[PATCH] D97487: Removing sqrtd2 and sqrtf4 from list of vectorizable function with MASSV

Masoud Ataei via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Thu Feb 25 10:14:26 PST 2021


masoud.ataei created this revision.
masoud.ataei added reviewers: pjeeva01, Whitney.
masoud.ataei added a project: PowerPC.
Herald added a subscriber: nemanjai.
masoud.ataei requested review of this revision.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

Under -O3 and -Ofast, the MASSV conversion prevents the sqrt call to be inlined. 
Inline sqrt is faster than MASSV call on leppc.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D97487

Files:
  llvm/include/llvm/Analysis/VecFuncs.def
  llvm/test/CodeGen/PowerPC/lower-massv.ll
  llvm/test/Transforms/LoopVectorize/PowerPC/massv-calls.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D97487.326435.patch
Type: text/x-patch
Size: 4017 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20210225/4f09f2f5/attachment.bin>


More information about the llvm-commits mailing list