[mlir][linalg] Update vectorization logic for linalg.pack (#149156) (#158926)

banach-space · web-flow · commit 2f6b433f2898 · 2025-09-18T10:32:46.000+01:00
NOTE: See #149156 for a smilar change for `linalg.unpack` This PR makes sure that we don't generate unnecessary `tensor.empty` when vectorizing `linalg.pack`. To better visualize the changes implemented here, consider this IR: ```mlir func.func @example( %src: tensor<64x4xf32>, %dest: tensor<2x4x16x2xf32>) -> tensor<2x4x16x2xf32> { %pack = linalg.pack %src outer_dims_perm = [1, 0] inner_dims_pos = [0, 1] inner_tiles = [16, 2] into %dest : tensor<64x4xf32> -> tensor<2x4x16x2xf32> return %pack : tensor<2x4x16x2xf32> } ``` Below is the output after vectorization, BEFORE and AFTER this PR. BEFORE (note `tensor.empty` and the fact that `%arg1` is not used): ```mlir func.func @example(%arg0: tensor<64x4xf32>, %arg1: tensor<2x4x16x2xf32>) -> tensor<2x4x16x2xf32> { %cst = arith.constant 0.000000e+00 : f32 %c0 = arith.constant 0 : index %0 = vector.transfer_read %arg0[%c0, %c0], %cst {in_bounds = [true, true]} : tensor<64x4xf32>, vector<64x4xf32> %1 = vector.shape_cast %0 : vector<64x4xf32> to vector<4x16x2x2xf32> %2 = vector.transpose %1, [2, 0, 1, 3] : vector<4x16x2x2xf32> to vector<2x4x16x2xf32> %3 = tensor.empty() : tensor<2x4x16x2xf32> %c0_0 = arith.constant 0 : index %4 = vector.transfer_write %2, %3[%c0_0, %c0_0, %c0_0, %c0_0] {in_bounds = [true, true, true, true]} : vector<2x4x16x2xf32>, tensor<2x4x16x2xf32> return %4 : tensor<2x4x16x2xf32> } ``` AFTER (note that `%arg1` is correctly used): ```mlir func.func @example(%arg0: tensor<64x4xf32>, %arg1: tensor<2x4x16x2xf32>) -> tensor<2x4x16x2xf32> { %cst = arith.constant 0.000000e+00 : f32 %c0 = arith.constant 0 : index %0 = vector.transfer_read %arg0[%c0, %c0], %cst {in_bounds = [true, true]} : tensor<64x4xf32>, vector<64x4xf32> %1 = vector.shape_cast %0 : vector<64x4xf32> to vector<4x16x2x2xf32> %2 = vector.transpose %1, [2, 0, 1, 3] : vector<4x16x2x2xf32> to vector<2x4x16x2xf32> %c0_0 = arith.constant 0 : index %3 = vector.transfer_write %2, %arg1[%c0_0, %c0_0, %c0_0, %c0_0] {in_bounds = [true, true, true, true]} : vector<2x4x16x2xf32>, tensor<2x4x16x2xf32> return %3 : tensor<2x4x16x2xf32> } ``` ADDITIONAL CHANGES: * Adds missing `CHECK-LABEL` in tests. * Capitalize LIT test variables names.
diff --git a/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp b/mlir/lib/Dialect/Linalg/Transforms/Vectorization.cpp
@@ -1776,12 +1776,6 @@ vectorizeAsTensorPackOp(RewriterBase &rewriter, linalg::PackOp packOp,
         rewriter, loc,
         rewriter.getZeroAttr(packOp.getSourceType().getElementType()));
   }
-  ReifiedRankedShapedTypeDims reifiedReturnShapes;
-  LogicalResult status =
-      cast<ReifyRankedShapedTypeOpInterface>(packOp.getOperation())
-          .reifyResultShapes(rewriter, reifiedReturnShapes);
-  (void)status; // prevent unused variable warning on non-assert builds.
-  assert(succeeded(status) && "failed to reify result shapes");
 
   // If the input vector sizes are not provided, then the vector sizes are
   // determined by the result tensor shape. In case the vector sizes aren't
@@ -1823,11 +1817,8 @@ vectorizeAsTensorPackOp(RewriterBase &rewriter, linalg::PackOp packOp,
       rewriter, loc, shapeCastOp.getResult(), destPermutation);
 
