Re: runtime math error handling

[FÖLDY Lajos]

On Tue, 14 Nov 2006, Zsely Istvan Gyula wrote:

> Dear Developers,
> I hope this is the most apropriate list for my problem.
> I developed a scientific program in F77 (compiled by linux g77 directly) and
> in a couple of cases
> mathematical errors occur (sqrt with a negative number, floating overflow).
> How can I handle these error in a source code procedure? I went through the
> online manual (http://gcc.gnu.org/onlinedocs/gcc-3.4.6/g77/), but did not
> find the soulution.
>
> Thank you for your help in advance!
> Best regards,
> Istvan Gy. Zsely

You can write an interface to the C99 fenv functions (see fenv.h, or
man fetestexcept)

regards,
lajos

ps: this list is about g++, so don't expect too many responses :-)