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