// padding is alwas binary 1 followed by binary 0s
        byte[] tail = new byte[padLen + 8];
        tail[0] = (byte)0x80;

        // append length before final transform:
        // save number of bits, casting the long to an array of 8 bytes
        // save low-order byte first.
        for (int i = 0; i < 8; i++)
            tail[padLen + i] = (byte)((count * 8) >>> (8 * i));

        engineUpdate(tail, 0, tail.length);