AVX-512: Implemented vector shuffle lowering for v8i64 and v8f64 types.