[clang][AArch64][SVE] Avoid going through memory for VLAT <-> VLST casts

[clang][AArch64][SVE] Avoid going through memory for VLAT <-> VLST casts

This change makes use of the llvm.vector.extract intrinsic to avoid
going through memory when performing bitcasts between vector-length
agnostic types and vector-length specific types.

Depends on D91362

Reviewed By: c-rhodes

Differential Revision: https://reviews.llvm.org/D92761