Commit Graph
10 Commits
Author SHA1 Message Date
J. Nick Koston 44e1a86819 [benchmark] Use TCP loopback sockets and correct message type
Use real TCP sockets instead of AF_UNIX socketpair so TCP_NODELAY
succeeds during init() and the benchmark exercises the full write
path. Replace hardcoded message type 38 with SensorStateResponse::MESSAGE_TYPE.
2026-03-21 09:02:17 -10:00
J. Nick Koston 135c599561 [benchmark] Pre-init APIBuffer to 1460 bytes in plaintext frame benchmarks
Avoid benchmarking heap allocation by pre-reserving the buffer
to typical TCP MSS size and reusing it across iterations, matching
real-world usage where the buffer persists across writes.
2026-03-21 08:48:15 -10:00
J. Nick Koston b9679f6caf Merge branch 'dev' into benchmark/plaintext-frame-v2 2026-03-21 08:09:51 -10:00
J. Nick Koston 1920d8a887 [benchmark] Add noise encryption benchmarks (#15037) 2026-03-20 17:35:17 -10:00
J. Nick Koston 4bf84b6c61 [benchmark] Enable codegen for api dependency chain (socket, network, mdns) 2026-03-20 16:21:31 -10:00
J. Nick Koston cf7b63da60 [benchmark] Add both USE_API_PLAINTEXT and USE_API_NOISE defines 2026-03-20 16:07:30 -10:00
J. Nick Koston bc2654379b [benchmark] Fix api init: use build flag instead of enable_codegen 2026-03-20 16:05:25 -10:00
J. Nick Koston 328ef819e6 [benchmark] Fix plaintext frame benchmark build: enable api codegen 2026-03-20 16:02:44 -10:00
J. Nick Koston 9bd5d91e61 [benchmark] Add plaintext API frame write benchmarks 2026-03-20 15:53:58 -10:00
J. Nick Koston 1670f04a87 [core] Add CodSpeed C++ benchmarks for protobuf, main loop, and helpers (#14878) 2026-03-17 12:29:38 -10:00