Split a double-precision floating-point number into a higher order word and a lower order word.
Multiply a double-precision floating-point number by an integer power of two.
The bias of a double-precision floating-point number's exponent.
The maximum biased base 2 exponent for a subnormal double-precision floating-point number.