[llvm] [NVPTX] Use PRMT instruction to lower i16 bswap (PR #168968)

[Artem Belevich via llvm-commits]

https://github.com/Artem-B approved this pull request.

I'm curious whether this makes any difference on the SASS level. The old shift/or pattern looks simple enough for ptxas to deal with.

https://github.com/llvm/llvm-project/pull/168968