[Mlir-commits] [mlir] [MLIR][Vector]: Generalize conversion of `vector.insert` to LLVM in line with `vector.extract` (PR #128915)

[Kunwar Grover]

================
@@ -793,10 +803,10 @@ func.func @insert_scalar_into_vec_2d_f32_dynamic_idx_scalable(%arg0: vector<1x[1
   return %0 : vector<1x[16]xf32>
 }
 
-// Multi-dim vectors are not supported but this test shouldn't crash.
-
 // CHECK-LABEL: @insert_scalar_into_vec_2d_f32_dynamic_idx_scalable(
-//       CHECK:   vector.insert
+//       CHECK:   llvm.extractvalue {{.*}} : !llvm.array<1 x vector<[16]xf32>>
+//       CHECK:   llvm.insertelement {{.*}} : vector<[16]xf32>
+//       CHECK:   llvm.insertvalue {{.*}} : !llvm.array<1 x vector<[16]xf32>>
----------------
Groverkss wrote:

Can we have a negative test for when dynamic indices fail?

https://github.com/llvm/llvm-project/pull/128915