[llvm] [RISCV] Attempt to widen SEW before generic shuffle lowering (PR #122311)

[Luke Lau via llvm-commits]

https://github.com/lukel97 approved this pull request.

LGTM + what Craig said. Test diff seems good so I don't think we need a heuristic. It looks like it has the benefit of shrinking the vrgather indices.

https://github.com/llvm/llvm-project/pull/122311