[llvm] [AArch64] Don't merge branch conditions that both compare memory loads (PR #206504)

[David Green via llvm-commits]

https://github.com/davemgreen approved this pull request.

I'm not sure I 100% buy the reasoning here or the heuristic (not that spills are the problem, just that they couldn't be scheduled better), but it only goes back to the previous codegen in some cases and seems fine enough. LGTM, thanks for putting together a fix.

https://github.com/llvm/llvm-project/pull/206504