[Mlir-commits] [mlir] [mlir][vector] Fold vector extract from insert when trailing unit dims (PR #192109)

[Mehdi Amini]

joker-eph wrote:

> Do you have an example of where this is ambigious? In case of vector.extract, yes this was ambigious as you couldn't tell if the extract type was f32/vector<1xf32>, which is why it was good to remove that behavior.

OK I played with more examples, and we already reject all the cases I was thinking of actually. It seems I was misled by reading the pattern: I thought it was written to target constructs where this "shape_cast" was implicit in the extract, but the verifier already reject them (and so the pattern don't need to guard against them). Sorry for the noise!

https://github.com/llvm/llvm-project/pull/192109