Convert the single-precision floating point value x to a signed 64-bit integer in round-to-nearest-even mode.
x