[Mlir-commits] [mlir] [mlir][AMX] Memory effects to amx.tile_zero (PR #155403)

[Arun Thangamani]

================
@@ -0,0 +1,30 @@
+// RUN: mlir-opt %s -cse -convert-vector-to-llvm="enable-amx" | mlir-opt | FileCheck %s
+
+// CHECK-LABEL: mem_effect(
+// CHECK: llvm.call_intrinsic "llvm.x86.tilezero.internal"
+// CHECK: llvm.call_intrinsic "llvm.x86.tilezero.internal"
+// CHECK: llvm.call_intrinsic "llvm.x86.tileloadd64.internal"
+// CHECK: llvm.call_intrinsic "llvm.x86.tileloadd64.internal"
+// CHECK: llvm.call_intrinsic "llvm.x86.tileloadd64.internal"
+// CHECK: llvm.call_intrinsic "llvm.x86.tileloadd64.internal"
+func.func @mem_effect(%arg0: memref<2x32x32xbf16>, %arg1: memref<2x16x32xbf16>) -> memref<16x32xf32> {
----------------
arun-thmn wrote:

Updated test-name + added a high level comment.

https://github.com/llvm/llvm-project/pull/155403