[llvm-branch-commits] [llvm] [AArch64] Fold zero-high vector inserts in MI peephole optimisation (PR #182835)

[via llvm-branch-commits]

Lukacma wrote:

It seems that codegen for some tests your patch influences is different upstream  than in this patch. It might be good idea to sync with upstream.

https://github.com/llvm/llvm-project/pull/182835