[llvm] [AMDGPU][InstCombine] Optimize constant shuffle patterns (PR #192246)

[Jay Foad via llvm-commits]

https://github.com/jayfoad approved this pull request.

LGTM now.

Is there any value in also matching the DS_SWIZZLE modes FFT and ROTATE? Can be done as a follow-up.

https://github.com/llvm/llvm-project/pull/192246