[llvm] [AMDGPU] [PeepholeOpt] Eliminate unnecessary packing in wider f16 vectors for sdwa/opsel-able instruction (PR #137137)

[Krzysztof Drewniak via llvm-commits]

https://github.com/krzysz00 commented:

Just one question, I'll let the compiler folks do the reviewing here

https://github.com/llvm/llvm-project/pull/137137