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