[PATCH] D96023: [LoopVectorize] Calculate Max Feasible Scalable VF.

[Sander de Smalen via Phabricator via llvm-commits]

sdesmalen created this revision.
Herald added a subscriber: hiraditya.
sdesmalen requested review of this revision.
Herald added a project: LLVM.

This patch adds a 'enable-scalable-vectorizaton' flag to the
loopvectorizer, which for now doesn't change the vectorization, but it
let's computeMaxVF call computeFeasibleMaxVF with 'ComputeScalableMaxVF =
true', which prints the output to a DEBUG ostream. This is then verified
by the corresponding test.

This patch is a preparatory patch with the ultimate goal of making
`computeMaxVF()` return both a max fixed VF and a max scalable VF,
so that `selectVectorizationFactor()` can pick the most cost-effective
vectorization factor.


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D96023

Files:
  llvm/include/llvm/Analysis/TargetTransformInfo.h
  llvm/include/llvm/Analysis/TargetTransformInfoImpl.h
  llvm/lib/Analysis/TargetTransformInfo.cpp
  llvm/lib/Target/AArch64/AArch64TargetTransformInfo.h
  llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
  llvm/test/Transforms/LoopVectorize/AArch64/scalable-vf-analysis.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D96023.321393.patch
Type: text/x-patch
Size: 12982 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20210204/f45c7890/attachment.bin>