fix sanity tests

Co-authored-by: drfloob <drfloob@users.noreply.github.com>

* Fix leak on ares resolver test

The pollset was not being properly shutdown. ASAN was failing 100% on
resolve_address_using_ares_resolver_test for a month (not identified in
CI).

* DoNothgin -> nullptr

A minor refactoring.

Added a couple of tests which run the baseline_test with all released
bootstrap generator versions on client and server. These tests will be
run on a continuous integration environment with gRPC servers and
clients built using the latest released version of gRPC in one selected
language.

* Add supported Node version ranges in xDS k8s url_map tests

This adds is_supported implementations for most of the url_map tests that didn't
already have them. The exception is metadata_filter_test because it doesn't use
any specific client features.

* Fix formatting

* Improve timeout test check order

* BinderTransport: Fix wire reader/writer issue with NDK

NDK sometimes call registered `AIBinder_Class_onTransact`
callback while we call its `AIBinder_transact`.

Sometimes this behavior happens consistently sometimes it
don't happen at all. Likely related to the short circuit
behavior happens when both gRPC server and gRPC client are in the same
process.

This commit also fixes some other existing issue in wire writer.

* Use atomic<int64_t> instead of atomic_int64_t

atomic_int64_t is not defined in older compilers.

* Lower log level to avoid excessive test output

* Add missing break

dropepd this while addressing reivew comments

* Use absl::make_unique

* Fix printf portability with PRId64

* clang-format

* generate projects.

* fix

* fix

* [event_engine] Fix initialization

* Automated change: Fix sanity tests

Co-authored-by: ctiller <ctiller@users.noreply.github.com>

* better api for ee up/down

* fixes

* mac fix

* review feedback

* nuke bm_cq_multiple_threads

* Revert "nuke bm_cq_multiple_threads"

This reverts commit dcda194a42729263a574b75df23e82eaf6567cd2.

* bleh-revert-fix

* fix tsan race in stats

* fix

* handle spurious wakeups

* disable on mac, win

* Automated change: Fix sanity tests

* Update timer_test.cc

Co-authored-by: ctiller <ctiller@users.noreply.github.com>

* mac fix

* Automated change: Fix sanity tests

Co-authored-by: ctiller <ctiller@users.noreply.github.com>

* Set uses polling to true for oracle_event_engine_posix_test

* add a TODO

* Automated change: Fix sanity tests

Co-authored-by: Vignesh2208 <Vignesh2208@users.noreply.github.com>

* [bloat] Prefer std::vector to absl::InlinedVector

* fix

* fix

* fix

* fix

* fixes

* fix

* fixes

* Use gRPC_INSTALL_LIBDIR for pkgconfig files

Fixes #25635.

If grpc libraries are installed in <prefix>/lib64, then .pc files should
be installed in <prefix>/lib64/pkgconfig. Before this commit, they were
always installed in <prefix>/lib/pkgconfig.

* Re-generate projects

This reverts commit 0979955f.

There's been an uptick of races against g_event_engine in iomgr (this is not the new event engine but rather the old one that is a component of iomgr on posix). It's of genuinely little interest if this shuts down or not, so instead, just initialize it once at the first grpc_init, and then leave it for the lifetime of the application.

Should solve the majority of our TSAN flakes on Linux, and *I suspect* many of our other flakiness problems at head.

As we move to event engine over the next small amount of time we can transition to a better init/shutdown story for that.

* Revert "Revert "Enable GRPC_ERROR_IS_ABSEIL_STATUS (#29869)" (#30031)"

This reverts commit 2d0d1775.

* Fix the test

Revert "Revert "Creating a posix oracle event engine and a suite of event engine client tests"" (#30060)

* Revert "Revert "Creating a posix oracle event engine and a suite of event engine client tests (#29714)" (#30042)"

This reverts commit 1630efd8.

* fix typos

* weighted_target and RLS: delegate to child picker on error

* security handshaker: include security connector type in error message

* update test

* fix sanity

* fix crash

* Automated change: Fix sanity tests

Co-authored-by: markdroth <markdroth@users.noreply.github.com>

* Revert "Revert "Tcp connect cancellation implementation for posix (#29976)" (#30037)"

This reverts commit 13dfc6b4.

* add tcp global init in iomgr_posix_cfstream