[llvm-branch-commits] [clang] [llvm] [AMDGPU] Introduce asyncmark/wait intrinsics (PR #180467)

[Pierre van Houtryve via llvm-branch-commits]

================
@@ -1034,6 +1030,10 @@ class SIInstrInfo final : public AMDGPUGenInstrInfo {
     return get(Opcode).TSFlags & SIInstrFlags::ASYNC_CNT;
   }
 
+  static bool usesLGKM_CNT(const MachineInstr &MI) {
----------------
Pierre-vh wrote:

function moves for no reason?

https://github.com/llvm/llvm-project/pull/180467