VPEXTRW $17, X20, AX              // 62b17d08c5c411 or 62e37d0815e011 or 62e3fd0815e011
	VPEXTRW $127, X20, AX             // 62b17d08c5c47f or 62e37d0815e07f or 62e3fd0815e07f
	VPEXTRW $17, X20, SP              // 62b17d08c5e411 or 62e37d0815e411 or 62e3fd0815e411
	VPEXTRW $127, X20, SP             // 62b17d08c5e47f or 62e37d0815e47f or 62e3fd0815e47f

	VPCOMPRESSB Y24, K1, Y24                           // 62027d2963c0
	VPCOMPRESSB Y21, K1, Y24                           // 62827d2963e8
	VPCOMPRESSB Y5, K1, -17(BP)(SI*8)                  // 62f27d29636cf5ef
	VPCOMPRESSB Y24, K1, -17(BP)(SI*8)                 // 62627d296344f5ef
	VPCOMPRESSB Y21, K1, -17(BP)(SI*8)                 // 62e27d29636cf5ef
	VPCOMPRESSB Y5, K1, (R15)                          // 62d27d29632f

	var size = 128
	if testing.Short() {
		size = 32
	}
	a := make([]byte, size)
	b := make([]byte, size)
	b_init := make([]byte, size)
	// randomish but deterministic data
	for i := 0; i < size; i++ {
		a[i] = byte(17 * i)
		b_init[i] = byte(23*i + 100)
	}

	for len := 0; len <= size; len++ {
		for x := 0; x <= size-len; x++ {
			for y := 0; y <= size-len; y++ {
				copy(b, b_init)
				copy(b[y:y+len], a[x:x+len])