[Mlir-commits] [mlir] [mlir][Vector] Reject scalable dst dim in contract-to-dot lowering (PR #224793)

[Federico Bruzzone]

================

----------------
FedericoBruzzone wrote:

1. Both negative tests have dim 0 scalable. 
    Can you please add a case with only the inner dst dim scalable, e.g., `vector<2x3xf32>, vector<3x[2]xf32> into vector<2x[2]xf32>`?

2. Can you please add a positive test where only the reduction dim is scalable (e.g., `vector<4x[8]xf32>` for a  matvec). It should still lower to `vector.reduction` on `vector<[8]xf32>`, right?

https://github.com/llvm/llvm-project/pull/224793