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