[llvm] [VectorCombine][AMDGPU] Narrow Phi of Shuffles. (PR #140188)

[Yingwei Zheng via llvm-commits]

================
@@ -3519,6 +3520,97 @@ bool VectorCombine::foldInterleaveIntrinsics(Instruction &I) {
   return true;
 }
 
+// Attempt to narrow a phi of shufflevector instructions where the two incoming
----------------
dtcxzyw wrote:

Use `///` for header comments.


https://github.com/llvm/llvm-project/pull/140188