[all-commits] [llvm/llvm-project] 9e97b2: [DAG] SimplifySetCC - relax fold (X^C1) == C2 --> ...

[Simon Pilgrim via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 9e97b2a477f3b9bdd1ef7954ce22e15d15fb54d8
      https://github.com/llvm/llvm-project/commit/9e97b2a477f3b9bdd1ef7954ce22e15d15fb54d8
  Author: Simon Pilgrim <llvm-dev at redking.me.uk>
  Date:   2022-04-06 (Wed, 06 Apr 2022)

  Changed paths:
    M llvm/lib/CodeGen/SelectionDAG/TargetLowering.cpp
    M llvm/test/CodeGen/Mips/countleading.ll
    M llvm/test/CodeGen/Thumb2/thumb2-teq.ll
    M llvm/test/CodeGen/X86/peep-test-3.ll

  Log Message:
  -----------
  [DAG] SimplifySetCC - relax fold (X^C1) == C2 --> X == C1^C2

https://alive2.llvm.org/ce/z/A_auBq

Remove limitation that wouldn't perform the fold if all the inverted bits are known zero

The thumb2 changes look to be benign, although it does show that the TEQ/TST isel patterns could probably be improved.

Fixes movmsk regression in D122754

Differential Revision: https://reviews.llvm.org/D123023