   // Create TransferWriteOp.
-  Value dest = tensor::EmptyOp::create(
-      rewriter, loc, reifiedReturnShapes[0],
-      transposeOp.getResult().getType().getElementType());
-  Operation *write =
-      createWriteOrMaskedWrite(rewriter, loc, transposeOp.getResult(), dest);
+  Operation *write = createWriteOrMaskedWrite(
+      rewriter, loc, transposeOp.getResult(), packOp.getDest());
   newResults.push_back(write->getResult(0));
   return success();
 }
diff --git a/mlir/test/Dialect/Linalg/vectorization/linalg-ops-with-patterns.mlir b/mlir/test/Dialect/Linalg/vectorization/linalg-ops-with-patterns.mlir
@@ -344,8 +344,7 @@ module attributes {transform.with_named_sequence} {
 // CHECK:           %[[VAL_4:.*]] = vector.transfer_read %[[VAL_0]]{{\[}}%[[VAL_3]], %[[VAL_3]], %[[VAL_3]]], %[[VAL_2]] {in_bounds = [true, true, true]} : tensor<32x8x16xf32>, vector<32x8x16xf32>
 // CHECK:           %[[VAL_5:.*]] = vector.shape_cast %[[VAL_4]] : vector<32x8x16xf32> to vector<32x4x2x1x16xf32>
 // CHECK:           %[[VAL_6:.*]] = vector.transpose %[[VAL_5]], [1, 3, 0, 4, 2] : vector<32x4x2x1x16xf32> to vector<4x1x32x16x2xf32>
-// CHECK:           %[[VAL_7:.*]] = tensor.empty() : tensor<4x1x32x16x2xf32>
-// CHECK:           %[[VAL_8:.*]] = vector.transfer_write %[[VAL_6]], %[[VAL_7]]{{\[}}%[[VAL_3]], %[[VAL_3]], %[[VAL_3]], %[[VAL_3]], %[[VAL_3]]] {in_bounds = [true, true, true, true, true]} : vector<4x1x32x16x2xf32>, tensor<4x1x32x16x2xf32>
+// CHECK:           %[[VAL_8:.*]] = vector.transfer_write %[[VAL_6]], %[[VAL_1]]{{\[}}%[[VAL_3]], %[[VAL_3]], %[[VAL_3]], %[[VAL_3]], %[[VAL_3]]] {in_bounds = [true, true, true, true, true]} : vector<4x1x32x16x2xf32>, tensor<4x1x32x16x2xf32>
 // CHECK:           return %[[VAL_8]] : tensor<4x1x32x16x2xf32>
 
 // -----
@@ -364,8 +363,7 @@ func.func @test_vectorize_padded_pack(%arg0: tensor<32x7x15xf32>, %arg1: tensor<
 // CHECK:           %[[VAL_4:.*]] = vector.transfer_read %[[VAL_0]]{{\[}}%[[VAL_3]], %[[VAL_3]], %[[VAL_3]]], %[[VAL_2]] {in_bounds = [true, false, false]} : tensor<32x7x15xf32>, vector<32x8x16xf32>
 // CHECK:           %[[VAL_5:.*]] = vector.shape_cast %[[VAL_4]] : vector<32x8x16xf32> to vector<32x4x2x1x16xf32>
 // CHECK:           %[[VAL_6:.*]] = vector.transpose %[[VAL_5]], [0, 1, 3, 4, 2] : vector<32x4x2x1x16xf32> to vector<32x4x1x16x2xf32>
-// CHECK:           %[[VAL_7:.*]] = tensor.empty() : tensor<32x4x1x16x2xf32>
-// CHECK:           %[[VAL_8:.*]] = vector.transfer_write %[[VAL_6]], %[[VAL_7]]{{\[}}%[[VAL_3]], %[[VAL_3]], %[[VAL_3]], %[[VAL_3]], %[[VAL_3]]] {in_bounds = [true, true, true, true, true]} : vector<32x4x1x16x2xf32>, tensor<32x4x1x16x2xf32>
+// CHECK:           %[[VAL_8:.*]] = vector.transfer_write %[[VAL_6]], %[[VAL_1]]{{\[}}%[[VAL_3]], %[[VAL_3]], %[[VAL_3]], %[[VAL_3]], %[[VAL_3]]] {in_bounds = [true, true, true, true, true]} : vector<32x4x1x16x2xf32>, tensor<32x4x1x16x2xf32>
 // CHECK:           return %[[VAL_8]] : tensor<32x4x1x16x2xf32>
 
