Hello,
I just received the most elegant solution from Stephan Zitz:
> Regarding your floating exception error:
>
> a 'Floating Exception' is merely a signal the kernel sends to the
> process. The signal is called 'SIGFPE', and, by default, a process
> dumps core.
>
> If you are absolutely certain that it should merely continue on it's
> way, ignoring this, you could tell the process to ignore the signal:
>
> #include <signal.h>
> ...
> signal( SIGFPE, SIG_IGN );
> /* here is the code that causes the sigfpe
> signal( SIGFPE, SIG_DFL );
> ...
>
Indeed, this works fine for me!
>
> Again, it is possible that the dec compiler would, under normal
> circumstances, do something to 'fixup' the FPE. I am not absolutely
> familiar with the internals of the dec compiler, but I could easily see
> the dec compiler doing something analogous to this with dividing by
> zero: i.e., if you attempt to do x/0, instead of dumping core, the dec
> compiler might 'fix up' that value to be Infinity.
Thanks again,
Erik-II
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Erik Nolf email: Erik.Nolf_at_rug.ac.be
RUG-ELIS (MEDISIP) tel : +32-9-264.66.28
Proeftuinstraat 86 fax : +32-9-264.66.99
B-9000 Gent, Belgium www :
http://petaxp.rug.ac.be/~nolf
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Received on Wed Apr 29 1998 - 16:23:38 NZST