This is an archive of the discontinued LLVM Phabricator instance.

[utils] Fix _run_benchmark in collect_and_build_with_pgo.py
ClosedPublic

Authored by MaskRay on Oct 27 2018, 1:55 PM.

Details

Summary

Also fix a FIXME in _build_stage1_clang: clang llvm-profdata profile are sufficient

Diff Detail

Repository
rL LLVM

Event Timeline

MaskRay created this revision.Oct 27 2018, 1:55 PM
MaskRay updated this revision to Diff 171411.Oct 27 2018, 2:00 PM
MaskRay removed a reviewer: gbiv.

.

LGTM assuming the _run_benchmark changes are undone.

Thank you!

utils/collect_and_build_with_pgo.py
57 ↗(On Diff #171411)

Please put this back; it should've been fixed by r345461. The intent was to run this in the instrumented clang's directory for extra coverage. Instrumentation makes the testing phase way slower, but given that this is already four clean builds of clang/llvm anyway, ...

If r345461 doesn't solve the problem you're seeing, more details would be appreciated. :)

This revision is now accepted and ready to land.Oct 27 2018, 3:25 PM
This revision was automatically updated to reflect the committed changes.