 module attributes {transform.with_named_sequence} {
diff --git a/mlir/test/Dialect/Linalg/vectorization/linalg-ops.mlir b/mlir/test/Dialect/Linalg/vectorization/linalg-ops.mlir
@@ -1301,25 +1301,27 @@ func.func @test_vectorize_unpack_no_vector_sizes_permute(%source: tensor<4x7x4xf
 // different - vector sizes are inferred (rather than user-specified) and hence
 // masking was used.
 
-func.func @test_vectorize_pack(%arg0: tensor<32x8x16xf32>, %arg1: tensor<4x1x32x16x2xf32>) -> tensor<4x1x32x16x2xf32> {
-  %pack = linalg.pack %arg0 outer_dims_perm = [1, 2, 0] inner_dims_pos = [2, 1] inner_tiles = [16, 2] into %arg1 : tensor<32x8x16xf32> -> tensor<4x1x32x16x2xf32>
+// CHECK-LABEL: func @test_vectorize_pack
+// CHECK-SAME:      %[[SRC:.*]]: tensor<32x8x16xf32>,
+// CHECK-SAME:      %[[DEST:.*]]: tensor<4x1x32x16x2xf32>
+func.func @test_vectorize_pack(%src: tensor<32x8x16xf32>, %dest: tensor<4x1x32x16x2xf32>) -> tensor<4x1x32x16x2xf32> {
+  %pack = linalg.pack %src outer_dims_perm = [1, 2, 0] inner_dims_pos = [2, 1] inner_tiles = [16, 2] into %dest : tensor<32x8x16xf32> -> tensor<4x1x32x16x2xf32>
   return %pack : tensor<4x1x32x16x2xf32>
 }
-//  CHECK-DAG: %[[cst:.*]] = arith.constant 0.000000e+00 : f32
-//  CHECK-DAG: %[[c0:.*]] = arith.constant 0 : index
-//      CHECK: %[[read:.*]] = vector.transfer_read %{{.*}}[%[[c0]], %[[c0]], %[[c0]]], %[[cst]]
+//  CHECK-DAG: %[[CST:.*]] = arith.constant 0.000000e+00 : f32
+//  CHECK-DAG: %[[C0:.*]] = arith.constant 0 : index
+//      CHECK: %[[READ:.*]] = vector.transfer_read %{{.*}}[%[[C0]], %[[C0]], %[[C0]]], %[[CST]]
 // CHECK-SAME:    {in_bounds = [true, true, true]} : tensor<32x8x16xf32>, vector<32x8x16xf32>
-//      CHECK: %[[shape_cast:.*]] = vector.shape_cast %[[read]] : vector<32x8x16xf32> to vector<32x4x2x1x16xf32>
-//      CHECK: %[[transpose:.*]] = vector.transpose %[[shape_cast]], [1, 3, 0, 4, 2] : vector<32x4x2x1x16xf32> to vector<4x1x32x16x2xf32>
-//  CHECK-DAG: %[[c0_1:.*]] = arith.constant 0 : index
-//  CHECK-DAG: %[[empty:.*]] = tensor.empty() : tensor<4x1x32x16x2xf32>
-//      CHECK: %[[write:.*]] = vector.transfer_write %[[transpose]], %[[empty]][%[[c0_1]], %[[c0_1]], %[[c0_1]], %[[c0_1]], %[[c0_1]]]
+//      CHECK: %[[SC:.*]] = vector.shape_cast %[[READ]] : vector<32x8x16xf32> to vector<32x4x2x1x16xf32>
+//      CHECK: %[[TR:.*]] = vector.transpose %[[SC]], [1, 3, 0, 4, 2] : vector<32x4x2x1x16xf32> to vector<4x1x32x16x2xf32>
+//  CHECK-DAG: %[[C0_1:.*]] = arith.constant 0 : index
+//      CHECK: %[[write:.*]] = vector.transfer_write %[[TR]], %[[DEST]][%[[C0_1]], %[[C0_1]], %[[C0_1]], %[[C0_1]], %[[C0_1]]]
 // CHECK-SAME:   {in_bounds = [true, true, true, true, true]} : vector<4x1x32x16x2xf32>, tensor<4x1x32x16x2xf32>
 //      CHECK: return %[[write]] : tensor<4x1x32x16x2xf32>
 
