Convert the single-precision floating point value x to an unsigned 64-bit integer in round-down (to negative infinity) mode.
x