-
Notifications
You must be signed in to change notification settings - Fork 12.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[mlir][tosa] Only match rfft2d of floats in linalg conversion #93432
Conversation
This prevents an assertion being triggered by the cast to FloatType.
@llvm/pr-subscribers-mlir @llvm/pr-subscribers-mlir-tosa Author: Kavan Bickerstaff (KB9) ChangesThis prevents an assertion being triggered by the cast to FloatType. Fixes #92064 Full diff: https://github.com/llvm/llvm-project/pull/93432.diff 2 Files Affected:
diff --git a/mlir/lib/Conversion/TosaToLinalg/TosaToLinalg.cpp b/mlir/lib/Conversion/TosaToLinalg/TosaToLinalg.cpp
index e6ba6e6bc602d..8ad8e41414656 100644
--- a/mlir/lib/Conversion/TosaToLinalg/TosaToLinalg.cpp
+++ b/mlir/lib/Conversion/TosaToLinalg/TosaToLinalg.cpp
@@ -2324,7 +2324,10 @@ struct RFFT2dConverter final : public OpRewritePattern<RFFT2dOp> {
auto loc = rfft2d.getLoc();
auto input = rfft2d.getInput();
auto elementType =
- cast<FloatType>(cast<ShapedType>(input.getType()).getElementType());
+ dyn_cast<FloatType>(cast<ShapedType>(input.getType()).getElementType());
+ if (!elementType)
+ return rewriter.notifyMatchFailure(rfft2d,
+ "only supports float element types");
// Compute the output type and set of dynamic sizes
llvm::SmallVector<Value> dynamicSizes;
diff --git a/mlir/test/Conversion/TosaToLinalg/tosa-to-linalg-invalid.mlir b/mlir/test/Conversion/TosaToLinalg/tosa-to-linalg-invalid.mlir
index ad65410e635e9..b78577275a52a 100644
--- a/mlir/test/Conversion/TosaToLinalg/tosa-to-linalg-invalid.mlir
+++ b/mlir/test/Conversion/TosaToLinalg/tosa-to-linalg-invalid.mlir
@@ -27,3 +27,12 @@ func.func @unranked_add(%arg0 : tensor<10x10xf32> , %arg1 : tensor<10x10xf32>, %
%2 = tosa.reshape %0 {new_shape = array<i64: 10, 10>} : (tensor<*xf32>) -> tensor<10x10xf32>
return %2 : tensor<10x10xf32>
}
+
+// -----
+
+// CHECK-LABEL: @rfft2d_with_non_float_type
+func.func @rfft2d_with_non_float_type(%arg0 : tensor<1x1x1xi32>) -> (tensor<1x1x1xi32>, tensor<1x1x1xi32>) {
+ // expected-error@+1 {{failed to legalize operation 'tosa.rfft2d'}}
+ %real, %imag = tosa.rfft2d %arg0 : (tensor<1x1x1xi32>) -> (tensor<1x1x1xi32>, tensor<1x1x1xi32>)
+ return %real, %imag : tensor<1x1x1xi32>, tensor<1x1x1xi32>
+}
|
Ping |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Sorry for the delay!
Thanks for the review! Would you mind merging this? I don't have the permissions to do so. |
Done! |
This prevents an assertion being triggered by the cast to FloatType.
Fixes #92064