 module attributes {transform.with_named_sequence} {
-  transform.named_sequence @__transform_main(%arg0: !transform.any_op {transform.readonly}) {
-    %0 = transform.structured.match ops{["linalg.pack"]} in %arg0 : (!transform.any_op) -> !transform.any_op
+  transform.named_sequence @__transform_main(%src: !transform.any_op {transform.readonly}) {
+    %0 = transform.structured.match ops{["linalg.pack"]} in %src : (!transform.any_op) -> !transform.any_op
     transform.structured.vectorize %0 vector_sizes [4, 1, 32] : !transform.any_op
     transform.yield
   }
@@ -1331,26 +1333,28 @@ module attributes {transform.with_named_sequence} {
 // different - vector sizes are inferred (rather than user-specified) and hence
 // masking was used.
 
-func.func @test_vectorize_padded_pack(%arg0: tensor<32x7x15xf32>, %arg1: tensor<32x4x1x16x2xf32>) -> tensor<32x4x1x16x2xf32> {
+// CHECK-LABEL: func @test_vectorize_padded_pack
+// CHECK-SAME:      %[[SRC:.*]]: tensor<32x7x15xf32>,
+// CHECK-SAME:      %[[DEST:.*]]: tensor<32x4x1x16x2xf32>
+func.func @test_vectorize_padded_pack(%src: tensor<32x7x15xf32>, %dest: tensor<32x4x1x16x2xf32>) -> tensor<32x4x1x16x2xf32> {
   %pad = arith.constant 0.000000e+00 : f32
-  %pack = linalg.pack %arg0 padding_value(%pad : f32) inner_dims_pos = [2, 1] inner_tiles = [16, 2] into %arg1 : tensor<32x7x15xf32> -> tensor<32x4x1x16x2xf32>
+  %pack = linalg.pack %src padding_value(%pad : f32) inner_dims_pos = [2, 1] inner_tiles = [16, 2] into %dest : tensor<32x7x15xf32> -> tensor<32x4x1x16x2xf32>
   return %pack : tensor<32x4x1x16x2xf32>
 }
-//  CHECK-DAG: %[[cst:.*]] = arith.constant 0.000000e+00 : f32
-//  CHECK-DAG: %[[c0:.*]] = arith.constant 0 : index
-//  CHECK-DAG: %[[c32:.*]] = arith.constant 32 : index
-//  CHECK-DAG: %[[c7:.*]] = arith.constant 7 : index
-//  CHECK-DAG: %[[c15:.*]] = arith.constant 15 : index
-//      CHECK: %[[mask:.*]] = vector.create_mask %[[c32]], %[[c7]], %[[c15]] : vector<32x8x16xi1>
-//      CHECK: %[[masked_read:.*]] = vector.mask %[[mask]] {
-// CHECK-SAME:   vector.transfer_read %{{.*}}[%[[c0]], %[[c0]], %[[c0]]], %[[cst]]
+//  CHECK-DAG: %[[CST:.*]] = arith.constant 0.000000e+00 : f32
+//  CHECK-DAG: %[[C0:.*]] = arith.constant 0 : index
+//  CHECK-DAG: %[[C32:.*]] = arith.constant 32 : index
+//  CHECK-DAG: %[[C7:.*]] = arith.constant 7 : index
+//  CHECK-DAG: %[[C15:.*]] = arith.constant 15 : index
+//      CHECK: %[[MASK:.*]] = vector.create_mask %[[C32]], %[[C7]], %[[C15]] : vector<32x8x16xi1>
+//      CHECK: %[[READ:.*]] = vector.mask %[[MASK]] {
+// CHECK-SAME:   vector.transfer_read %{{.*}}[%[[C0]], %[[C0]], %[[C0]]], %[[CST]]
 // CHECK-SAME:   {in_bounds = [true, true, true]} : tensor<32x7x15xf32>, vector<32x8x16xf32>
 // CHECK-SAME: } : vector<32x8x16xi1> -> vector<32x8x16xf32>
-//      CHECK: %[[shape_cast:.*]] = vector.shape_cast %[[masked_read]] : vector<32x8x16xf32> to vector<32x4x2x1x16xf32>
-//      CHECK: %[[transpose:.*]] = vector.transpose %[[shape_cast]], [0, 1, 3, 4, 2] : vector<32x4x2x1x16xf32> to vector<32x4x1x16x2xf32>
-//  CHECK-DAG: %[[c0_1:.*]] = arith.constant 0 : index
-//  CHECK-DAG: %[[empty:.*]] = tensor.empty() : tensor<32x4x1x16x2xf32>
-//      CHECK: %[[write:.*]] = vector.transfer_write %[[transpose]], %[[empty]][%[[c0_1]], %[[c0_1]], %[[c0_1]], %[[c0_1]], %[[c0_1]]]
+//      CHECK: %[[SC:.*]] = vector.shape_cast %[[READ]] : vector<32x8x16xf32> to vector<32x4x2x1x16xf32>
+//      CHECK: %[[TR:.*]] = vector.transpose %[[SC]], [0, 1, 3, 4, 2] : vector<32x4x2x1x16xf32> to vector<32x4x1x16x2xf32>
+//  CHECK-DAG: %[[C0_1:.*]] = arith.constant 0 : index
+//      CHECK: %[[write:.*]] = vector.transfer_write %[[TR]], %[[DEST]][%[[C0_1]], %[[C0_1]], %[[C0_1]], %[[C0_1]], %[[C0_1]]]
 // CHECK-SAME:   {in_bounds = [true, true, true, true, true]} : vector<32x4x1x16x2xf32>, tensor<32x4x1x16x2xf32>
 //      CHECK: return %[[write]] : tensor<32x4x1x16x2xf32>
 
