Convert an integer-valued double-precision floating-point number to a signed 64-bit integer byte array according to host byte order.
