[Mlir-commits] [mlir] [MLIR][Linalg] Introduce broadcast/transpose semantic to 'linalg.batc… (PR #122275)

[Adam Siemieniuk]

================
@@ -1142,3 +1142,121 @@ func.func @winograd_output_transform_output_width(%arg0: tensor<6x6x3x3x2x2xf32>
   %0 = linalg.winograd_output_transform m(4) r(3) ins(%arg0 : tensor<6x6x3x3x2x2xf32>) outs(%arg1 : tensor<2x12x11x2xf32>) -> tensor<2x12x11x2xf32>
   return %0 : tensor<2x12x11x2xf32>
 }
+
+// -----
+
+func.func @missing_indexing_map_batch_matmul(%arg0: tensor<?x?x?xf32>, %arg1: tensor<?x?x?xf32>, %arg2: tensor<?x?x?xf32>) {
----------------
adam-smnk wrote:

Args should have memref type, or op should be on tensors and have a result (same for the other cases).

https://github.com/llvm/llvm-project/pull/122275