[llvm] [AMDGPU] Add pattern to select scalar ops for fshr with uniform operands (PR #165295)

[Jay Foad via llvm-commits]

https://github.com/jayfoad requested changes to this pull request.

Please don't add three copies of your new patterns. Just put them somewhere outside of any `True16Predicate = ...` clause.

https://github.com/llvm/llvm-project/pull/165295