Implementation of "f64_pow" is incomplete

[chikuwait]

The pow function existing in math.h in libC is 2.0 when passed as 4.0 and 0.5 as arguments. However, in my implementation there is a problem that the values ​​do not match because the value of the second argument is cast to an integer type.

[chikuwait]

This problem can be solved by making full operation by porting fdlibm's pow function using the SoftFloat library