This is an archive of the discontinued LLVM Phabricator instance.

[AMDGPU][CostModel] Improve cost estimation for fused {fadd|fsub}(a,fmul(b,c))
ClosedPublic

Authored by dfukalov on Jul 15 2020, 4:25 PM.

Details

Summary

If result of fmul(b,c) has one use, in almost all cases (except denormals are
IEEE) the pair of operations will be fused in one fma/mad/mac/etc.

Diff Detail

Event Timeline

dfukalov created this revision.Jul 15 2020, 4:25 PM
Herald added a project: Restricted Project. · View Herald TranscriptJul 15 2020, 4:25 PM
This revision is now accepted and ready to land.Jul 15 2020, 4:28 PM
arsenm added inline comments.Jul 15 2020, 4:29 PM
llvm/test/Analysis/CostModel/AMDGPU/fused_costs.ll
49

Extra newlines

dfukalov updated this revision to Diff 278338.Jul 15 2020, 5:05 PM

removed extra new lines

This revision was automatically updated to reflect the committed changes.
dfukalov marked an inline comment as done.