[Mlir-commits] [mlir] [MLIR][AMDGPU] Redirect transfer read to masked load lowering (PR #146705)

[Zhuoran Yin]

================
@@ -51,8 +51,8 @@ def AmdgpuResolveStridedMetadataPass : Pass<"amdgpu-resolve-strided-metadata"> {
   ];
 }
 
-def AmdgpuTransferReadToLoadPass : Pass<"amdgpu-transfer-read-to-load"> {
-  let summary = "Lower the operations from the vector transfer_read to vector load";
+def AmdgpuMaskedloadToLoadPass : Pass<"amdgpu-maskedload-to-load"> {
+  let summary = "Lower the operations from the vector maskedload to vector load";
   let description = [{
     This pass creates a transfer read op lowering optimization. The lowering
----------------
jerryyin wrote:

Yep agreed!

https://github.com/llvm/llvm-project/pull/146705