Use rocprofiler-sdk for RCCL-API tracing (#126)

- Add support for RCCL API tracing through rocprofiler-sdk.
- Refactored the comm_data code to use the SDK RCCL_API callbacks.
- Add a runtime version check for SDK to gate callback enablement, rather than just the compile-time check.
- Fixed: SAMPLING_TIMEOUT was not being handled correctly in add_test.

[ROCm/rocprofiler-systems commit: af77d93f75]
Tento commit je obsažen v:
David Galiffi
2025-06-06 11:36:17 -04:00
odevzdal GitHub
rodič f718bd907c
revize c7c3c3f97e
31 změnil soubory, kde provedl 350 přidání a 1252 odebrání
@@ -834,10 +834,6 @@ parse_args(int argc, char** argv, std::vector<char*>& _env,
(defined(ROCPROFSYS_USE_MPI_HEADERS) && ROCPROFSYS_USE_MPI_HEADERS > 0)
add_default_env(_env, "ROCPROFSYS_USE_MPIP", true);
#endif
#if defined(ROCPROFSYS_USE_RCCL) && ROCPROFSYS_USE_RCCL > 0
add_default_env(_env, "ROCPROFSYS_USE_RCCLP", true);
#endif
}
_fill("ROCPROFSYS_CAUSAL_BINARY_EXCLUDE", _binary_excludes, _generate_configs);