[llvm-branch-commits] [llvm] [AMDGPU] Use reverse iteration in CodeGenPrepare (PR #145484)

[Matt Arsenault via llvm-branch-commits]

================
@@ -109,6 +110,7 @@ class AMDGPUCodeGenPrepareImpl
   bool FlowChanged = false;
   mutable Function *SqrtF32 = nullptr;
   mutable Function *LdexpF32 = nullptr;
+  mutable SetVector<Value *> DeadVals;
----------------
arsenm wrote:

I don't think this needs to be a set, the iteration shouldn't revisit the same instruction twice (at least other IR combiner passes seem to all use vectors)

https://github.com/llvm/llvm-project/pull/145484