[all-commits] [llvm/llvm-project] dcb822: [LoopVectorize] Propagate fast-math flags for inlo...

[RosieSumpter via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: dcb8222d87776a9e0f88344f156c1c95ee0c878f
      https://github.com/llvm/llvm-project/commit/dcb8222d87776a9e0f88344f156c1c95ee0c878f
  Author: Rosie Sumpter <rosie.sumpter at arm.com>
  Date:   2021-11-02 (Tue, 02 Nov 2021)

  Changed paths:
    M llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
    M llvm/test/Transforms/LoopVectorize/AArch64/strict-fadd.ll
    M llvm/test/Transforms/LoopVectorize/reduction-inloop-pred.ll
    M llvm/test/Transforms/LoopVectorize/reduction-inloop.ll

  Log Message:
  -----------
  [LoopVectorize] Propagate fast-math flags for inloop reductions

This patch updates VPReductionRecipe::execute so that the fast-math
flags associated with the underlying instruction of the VPRecipe are
propagated through to the reductions which are created.

Differential Revision: https://reviews.llvm.org/D112548