Convert the single-precision floating point value x to a signed integer in round-down (to negative infinity) mode.
x