This is an archive of the discontinued LLVM Phabricator instance.

AMDGPU/GlobalISel: Use more wide vector load/stores
ClosedPublic

Authored by arsenm on Jan 21 2020, 9:29 AM.

Details

Reviewers
nhaehnle
kerbowa
Summary

This improves the type breakdown for some large vectors. For example,
we now get a <4 x s32> and s32 store instead of 5 s32 stores for
<5 x s32>.

Diff Detail