RE: [photran] Problems looking at variables during debugging

["Simon Rees" <sjrees@xxxxxxxxx>]

Ralph,
Thanks for that. Its good to hear somebody is working on this. Another
error message I regularly get, that you might mention to your team, is
"Invalid C/C++ type code 12 in symbol table."

I guess I'll watch this space and see how your students get on.
Thanks, Simon

-----Original Message-----
From: photran-bounces@xxxxxxxxxxx [mailto:photran-bounces@xxxxxxxxxxx]
On Behalf Of Ralph Johnson
Sent: 15 February 2007 11:44
To: Photran Information
Subject: Re: [photran] Problems looking at variables during debugging


> Can anybody help with my debugging problem - below - as I haven't been

> able to figure this out myself. I am hoping to use Photran but being 
> able to use the debugging tool is something of a make-or-break issue 
> for me.

The debugger has a lot of problem all stemming from the fact that gdb
isn't that good with Fortran.  It can be better, though.

There is a group of undergrads working on fixing the Photran debugger.
They have a fix for the U variable problem.  The U variables are
termpoararies inserted by the compiler.  They have fixed a number of
other problems, as well, though they haven't fixed them all.  The big
problem is one of packaging.  Photran shares its debugger with CDT, so
they are figuring out how to package their changes in such a wa that
they can fix the Photran debugger without breaking the CDT debugger.
When they figure that out, they will release their changes.

I don't recall hearing about the problem involving extracting a
component from something that is not a structure.  I'll talk to them and
see if it is one of the bugs that they fixed.

This is a class project, so it ought to be done by the end of semester,
but probably not in the next month.

-Ralph Johnson
_______________________________________________
photran mailing list
photran@xxxxxxxxxxx https://dev.eclipse.org/mailman/listinfo/photran