@@ -1364,38 +1368,37 @@ module attributes {transform.with_named_sequence} {
 
 // -----
 
-func.func @test_vectorize_dynamic_pack(%arg0: tensor<?x?xf32>, %arg1: tensor<?x?x16x2xf32>) -> tensor<?x?x16x2xf32> {
-  %pack = linalg.pack %arg0 inner_dims_pos = [1, 0] inner_tiles = [16, 2] into %arg1 : tensor<?x?xf32> -> tensor<?x?x16x2xf32>
+// CHECK-LABEL: func @test_vectorize_dynamic_pack
+// CHECK-SAME:      %[[SRC:.*]]: tensor<?x?xf32>,
+// CHECK-SAME:      %[[DEST:.*]]: tensor<?x?x16x2xf32>
+func.func @test_vectorize_dynamic_pack(%src: tensor<?x?xf32>, %dest: tensor<?x?x16x2xf32>) -> tensor<?x?x16x2xf32> {
+  %pack = linalg.pack %src inner_dims_pos = [1, 0] inner_tiles = [16, 2] into %dest : tensor<?x?xf32> -> tensor<?x?x16x2xf32>
   return %pack : tensor<?x?x16x2xf32>
 }
-//  CHECK-DAG: %[[cst:.*]] = arith.constant 0.000000e+00 : f32
-//  CHECK-DAG: %[[c0:.*]] = arith.constant 0 : index
-//  CHECK-DAG: %[[c1:.*]] = arith.constant 1 : index
-//  CHECK-DAG: %[[d0:.*]] = tensor.dim {{.*}} %[[c0]] : tensor<?x?x16x2xf32>
-//  CHECK-DAG: %[[d1:.*]] = tensor.dim {{.*}} %[[c1]] : tensor<?x?x16x2xf32>
-//  CHECK-DAG: %[[c0_1:.*]] = arith.constant 0 : index
-//  CHECK-DAG: %[[c0_0:.*]] = arith.constant 0 : index
-//  CHECK-DAG: %[[c1_0:.*]] = arith.constant 1 : index
-//  CHECK-DAG: %[[d0_0:.*]] = tensor.dim {{.*}} %[[c0_0]] : tensor<?x?xf32>
-//  CHECK-DAG: %[[d1_0:.*]] = tensor.dim {{.*}} %[[c1_0]] : tensor<?x?xf32>
-//      CHECK: %[[mask:.*]] = vector.create_mask %[[d0_0]], %[[d1_0]] : vector<8x16xi1>
-//      CHECK: %[[masked_read:.*]] = vector.mask %[[mask]] {
-// CHECK-SAME:   vector.transfer_read %{{.*}}[%[[c0_1]], %[[c0_1]]], %[[cst]]
+
+//  CHECK-DAG: %[[CST:.*]] = arith.constant 0.000000e+00 : f32
+//  CHECK-DAG: %[[C0_1:.*]] = arith.constant 0 : index
+//  CHECK-DAG: %[[C0_0:.*]] = arith.constant 0 : index
+//  CHECK-DAG: %[[C1_0:.*]] = arith.constant 1 : index
+//  CHECK-DAG: %[[D0_0:.*]] = tensor.dim {{.*}} %[[C0_0]] : tensor<?x?xf32>
+//  CHECK-DAG: %[[D1_0:.*]] = tensor.dim {{.*}} %[[C1_0]] : tensor<?x?xf32>
+//      CHECK: %[[MASK:.*]] = vector.create_mask %[[D0_0]], %[[D1_0]] : vector<8x16xi1>
+//      CHECK: %[[READ:.*]] = vector.mask %[[MASK]] {
+// CHECK-SAME:   vector.transfer_read %{{.*}}[%[[C0_1]], %[[C0_1]]], %[[CST]]
