This is the mail archive of the fortran@gcc.gnu.org mailing list for the GNU Fortran project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: single to double conversion


On Wed, 27 Sep 2006, Steve Kargl wrote:

> >  If it is intentional behaviour,
> > can code like this, __which will never work__, be caught with a
warning?
>
> Get ftnchek.  It is an excellent tool for finding bugs.

I had not seen ftnchek before; it looks like a good tool.  I hope they
are able to extend it to f90 / f95 error checking.  For my everyday
programming, I'll be happy for gfortran to catch as many errors as
possible. :)

Re: Tim Prince's comments about -f2c, I think that is the heart of
the matter!  I am no C (or -f2c) expert, and did not realize that is why
g77 functions the way it does.  I'm happy to learn some things from this
discussion.  Thanks to Tobias Schlüte for pointing that out.

Thanks also to Tobias Burnus, for running small tests with different
compilers.  After some information and examples from the many replies, I
fully agree that gfortran is handling things the proper way.

So that it is more unlikely for anyone else to encounter this problem in
the future, I am one user voicing support for gfortran to one day have
the ability to warn about this error, similar to ftnchek, ifort, or NAG
f95.  It seems I was just "lucky" with g77 and DEC f90.

nolo contendere,

Brian


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]