Page MenuHomePhabricator

AVX1 : Enable vector masked_load/store to AVX1
ClosedPublic

Authored by igorb on Jan 25 2016, 1:33 AM.

Details

Summary

AVX1 : Enable vector masked_load/store to AVX1.
Use AVX1 FP instructions (vmaskmovps/pd) in place of the AVX2 int instructions (vpmaskmovd/q).

Diff Detail

Repository
rL LLVM

Event Timeline

igorb updated this revision to Diff 45845.Jan 25 2016, 1:33 AM
igorb retitled this revision from to AVX1 : Enable vector masked_load/store to AVX1.
igorb updated this object.
igorb added a reviewer: delena.
igorb set the repository for this revision to rL LLVM.
igorb added a subscriber: llvm-commits.
delena added inline comments.Jan 25 2016, 1:55 AM
lib/Target/X86/X86InstrSSE.td
8740 ↗(On Diff #45845)

Could you, please, change it to VPBLEND* for integer vectors?

delena accepted this revision.Jan 25 2016, 2:16 AM
delena edited edge metadata.

LGTM

This revision is now accepted and ready to land.Jan 25 2016, 2:16 AM
This revision was automatically updated to reflect the committed changes.