use core::arch::x86_64 as simd;

#[no_mangle]
fn stuff(bytes: &[u8]) -> u16 {
    unsafe {
        let bytes: [i32; 4] = *(bytes.as_ptr() as *const [i32; 4]);

        let vec = simd::_mm_set_epi32(
            bytes[3], bytes[2], bytes[1], bytes[0],
        );

        simd::_mm_movemask_epi8(vec) as u16
    }
}

static BYTES: &[u8] = &[
    // We ignore this byte, it's only here to screw with alignment
    0x00,

    0xFF,0x00,0x00,0x00,
    0x00,0x00,0x00,0x00,
    0xFF,0xFF,0xFF,0xFF,
    0xFF,0xFF,0xFF,0x00,
];

fn main() {
    let bits = stuff(&BYTES[1..]);

    // Bits are in reverse order due to LE architecture
    assert_eq!(bits, 0b0111_1111_0000_0001);

    println!("{:016b}", bits);
}