[llvm] [LLVM][CodeGen][SVE] Add lowering for 3-way VECTOR_(DE)INTERLEAVE operations. (PR #162502)

[Paul Walker via llvm-commits]

paulwalker-arm wrote:

> > NOTE: I plan to follow up with a combine to spot cases where data is either read from (VECTOR_DEINTERLEAVE) or written to (VECTOR_INTERLEAVE) memory to remove the need to use a stack temporary.
> 
> Looks like there's already another PR for that: #162475

That's at the IR level.  I'm planning to remove all that (well the part that works with the vector intrinsics) in favour of a codegen based solution.

https://github.com/llvm/llvm-project/pull/162502