This is an archive of the discontinued LLVM Phabricator instance.

AMDGPU: Correct constants used in fast math log expansion
ClosedPublic

Authored by arsenm on Jun 12 2023, 1:31 PM.

Details

Reviewers
foad
rampitec
Pierre-vh
cdevadas
Group Reviewers
Restricted Project
Summary

The division between float constants was done with less
precision. Performing the divide in double and truncating to float
provides the same value as used in the library fast math expansion.

Diff Detail

Event Timeline

arsenm created this revision.Jun 12 2023, 1:31 PM
Herald added a project: Restricted Project. · View Herald TranscriptJun 12 2023, 1:31 PM
arsenm requested review of this revision.Jun 12 2023, 1:32 PM
Herald added a project: Restricted Project. · View Herald TranscriptJun 12 2023, 1:32 PM
Herald added a subscriber: wdng. · View Herald Transcript
cdevadas accepted this revision.Jun 12 2023, 1:43 PM
This revision is now accepted and ready to land.Jun 12 2023, 1:43 PM
arsenm updated this revision to Diff 530660.Jun 12 2023, 1:50 PM

Also globalisel

cdevadas accepted this revision.Jun 12 2023, 1:53 PM