 // CHECK-SAME:   {in_bounds = [true, true]} : tensor<?x?xf32>, vector<8x16xf32>
 // CHECK-SAME: } : vector<8x16xi1> -> vector<8x16xf32>
-//      CHECK: %[[shape_cast:.*]] = vector.shape_cast %[[masked_read]] : vector<8x16xf32> to vector<4x2x1x16xf32>
-//      CHECK: %[[transpose:.*]] = vector.transpose %[[shape_cast]], [0, 2, 3, 1] : vector<4x2x1x16xf32> to vector<4x1x16x2xf32>
-//  CHECK-DAG: %[[c0_2:.*]] = arith.constant 0 : index
-//  CHECK-DAG: %[[c16:.*]] = arith.constant 16 : index
-//  CHECK-DAG: %[[c2:.*]] = arith.constant 2 : index
-//  CHECK-DAG: %[[empty:.*]] = tensor.empty(%[[d0]], %[[d1]]) : tensor<?x?x16x2xf32>
-//  CHECK-DAG: %[[d2:.*]] = tensor.dim %[[empty]], {{.*}} : tensor<?x?x16x2xf32>
-//  CHECK-DAG: %[[d3:.*]] = tensor.dim %[[empty]], {{.*}} : tensor<?x?x16x2xf32>
-//      CHECK: %[[mask_0:.*]] = vector.create_mask %[[d2]], %[[d3]], %[[c16]], %[[c2]] : vector<4x1x16x2xi1>
-//      CHECK: %[[masked_write:.*]] = vector.mask %[[mask_0]] {
-// CHECK-SAME:   vector.transfer_write %[[transpose]], %[[empty]][%[[c0_2]], %[[c0_2]], %[[c0_2]], %[[c0_2]]]
+//      CHECK: %[[SC:.*]] = vector.shape_cast %[[READ]] : vector<8x16xf32> to vector<4x2x1x16xf32>
+//      CHECK: %[[TR:.*]] = vector.transpose %[[SC]], [0, 2, 3, 1] : vector<4x2x1x16xf32> to vector<4x1x16x2xf32>
+//  CHECK-DAG: %[[C0_2:.*]] = arith.constant 0 : index
+//  CHECK-DAG: %[[C16:.*]] = arith.constant 16 : index
+//  CHECK-DAG: %[[C2:.*]] = arith.constant 2 : index
+//  CHECK-DAG: %[[D2:.*]] = tensor.dim %[[DEST]], {{.*}} : tensor<?x?x16x2xf32>
+//  CHECK-DAG: %[[D3:.*]] = tensor.dim %[[DEST]], {{.*}} : tensor<?x?x16x2xf32>
+//      CHECK: %[[MASK_0:.*]] = vector.create_mask %[[D2]], %[[D3]], %[[C16]], %[[C2]] : vector<4x1x16x2xi1>
+//      CHECK: %[[WRITE:.*]] = vector.mask %[[MASK_0]] {
+// CHECK-SAME:   vector.transfer_write %[[TR]], %[[DEST]][%[[C0_2]], %[[C0_2]], %[[C0_2]], %[[C0_2]]]
 // CHECK-SAME:   {in_bounds = [true, true, true, true]} : vector<4x1x16x2xf32>, tensor<?x?x16x2xf32>
-//      CHECK: return %[[masked_write]] : tensor<?x?x16x2xf32>
+//      CHECK: return %[[WRITE]] : tensor<?x?x16x2xf32>
 
 module attributes {transform.with_named_sequence} {
   transform.named_sequence @__transform_main(%arg0: !transform.any_op {transform.readonly}) {
@@ -1407,22 +1410,23 @@ module attributes {transform.with_named_sequence} {
 
