[Mlir-commits] [mlir] [MLIR][TOSA] Enable strict property assembly format (PR #217291)

[Luke Hutton]

================
@@ -74,9 +74,9 @@ func.func @unary_resize_bilinear_fp16(%arg0 : tensor<3x1x1x7xf16>) -> tensor<3x1
 
 // CHECK-LABEL: @unary_resize_nearest_i8
 func.func @unary_resize_nearest_i8(%arg0 : tensor<3x1x1x7xi8>) -> tensor<3x1x1x7xi8> {
-  %scale = tosa.const_shape { values = dense<[2, 1, 3, 1]> : tensor<4xindex> } : () -> !tosa.shape<4>
-  %offset = tosa.const_shape { values = dense<0> : tensor<2xindex> } : () -> !tosa.shape<2>
-  %border = tosa.const_shape { values = dense<0> : tensor<2xindex> } : () -> !tosa.shape<2>
+  %scale = tosa.const_shape values(dense<[2, 1, 3, 1]> : tensor<4xindex>) : () -> !tosa.shape<4>
+  %offset = tosa.const_shape values(dense<0> : tensor<2xindex>) : () -> !tosa.shape<2>
+  %border = tosa.const_shape values(dense<0> : tensor<2xindex>) : () -> !tosa.shape<2>
   %resize = tosa.resize %arg0, %scale, %offset, %border {mode = NEAREST_NEIGHBOR} : (tensor<3x1x1x7xi8>, !tosa.shape<4>, !tosa.shape<2>, !tosa.shape<2>) -> tensor<3x1x1x7xi8>
----------------
lhutton1 wrote:

I think it would be good to also update operations that use a custom parser to emulate the same assembly format as other operations. I'd also support removing these custom parsers which I think were only being used to avoid writing `#tosa.mode<NEAREST_NEIGHBOR>`

https://github.com/llvm/llvm-project/pull/217291