[llvm] [AMDGPU] Allow merging unordered and monotonic atomic loads in SILoadStoreOptimizer (PR #189932)

[Matt Arsenault via llvm-commits]

https://github.com/arsenm commented:

SILoadStoreOptimizer should not be performing these sorts of optimizations. This should be done on the IR (really this applies to most things here) 

https://github.com/llvm/llvm-project/pull/189932