[PATCH] D130618: [AArch64][LoopVectorize] Enable tail-folding of simple loops on neoverse-v1

David Sherwood via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Fri May 12 06:38:51 PDT 2023


david-arm updated this revision to Diff 521631.
david-arm edited the summary of this revision.
david-arm added a comment.

- Bumped up the instruction threshold from 10 -> 15 in order to reduce the risk of causing possible regressions for tight loops. This still leaves us with a 7% win for x264, but there is also no longer a regression for parest because the loops are now too small to be tail-folded. This is more by chance however, since the problem with tail-folding for parest seems unrelated to loop size and due to problems with code quality.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D130618/new/

https://reviews.llvm.org/D130618

Files:
  llvm/lib/Target/AArch64/AArch64Subtarget.cpp
  llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
  llvm/test/Transforms/LoopVectorize/AArch64/sve-epilog-vect-vscale-tune.ll
  llvm/test/Transforms/LoopVectorize/AArch64/sve-tail-folding-option.ll
  llvm/test/Transforms/LoopVectorize/AArch64/sve-tail-folding-overflow-checks.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D130618.521631.patch
Type: text/x-patch
Size: 10898 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20230512/a497acc9/attachment.bin>


More information about the llvm-commits mailing list