llvm-project

[AMDGPU] Use generic operand swapping in SIInstrInfo::commuteInstructionImpl

#238393

Open