[Mlir-commits] [mlir] [mlir][vector] Teach `TransferOptimization` to forward masked stores (PR #87794)

[Cullen Rhodes]

================
@@ -360,3 +360,53 @@ func.func @forward_dead_store_dynamic_non_overlap_trailing_dim(
   vector.transfer_write %x, %buffer[%i0, %i0] {in_bounds = [true]} : vector<4xf32>, memref<?x?xf32>
   return
 }
+
+// CHECK-LABEL: func @forward_dead_constant_splat_store_with_masking
+//   CHECK-NOT:   vector.transfer_write
+//   CHECK-NOT:   vector.transfer_read
+//       CHECK:   scf.for
+//       CHECK:   }
+//       CHECK:   vector.transfer_write
+//       CHECK:   return
+func.func @forward_dead_constant_splat_store_with_masking(%buffer : memref<?x?xf32>, %mask: vector<[8]x[8]xi1>) {
+  %cst = arith.constant dense<0.0> : vector<[8]x[8]xf32>
+  %cst_f32 = arith.constant 0.0 : f32
+  %c1 = arith.constant 1 : index
+  %c0 = arith.constant 0 : index
+  %c512 = arith.constant 512 : index
+  %vscale = vector.vscale
----------------
c-rhodes wrote:

unused

https://github.com/llvm/llvm-project/pull/87794