While doing some benchmark performance comparison analysis I noticed that
one version was performing noticeably worse due to additional calls to
memcpy() being generated. These statistics have been useful when looking
into memcpy() inlining behaviour for the CHERI LLVM fork, and are hopefully
also useful upstream.
Details
Details
- Reviewers
- None
Diff Detail
Diff Detail
- Repository
- rG LLVM Github Monorepo
Unit Tests
Unit Tests
Time | Test | |
---|---|---|
60,060 ms | x64 debian > libomp.worksharing/for::omp_for_schedule_runtime.c |