[llvm] [AArch64] Use vecshiftL64 instead of vecshiftR64 to match scalar SLI imm. (PR #139904)

[Ricardo Jesus via llvm-commits]

rj-jesus wrote:

> LGTM, but it would be good to remove the instructions that do not do anything.

Indeed, would you rather it done here or separately?

https://github.com/llvm/llvm-project/pull/139904