[PATCH] D120023: [AMDGPU] Return better Changed status from SIFoldOperands

[Jay Foad via Phabricator via llvm-commits]

foad added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/SIFoldOperands.cpp:1240
       // be folded due to multiple uses or operand constraints.
       if (tryConstantFoldOp(*MRI, TII, &UseMI))
         LLVM_DEBUG(dbgs() << "Constant folded " << UseMI);
----------------
piotr wrote:
> Looks `tryConstantFoldOp` may make changes?
Wow, good catch! Should be fixed now. But I think we need some better testing for this kind of problem. I will see if D80916 can be extended to do hashing on MachineIR as well as IR.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D120023/new/

https://reviews.llvm.org/D120023