[llvm] [LLVM][CodeGen][AArch64] Improve generated code for scalar min(a,b)-max(a,b). (PR #223697)

[Paul Walker via llvm-commits]

paulwalker-arm wrote:

Abandoned in favour of https://github.com/llvm/llvm-project/pull/223725.

https://github.com/llvm/llvm-project/pull/223697