HUGE_VAL is defined as DBL_MAX.
XGC supports all the math functions defined in the ANSI standard.
XGC does not support math functions for long double arguments. However because long double is defined to use the same machine representation as double, the ANSI standard functions may be used.
Functions generally do not raise exceptions for domain errors. Instead they return a value as if the argument was at the nearest limit of the domain. For example, asin (1.5) will return pi/2. Thus by ignoring the value of errno, and by masking the interrupts for arithmetic overflow, we get a saturated arithmetic behavior.
Where applicable, the maximum and RMS error values are given.