[all-commits] [llvm/llvm-project] 20b0fa: [SVE] Add support for folding for select + masked ...

[DylanFleming-arm via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 20b0fa91c9eebc7501e280049b61e8de352f3c94
      https://github.com/llvm/llvm-project/commit/20b0fa91c9eebc7501e280049b61e8de352f3c94
  Author: Dylan Fleming <Dylan.Fleming at arm.com>
  Date:   2021-07-26 (Mon, 26 Jul 2021)

  Changed paths:
    M llvm/include/llvm/IR/PatternMatch.h
    M llvm/lib/Transforms/InstCombine/InstCombineSelect.cpp
    A llvm/test/Transforms/InstCombine/select-masked_load.ll
    M llvm/test/Transforms/LoopVectorize/ARM/mve-reduction-predselect.ll
    M llvm/test/Transforms/LoopVectorize/ARM/mve-reductions.ll

  Log Message:
  -----------
  [SVE] Add support for folding for select + masked loads

Add folds to instcombine to support the removal of select instruction when the masked_load is guaranteed to zero the same lanes, i.e. select(mask, mload(,,mask,0), 0) -> mload(,,mask,0).

Patch originally authored by @paulwalker-arm

Reviewed By: david-arm

Differential Revision: https://reviews.llvm.org/D106376