This is the mail archive of the
gsl-discuss@sources.redhat.com
mailing list for the GSL project.
Re: exception handling complex elementary functions (fwd)
Brian Gough writes:
> > I've done as requested. I decided not to use exact same format as
> > specfunc/error.h so the returned result is always similar to that of the
> > original functions. Attached are complex\math.c (with some #ifndef
> > additions) and complex\math_eh.c. Please let me know if you have any
> > other requests.
Hello,
Could you write some test cases to exercise the new parts of the code.
The existing test suite (test.c) doesn't cause any floating-point
overflows or underflows in the new functions.
I made the file into a drop-in replacement for math.c:
http://www.network-theory.co.uk/gsl/dev/math_eh.c
Here is the output of coverage testing with gcov for 'make check':
http://www.network-theory.co.uk/gsl/dev/math_eh.c.gcov
--
Brian Gough