[Mlir-commits] [mlir] [mlir][tensor] Add runtime verification for expand_shape/concat/reshape (PR #236783)

[Matthias Springer]

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

----------------
matthias-springer wrote:

I'm surprised that this is even allowed. IMO, the verifier should enforce that the result type is `tensor<?x3xf32>`. To add more static type information, `tensor.cast` can be used.

https://github.com/llvm/llvm-project/pull/236783