[Mlir-commits] [mlir] [mlir][linalg] Fix splat fold crash on non-TypedAttr element types (PR #218012)

[Renato Golin]

================
@@ -1350,6 +1367,23 @@ func.func @transpose_non_splat_constant(%init: tensor<3x2xf32>) -> tensor<3x2xf3
   func.return %transpose : tensor<3x2xf32>
 }
 
+
+// -----
+
+// Same as @broadcast_splat_constant_complex, for the transpose pattern that
----------------
rengolin wrote:

This comment is contextual to the change, that only makes sense to people reading the PR, not the code afterwards. Please replicate the comment above here.

https://github.com/llvm/llvm-project/pull/218012