[Mlir-commits] [mlir] Fixes in 'tosa.reshape' lowering and folder (PR #85798)

[Spenser Bauman]

================
@@ -1,95 +1,282 @@
 // RUN: mlir-opt --split-input-file --tosa-to-tensor %s -o -| FileCheck %s
 
-// CHECK-LABEL: @test_reshape_downrank
-// CHECK-SAME: (%[[ARG0:[0-9a-zA-Z_]*]]
-func.func @test_reshape_downrank(%arg0: tensor<2x3xf32>) -> tensor<6xf32> {
-  // CHECK: [[RESHAPE:%.+]] = tensor.collapse_shape %[[ARG0]] {{\[}}[0, 1]]
-  %0 = "tosa.reshape"(%arg0) {new_shape = array<i64: 6>} : (tensor<2x3xf32>) -> tensor<6xf32>
-  // CHECK: return [[RESHAPE]]
-  return %0 : tensor<6xf32>
+// -----
+
+// CHECK-LABEL: test_reshape_0d_same_s2s_explicit
+// CHECK-SAME: %[[ARG_0:[a-zA-Z0-9_]+]]: tensor<f32>
+// CHECK: return %[[ARG_0]] : tensor<f32>
+func.func @test_reshape_0d_same_s2s_explicit(%arg0: tensor<f32>) -> tensor<f32> {
+  %0 = "tosa.reshape"(%arg0) {new_shape = array<i64>} : (tensor<f32>) -> tensor<f32>
----------------
sabauma wrote:

Could you add a test case for a reshape of a multi-dimensional (size 1) tensor down do a rank-0 tensor. I don't see any case like that. Possibly an example of the reverse as well. In my experience, rank-0 tensors tend to be edge cases in a lot of conversions.

https://github.com/llvm/llvm-project/pull/85798