[llvm] [AMDGPU] Fold saddsat(x, -C) to ssubsat(x, C) when C is an inline constant (PR #216592)

[Matt Arsenault via llvm-commits]

https://github.com/arsenm commented:

Also handle the 16-bit case? what about unsigned cases?

https://github.com/llvm/llvm-project/pull/216592