Parameterized floating-point logarithm and exponential functions for FPGAs