pub fn u16x8_sub(a: v128, b: v128) -> v128
simd128
Subtracts two 128-bit vectors as if they were two packed eight 16-bit integers.