[llvm] AMDGPU: Fold frame indexes into s_or_b32 and s_and_b32 (PR #102345)

[Matt Arsenault via llvm-commits]

arsenm wrote:

For the or case, I think we need to check the disjoint flag on the instruction (and also transform the final or into an add, if we can fold the offsets) 

https://github.com/llvm/llvm-project/pull/102345