 // -----
 
-// CHECK-LABEL: test_vectorize_pack_no_vector_sizes
-func.func @test_vectorize_pack_no_vector_sizes(%arg0: tensor<64x4xf32>, %arg1: tensor<2x4x16x2xf32>) -> tensor<2x4x16x2xf32> {
-  %pack = linalg.pack %arg0 outer_dims_perm = [1, 0] inner_dims_pos = [0, 1] inner_tiles = [16, 2] into %arg1 : tensor<64x4xf32> -> tensor<2x4x16x2xf32>
+// CHECK-LABEL: func @test_vectorize_pack_no_vector_sizes
+// CHECK-SAME:      %[[SRC:.*]]: tensor<64x4xf32>,
+// CHECK-SAME:      %[[DEST:.*]]: tensor<2x4x16x2xf32>
+func.func @test_vectorize_pack_no_vector_sizes(%src: tensor<64x4xf32>, %dest: tensor<2x4x16x2xf32>) -> tensor<2x4x16x2xf32> {
+  %pack = linalg.pack %src outer_dims_perm = [1, 0] inner_dims_pos = [0, 1] inner_tiles = [16, 2] into %dest : tensor<64x4xf32> -> tensor<2x4x16x2xf32>
   return %pack : tensor<2x4x16x2xf32>
 }
-//  CHECK-DAG: %[[cst:.*]] = arith.constant 0.000000e+00 : f32
-//  CHECK-DAG: %[[c0:.*]] = arith.constant 0 : index
-//      CHECK: %[[read:.*]] = vector.transfer_read %{{.*}}[%[[c0]], %[[c0]]], %[[cst]]
+//  CHECK-DAG: %[[CST:.*]] = arith.constant 0.000000e+00 : f32
+//  CHECK-DAG: %[[C0:.*]] = arith.constant 0 : index
+//      CHECK: %[[READ:.*]] = vector.transfer_read %{{.*}}[%[[C0]], %[[C0]]], %[[CST]]
 // CHECK-SAME:    {in_bounds = [true, true]} : tensor<64x4xf32>, vector<64x4xf32>
-//      CHECK: %[[shape_cast:.*]] = vector.shape_cast %[[read]] : vector<64x4xf32> to vector<4x16x2x2xf32>
-//      CHECK: %[[transpose:.*]] = vector.transpose %[[shape_cast]], [2, 0, 1, 3] : vector<4x16x2x2xf32> to vector<2x4x16x2xf32>
-//  CHECK-DAG: %[[c0_1:.*]] = arith.constant 0 : index
-//  CHECK-DAG: %[[empty:.*]] = tensor.empty() : tensor<2x4x16x2xf32>
-//      CHECK: %[[write:.*]] = vector.transfer_write %[[transpose]], %[[empty]][%[[c0_1]], %[[c0_1]], %[[c0_1]], %[[c0_1]]]
+//      CHECK: %[[SC:.*]] = vector.shape_cast %[[READ]] : vector<64x4xf32> to vector<4x16x2x2xf32>
+//      CHECK: %[[TR:.*]] = vector.transpose %[[SC]], [2, 0, 1, 3] : vector<4x16x2x2xf32> to vector<2x4x16x2xf32>
+//  CHECK-DAG: %[[C0_1:.*]] = arith.constant 0 : index
+//      CHECK: %[[WRITE:.*]] = vector.transfer_write %[[TR]], %[[DEST]][%[[C0_1]], %[[C0_1]], %[[C0_1]], %[[C0_1]]]
 // CHECK-SAME:   {in_bounds = [true, true, true, true]} : vector<2x4x16x2xf32>, tensor<2x4x16x2xf32>
-//      CHECK: return %[[write]] : tensor<2x4x16x2xf32>
+//      CHECK: return %[[WRITE]] : tensor<2x4x16x2xf32>
 
 module attributes {transform.with_named_sequence} {
   transform.named_sequence @__transform_main(%arg0: !transform.any_op {transform.readonly}) {
@@ -1435,22 +1439,23 @@ module attributes {transform.with_named_sequence} {
 // -----
 
