[llvm] [AMDGPU] fix SIPeepholeSDWA optimization for fp16 (PR #109395)

[Pankaj Dwivedi via llvm-commits]

PankajDwivedi-25 wrote:

Right, After Including this patch sorting output for some int8 cases has output values as  `-1`.
These values appear to have been converted to `-1` from the original input provided which originally does not have any `-1` in the list.

https://github.com/llvm/llvm-project/pull/109395