[all-commits] [llvm/llvm-project] 636efd: [SVE][LoopVectorize] Add option to disable tail-fo...

[david-arm via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 636efd2e3508f4a39717cee75c1255a48acb8f4e
      https://github.com/llvm/llvm-project/commit/636efd2e3508f4a39717cee75c1255a48acb8f4e
  Author: David Sherwood <david.sherwood at arm.com>
  Date:   2023-03-27 (Mon, 27 Mar 2023)

  Changed paths:
    M llvm/include/llvm/Transforms/Vectorize/LoopVectorizationLegality.h
    M llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
    M llvm/test/Transforms/LoopVectorize/AArch64/sve-tail-folding-option.ll

  Log Message:
  -----------
  [SVE][LoopVectorize] Add option to disable tail-folding for reverse loops

If we use tail-folding for reverse loops that contain loads
and stores then we will need to reverse the loop predicate.
This patch adds a new 'reverse' sve-tail-folding option and
ensures they are not considered 'simple'.

I did this by adding a function called
containsDecreasingPointers to AArch64TargetTransformInfo.cpp
that searches all instructions in the loop for loads or
stores with negative strides.

Differential Revision: https://reviews.llvm.org/D146128