 // CHECK-LABEL: test_vectorize_padded_pack_no_vector_sizes
-func.func @test_vectorize_padded_pack_no_vector_sizes(%arg0: tensor<32x7x15xf32>, %arg1: tensor<32x4x1x16x2xf32>) -> tensor<32x4x1x16x2xf32> {
+// CHECK-SAME:      %[[SRC:.*]]: tensor<32x7x15xf32>,
+// CHECK-SAME:      %[[DEST:.*]]: tensor<32x4x1x16x2xf32>
+func.func @test_vectorize_padded_pack_no_vector_sizes(%src: tensor<32x7x15xf32>, %dest: tensor<32x4x1x16x2xf32>) -> tensor<32x4x1x16x2xf32> {
   %pad = arith.constant 0.000000e+00 : f32
-  %pack = linalg.pack %arg0 padding_value(%pad : f32) inner_dims_pos = [2, 1] inner_tiles = [16, 2] into %arg1 : tensor<32x7x15xf32> -> tensor<32x4x1x16x2xf32>
+  %pack = linalg.pack %src padding_value(%pad : f32) inner_dims_pos = [2, 1] inner_tiles = [16, 2] into %dest : tensor<32x7x15xf32> -> tensor<32x4x1x16x2xf32>
   return %pack : tensor<32x4x1x16x2xf32>
 }
-//  CHECK-DAG: %[[cst:.*]] = arith.constant 0.000000e+00 : f32
-//  CHECK-DAG: %[[c0:.*]] = arith.constant 0 : index
-//      CHECK: %[[transfer_read:.*]] =  vector.transfer_read %{{.*}}[%[[c0]], %[[c0]], %[[c0]]], %[[cst]]
+//  CHECK-DAG: %[[CST:.*]] = arith.constant 0.000000e+00 : f32
+//  CHECK-DAG: %[[C0:.*]] = arith.constant 0 : index
+//      CHECK: %[[READ:.*]] =  vector.transfer_read %{{.*}}[%[[C0]], %[[C0]], %[[C0]]], %[[CST]]
 // CHECK-SAME:   {in_bounds = [true, false, false]} : tensor<32x7x15xf32>, vector<32x8x16xf32>
-//      CHECK: %[[shape_cast:.*]] = vector.shape_cast %[[transfer_read]] : vector<32x8x16xf32> to vector<32x4x2x1x16xf32>
-//      CHECK: %[[transpose:.*]] = vector.transpose %[[shape_cast]], [0, 1, 3, 4, 2] : vector<32x4x2x1x16xf32> to vector<32x4x1x16x2xf32>
-//  CHECK-DAG: %[[c0_1:.*]] = arith.constant 0 : index
-//  CHECK-DAG: %[[empty:.*]] = tensor.empty() : tensor<32x4x1x16x2xf32>
-//      CHECK: %[[write:.*]] = vector.transfer_write %[[transpose]], %[[empty]][%[[c0_1]], %[[c0_1]], %[[c0_1]], %[[c0_1]], %[[c0_1]]]
+//      CHECK: %[[SC:.*]] = vector.shape_cast %[[READ]] : vector<32x8x16xf32> to vector<32x4x2x1x16xf32>
+//      CHECK: %[[TR:.*]] = vector.transpose %[[SC]], [0, 1, 3, 4, 2] : vector<32x4x2x1x16xf32> to vector<32x4x1x16x2xf32>
+//  CHECK-DAG: %[[C0_1:.*]] = arith.constant 0 : index
+//      CHECK: %[[WRITE:.*]] = vector.transfer_write %[[TR]], %[[DEST]][%[[C0_1]], %[[C0_1]], %[[C0_1]], %[[C0_1]], %[[C0_1]]]
 // CHECK-SAME:   {in_bounds = [true, true, true, true, true]} : vector<32x4x1x16x2xf32>, tensor<32x4x1x16x2xf32>
-//      CHECK: return %[[write]] : tensor<32x4x1x16x2xf32>
+//      CHECK: return %[[WRITE]] : tensor<32x4x1x16x2xf32>
 
 module attributes {transform.with_named_sequence} {
   transform.named_sequence @__transform_main(%arg0: !transform.any_op {transform.readonly}) {
