This is an archive of the discontinued LLVM Phabricator instance.

[RISCV][VP] Lower mask vector VP AND/OR/XOR to RVV instructions
ClosedPublic

Authored by victor-eds on Dec 8 2021, 6:15 AM.

Details

Summary

For fixed and scalable vectors, each intrinsic x is lowered to vmx.mm,
dropping the mask, which is safe to do as masked-off elements are
undef anyway.

Diff Detail

Event Timeline

victor-eds created this revision.Dec 8 2021, 6:15 AM
victor-eds requested review of this revision.Dec 8 2021, 6:15 AM
Herald added a project: Restricted Project. · View Herald TranscriptDec 8 2021, 6:15 AM

Format file

victor-eds edited reviewers, added: rogfer01; removed: efocht.Dec 14 2021, 12:32 AM

LGTM but I'll wait for an "external" approval to formalize it.

This revision is now accepted and ready to land.Dec 17 2021, 9:19 AM

Thanks. I don't have permissions to merge on main yet, so I need someone to land this patch for me now that it's been accepted. @craig.topper