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

[via cfe-commits]

================
@@ -3082,11 +3082,28 @@ void GCNScheduleDAGMILive::setTargetOccupancy(unsigned TargetOccupancy) {
     MFI.limitOccupancy(MinOccupancy);
 }
 
-static bool hasIGLPInstrs(ScheduleDAGInstrs *DAG) {
+/// Scan for IGLP "mutation-only" instructions (SCHED_GROUP_BARRIER or
+/// IGLP_OPT). Returns {any in region, strategy ID if such an instruction is
+/// IGLP_OPT}.
+static std::pair<bool, std::optional<AMDGPU::IGLPStrategyID>>
----------------
hidekisaito wrote:

Fixed. Two functions merged and for-loop over MIs exits as soon as it can.

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