Convert the double-precision floating point value x to a signed integer value in round-towards-zero mode.
x