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

[Ricardo Jesus via llvm-commits]

rj-jesus wrote:

> It can certainly be separate, this makes sense on its own. It is not necessary either way - it is a very minor missed optimization.

Thanks very much, that makes sense - in that case I'll merge this soon and look into the missed optimisation later. :)

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