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