Add benchmarks for inproc & chaotic good transports. Initial inproc results: ``` ----------------------------------------------------------------------------------------------------- Benchmark Time CPU Iterations ----------------------------------------------------------------------------------------------------- BM_UnaryWithSpawnPerEnd<TransportFixture<InprocTraits>> 2017 ns 2017 ns 6999359 BM_UnaryWithSpawnPerOp<TransportFixture<InprocTraits>> 2201 ns 2201 ns 6343162 BM_ClientToServerStreaming<TransportFixture<InprocTraits>> 388 ns 388 ns 36472157 ``` Initial chaotic good results: ``` ---------------------------------------------------------------------------------------------------------- Benchmark Time CPU Iterations ---------------------------------------------------------------------------------------------------------- BM_UnaryWithSpawnPerEnd<TransportFixture<ChaoticGoodTraits>> 87201 ns 36978 ns 392538 BM_UnaryWithSpawnPerOp<TransportFixture<ChaoticGoodTraits>> 86533 ns 36676 ns 369698 BM_ClientToServerStreaming<TransportFixture<ChaoticGoodTraits>> 20843 ns 10253 ns 1000000 ``` (chaotic good is jumping threads all over the place here... we'll want to address that as future work) Closes #37054 COPYBARA_INTEGRATE_REVIEW=https://github.com/grpc/grpc/pull/37054 from ctiller:moar-transporty 27511cd483 PiperOrigin-RevId: 647319669
27511cd483
fbd8ebe7cd
Craig Tiller