[PATCH] D74936: AMDGPU/GlobalISel: Legalize G_FPOW

Jay Foad via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Fri Feb 21 01:35:33 PST 2020


foad accepted this revision.
foad added a comment.
This revision is now accepted and ready to land.

LGTM.

As discussed off-list, perhaps we could use a normal fp16 mul instead of converting to fp32 for a legacy mul, at least for graphics.

As an extension it would be nice to catch cases like pow(3,y) and constant-fold the log(3). SelectionDAG doesn't do this.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D74936/new/

https://reviews.llvm.org/D74936





More information about the llvm-commits mailing list