Function cross_block_alignr_128x4 

unsafe fn cross_block_alignr_128x4(
    a: [__m128i; 4],
    b: [__m128i; 4],
    shift_bytes: usize,
) -> [__m128i; 4]

Concatenates b and a (each N blocks) and extracts N blocks starting at byte offset shift_bytes. Extracts from [b : a] (b in low bytes, a in high bytes), matching alignr semantics.