[llvm] [AArch64] Use SVE for load-extend-broadcast (PR #221689)

[Jack Styles via llvm-commits]

Stylie777 wrote:

> It should be the other way around, if I understand this correctly. As there is a native instruction to support this pattern with SVE it should be the default codegen and CPUs for which it is slow for some reason should have it disabled. Using the existing target feature is probably the best bet.

Understood, thanks. There is a feature with @sdesmalen-arm mentioned, but this is targeting FP operations, whereas for this we would want to stop it regardless of type on some platforms. I think this still needs the new feature but I will restructure the new feature to reflect this change.

https://github.com/llvm/llvm-project/pull/221689