Page MenuHomePhabricator

Optimize performance of Thread sanitizer memory access functions
ClosedPublic

Authored by yuri on Thu, Feb 7, 3:15 AM.

Details

Reviewers
dvyukov
Summary
  • Manually unwind code in MemoryAccessImpl1() because clang do not optimize it
  • Check for .rodata section only in read operations
  • Place LIKELY/UNLIKELY on fast paths

Diff Detail

Event Timeline

yuri created this revision.Thu, Feb 7, 3:15 AM
Herald added a project: Restricted Project. · View Herald TranscriptThu, Feb 7, 3:15 AM
Herald added subscribers: Restricted Project, llvm-commits, jfb, kubamracek. · View Herald Transcript
dvyukov accepted this revision.Thu, Feb 7, 4:40 AM
This revision is now accepted and ready to land.Thu, Feb 7, 4:40 AM