[Mlir-commits] [mlir] [mlir][vector] Don't fold in_bounds for scalable vector dimensions (PR #213506)

[Andrzej Warzyński]

https://github.com/banach-space approved this pull request.

Thank you, LGTM % nit

> Confirmed on SVE hardware (AWS Graviton3, 256-bit vectors, vscale = 2)

Do you mind sharing the context in which you have been using that machine? You are the first person to be testing things on that hardware, so it made me super curious. Absolutely no problem if you can't share publicly :) 

https://github.com/llvm/llvm-project/pull/213506