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

[Sameer Sahasrabuddhe via llvm-branch-commits]

================
@@ -1395,6 +1450,53 @@ void WaitcntBrackets::print(raw_ostream &OS) const {
   }
   OS << '\n';
 
+  OS << "Async score: ";
+  if (AsyncScore.empty())
+    OS << "none";
+  else
+    llvm::interleaveComma(AsyncScore, OS);
+  OS << '\n';
+
+  OS << "Async marks: " << AsyncMarks.size() << '\n';
+
+  for (const auto &Mark : AsyncMarks) {
+    for (auto T : inst_counter_types()) {
+      unsigned MarkedScore = Mark[T];
+      switch (T) {
+      case LOAD_CNT:
----------------
ssahasra wrote:

Yeah that always looks tempting on first sight. But the existing place and this new place are printing information at a different level of detail, with a different formatting of the lines. I am pretty sure the refactoring will not gain us much.

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