[llvm] [CodeGen] Don't include aliases in RegisterClassInfo::IgnoreCSRForAllocOrder (PR #80015)

[Jay Foad via llvm-commits]

jayfoad wrote:

> It looks like this approach is faster on AMDGPU, but slower on more conventional architectures:

Thanks for keeping me honest! I will revert and rethink.

https://github.com/llvm/llvm-project/pull/80015