[llvm] [AMDGPU] Fix negative immediate offset for unbuffered smem loads (PR #89165)

[Matt Arsenault via llvm-commits]

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

I think this code needs a cleanup. I have a hard time following what's going on with the recursion and splitting up the offsets. The addressing isn't that complex 

https://github.com/llvm/llvm-project/pull/89165