[PATCH] D111555: [LoopVectorize] Add strict reduction support for fmuladd intrinsic

[Rosie Sumpter via Phabricator via llvm-commits]

RosieSumpter created this revision.
Herald added subscribers: rogfer01, hiraditya.
RosieSumpter requested review of this revision.
Herald added subscribers: llvm-commits, vkmr.
Herald added a project: LLVM.

Enables LoopVectorize to handle strict reduction patterns involving
the llvm.fmuladd intrinsic. Previously such loops were not being
vectorized because only strict reduction patterns involving the FAdd
instruction were detected.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D111555

Files:
  llvm/include/llvm/Analysis/IVDescriptors.h
  llvm/lib/Analysis/IVDescriptors.cpp
  llvm/lib/Transforms/Utils/LoopUtils.cpp
  llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
  llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
  llvm/lib/Transforms/Vectorize/VPlan.cpp
  llvm/lib/Transforms/Vectorize/VPlan.h
  llvm/test/Transforms/LoopVectorize/AArch64/scalable-strict-fadd.ll
  llvm/test/Transforms/LoopVectorize/AArch64/strict-fadd.ll
  llvm/test/Transforms/LoopVectorize/reduction-inloop.ll
  llvm/unittests/Transforms/Vectorize/VPlanTest.cpp

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D111555.378687.patch
Type: text/x-patch
Size: 41443 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20211011/f6ad2a3c/attachment-0001.bin>