[mlir][tosa] Fix transpose_conv2d verifier when output channels are dynamic #147062
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…ynamic This commit fixes a transpose_conv2d verifier check which compares the output channels size to the bias size. The check didn't make sure output channels were static before performing the comparison. This lead to failures such as: ``` 'tosa.transpose_conv2d' op bias channels expected to be equal to output channels (-9223372036854775808) or 1, got 5 ``` when the output channels size was dynamic. Change-Id: If10625e46be3d3ac5577aa8913621a41cda7b8f3
|
@llvm/pr-subscribers-mlir-tosa Author: Luke Hutton (lhutton1) ChangesThis commit fixes a transpose_conv2d verifier check which compares the output channels size to the bias size. The check didn't make sure output channels were static before performing the comparison. This lead to failures such as: when the output channels size was dynamic. Full diff: https://github.com/llvm/llvm-project/pull/147062.diff 2 Files Affected:
diff --git a/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp b/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp
index a0dc8964f7a02..85ab48b37813f 100644
--- a/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp
+++ b/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp
@@ -3429,7 +3429,8 @@ LogicalResult TransposeConv2DOp::verify() {
return success();
const int64_t outputChannels = outputType.getDimSize(3);
- if (biasChannels != outputChannels && biasChannels != 1)
+ if (!ShapedType::isDynamic(outputChannels) &&
+ biasChannels != outputChannels && biasChannels != 1)
return emitOpError(
"bias channels expected to be equal to output channels (")
<< outputChannels << ") or 1, got " << biasChannels;
diff --git a/mlir/test/Dialect/Tosa/tosa-infer-shapes.mlir b/mlir/test/Dialect/Tosa/tosa-infer-shapes.mlir
index e280c1155f526..d0f40279421f4 100644
--- a/mlir/test/Dialect/Tosa/tosa-infer-shapes.mlir
+++ b/mlir/test/Dialect/Tosa/tosa-infer-shapes.mlir
@@ -1032,6 +1032,15 @@ func.func @transpose_conv2d_strided(%arg0: tensor<1x5x7x1xf32>, %arg1: tensor<1x
// -----
+// CHECK-LABEL: @transpose_conv2d_dynamic_out_channels
+func.func @transpose_conv2d_dynamic_out_channels(%arg0: tensor<2x1x1x3xf32>, %arg1: tensor<5x3x6x3xf32>, %arg2: tensor<5xf32>, %arg3: tensor<1xf32>, %arg4: tensor<1xf32>) {
+ // CHECK: -> tensor<2x3x6x5xf32>
+ %0 = tosa.transpose_conv2d %arg0, %arg1, %arg2, %arg3, %arg4 {acc_type = f32, out_pad = array<i64: 0, 0, 0, 0>, stride = array<i64: 1, 1>} : (tensor<2x1x1x3xf32>, tensor<5x3x6x3xf32>, tensor<5xf32>, tensor<1xf32>, tensor<1xf32>) -> tensor<2x3x6x?xf32>
+ return
+}
+
+// -----
+
// CHECK-LABEL: @resize_int_horizontal
func.func @resize_int_horizontal(%arg0: tensor<1x15x13x1xi8>) {
%scale = tosa.const_shape { values = dense<[11, 7, 89, 6]> : tensor<4xindex> } : () -> !tosa.shape<4>
|
|
@llvm/pr-subscribers-mlir Author: Luke Hutton (lhutton1) ChangesThis commit fixes a transpose_conv2d verifier check which compares the output channels size to the bias size. The check didn't make sure output channels were static before performing the comparison. This lead to failures such as: when the output channels size was dynamic. Full diff: https://github.com/llvm/llvm-project/pull/147062.diff 2 Files Affected:
diff --git a/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp b/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp
index a0dc8964f7a02..85ab48b37813f 100644
--- a/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp
+++ b/mlir/lib/Dialect/Tosa/IR/TosaOps.cpp
@@ -3429,7 +3429,8 @@ LogicalResult TransposeConv2DOp::verify() {
return success();
const int64_t outputChannels = outputType.getDimSize(3);
- if (biasChannels != outputChannels && biasChannels != 1)
+ if (!ShapedType::isDynamic(outputChannels) &&
+ biasChannels != outputChannels && biasChannels != 1)
return emitOpError(
"bias channels expected to be equal to output channels (")
<< outputChannels << ") or 1, got " << biasChannels;
diff --git a/mlir/test/Dialect/Tosa/tosa-infer-shapes.mlir b/mlir/test/Dialect/Tosa/tosa-infer-shapes.mlir
index e280c1155f526..d0f40279421f4 100644
--- a/mlir/test/Dialect/Tosa/tosa-infer-shapes.mlir
+++ b/mlir/test/Dialect/Tosa/tosa-infer-shapes.mlir
@@ -1032,6 +1032,15 @@ func.func @transpose_conv2d_strided(%arg0: tensor<1x5x7x1xf32>, %arg1: tensor<1x
// -----
+// CHECK-LABEL: @transpose_conv2d_dynamic_out_channels
+func.func @transpose_conv2d_dynamic_out_channels(%arg0: tensor<2x1x1x3xf32>, %arg1: tensor<5x3x6x3xf32>, %arg2: tensor<5xf32>, %arg3: tensor<1xf32>, %arg4: tensor<1xf32>) {
+ // CHECK: -> tensor<2x3x6x5xf32>
+ %0 = tosa.transpose_conv2d %arg0, %arg1, %arg2, %arg3, %arg4 {acc_type = f32, out_pad = array<i64: 0, 0, 0, 0>, stride = array<i64: 1, 1>} : (tensor<2x1x1x3xf32>, tensor<5x3x6x3xf32>, tensor<5xf32>, tensor<1xf32>, tensor<1xf32>) -> tensor<2x3x6x?xf32>
+ return
+}
+
+// -----
+
// CHECK-LABEL: @resize_int_horizontal
func.func @resize_int_horizontal(%arg0: tensor<1x15x13x1xi8>) {
%scale = tosa.const_shape { values = dense<[11, 7, 89, 6]> : tensor<4xindex> } : () -> !tosa.shape<4>
|
GeorgeARM
approved these changes
Jul 15, 2025
udaya-ranga
reviewed
Jul 16, 2025
| // CHECK-LABEL: @transpose_conv2d_dynamic_out_channels | ||
| func.func @transpose_conv2d_dynamic_out_channels(%arg0: tensor<2x1x1x3xf32>, %arg1: tensor<5x3x6x3xf32>, %arg2: tensor<5xf32>, %arg3: tensor<1xf32>, %arg4: tensor<1xf32>) { | ||
| // CHECK: -> tensor<2x3x6x5xf32> | ||
| %0 = tosa.transpose_conv2d %arg0, %arg1, %arg2, %arg3, %arg4 {acc_type = f32, out_pad = array<i64: 0, 0, 0, 0>, stride = array<i64: 1, 1>} : (tensor<2x1x1x3xf32>, tensor<5x3x6x3xf32>, tensor<5xf32>, tensor<1xf32>, tensor<1xf32>) -> tensor<2x3x6x?xf32> |
There was a problem hiding this comment.
Worth having a negative test as well ? (Or is it already one of the tests in this file ?)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This commit fixes a transpose_conv2d verifier check which compares the output channels size to the bias size. The check didn't make sure output channels were static before performing the comparison. This lead to failures such as:
when the output channels size was dynamic.