[PATCH] D147945: [AArch64] Add a basic pattern for FACGE and FACGT
Dave Green via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Wed Apr 12 10:47:56 PDT 2023
dmgreen added a comment.
Yeah I had seen the scalar instruction. Scalar usually uses `fcmp` instructions that set flags though, they don't set lanes to all-ones like vector instructions. We could use the scalar vacgt/vacge for the pattern you describe (so long as the result is moved from the vpr to a gpr), but I'm not sure that would come up a lot and should probably at least be handled as a separate patch.
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D147945/new/
https://reviews.llvm.org/D147945
More information about the llvm-commits
mailing list