[llvm] [NVPTX] Fix BFE selection for i64 values (PR #228685)

[Alex MacLean via llvm-commits]

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

Does this change improve performance or final generated SASS on any supported architecture? It seems like this instruction is not natively supported on any architecture. If anything, I'd lean towards disabling BFE/BFI folding on all modern SMs.

https://github.com/llvm/llvm-project/pull/228685