[llvm] [AMDGPU] Model WMMA co-execution windows in the scheduler for gfx1250 (PR #204077)

[Quentin Colombet via llvm-commits]

================
@@ -4314,12 +4310,10 @@ void PostGenericScheduler::initialize(ScheduleDAGMI *Dag) {
   // Initialize the HazardRecognizers. If itineraries don't exist, are empty,
   // or are disabled, then these HazardRecs will be disabled.
   const InstrItineraryData *Itin = SchedModel->getInstrItineraries();
-  if (!Top.HazardRec) {
-    Top.HazardRec = DAG->TII->CreateTargetMIHazardRecognizer(Itin, DAG);
-  }
-  if (!Bot.HazardRec) {
-    Bot.HazardRec = DAG->TII->CreateTargetMIHazardRecognizer(Itin, DAG);
-  }
+  if (!Top.HazardRec)
+    Top.HazardRec.reset(DAG->TII->CreateTargetMIHazardRecognizer(Itin, DAG));
+  if (!Bot.HazardRec)
+    Bot.HazardRec.reset(DAG->TII->CreateTargetMIHazardRecognizer(Itin, DAG));
----------------
qcolombet wrote:

You don't have too if it makes your life more complicated, but all the changes here could be committed as NFC.

https://github.com/llvm/llvm-project/pull/204077