[clang] [llvm] [AMDGPU] First installment of IGLP_OPT 4 (MFMAValuSpacingOpt) implementation (PR #190916)

[via cfe-commits]

================
@@ -3244,6 +3244,21 @@ static bool hasIGLPInstrs(ScheduleDAGInstrs *DAG) {
   });
 }
 
+static bool hasIGLPOpt(ScheduleDAGInstrs *DAG, int64_t StrategyImm) {
+  return any_of(*DAG, [StrategyImm](MachineBasicBlock::iterator MI) {
+    return MI->getOpcode() == AMDGPU::IGLP_OPT && MI->getNumOperands() >= 1 &&
----------------
hidekisaito wrote:

I'll drop NumOperand and isImm checks in the next push. After opcode check, they seem redundant.

https://github.com/llvm/llvm-project/pull/190916