[llvm] [AMDGPU][Attributor] Add `ThinOrFullLTOPhase` as an argument (PR #123994)

[Matt Arsenault via llvm-commits]

================
@@ -1365,9 +1366,30 @@ static bool runImpl(Module &M, AnalysisGetter &AG, TargetMachine &TM,
 
   Attributor A(Functions, InfoCache, AC);
 
-  LLVM_DEBUG(dbgs() << "[AMDGPUAttributor] Module " << M.getName() << " is "
-                    << (AC.IsClosedWorldModule ? "" : "not ")
-                    << "assumed to be a closed world.\n");
+  LLVM_DEBUG({
+    StringRef LTOPhaseStr;
+    switch (LTOPhase) {
+    case ThinOrFullLTOPhase::None:
+      LTOPhaseStr = "None";
+      break;
+    case ThinOrFullLTOPhase::ThinLTOPreLink:
+      LTOPhaseStr = "ThinLTOPreLink";
+      break;
+    case ThinOrFullLTOPhase::ThinLTOPostLink:
+      LTOPhaseStr = "ThinLTOPostLink";
+      break;
+    case ThinOrFullLTOPhase::FullLTOPreLink:
+      LTOPhaseStr = "FullLTOPreLink";
+      break;
+    case ThinOrFullLTOPhase::FullLTOPostLink:
+      LTOPhaseStr = "FullLTOPostLink";
+      break;
+    }
+    dbgs() << "[AMDGPUAttributor] Running at phase " << LTOPhaseStr << '\n';
+    dbgs() << "[AMDGPUAttributor] Module " << M.getName() << " is "
----------------
arsenm wrote:

Can also make this one continuous dbgs() << statement 

https://github.com/llvm/llvm-project/pull/123994