Ok, I am working on the valgrind stuff, its my first time with this tool, so
i'll try to get information from it and contact you soon with as much
information from the error i can get.

Thanks
--Diego--

On Wed, May 7, 2008 at 9:15 PM, James Scott Jr <[EMAIL PROTECTED]> wrote:

>
>
> On Wed, 2008-05-07 at 21:02 -0600, Diego Rivera wrote:
> > Thanks for the quick answer, however, the GTK+ component is the same
> > process as the Console application. The console application checks for
> > some arguments and finishes by calling the working function of the
> > application with parameters that reflect the arguments received from
> > the console. My GtkButton callback just calls this last function with
> > some parameters i have selected.
> >
> > So no inter-process communication is requiered
> >
> > Thanks for the suggestion anyway
> >
> OK, then you need to post some relevant portion of the code if possible.
> These type of issues can be as casued by simple things like - using
> local stack memory for structures vs malloc or g_new0() memory for them.
>
> Post your main() and the callback for us to look at.  And please do
> consider the valgrind suggestion, or retry the gdb exercise.
>
> James,
>
> > -Diego--
> >
> > On Wed, May 7, 2008 at 8:55 PM, James Scott Jr <[EMAIL PROTECTED]>
> > wrote:
> >
> >
> >
> >         On Wed, 2008-05-07 at 20:33 -0600, Diego Rivera wrote:
> >         > Hi! I am not exactly new to GTK+, however I have come across
> >         a problem i
> >         > have no explanation for, and I'll be glad if someone could
> >         point me towards
> >         > a possible solution.
> >         >
> >         > I am currently developing an application that uses various
> >         algorithms for
> >         > sequence alignment in biological applications (DNA, RNA,
> >         Proteins and so
> >         > on). This application is mainly for scientific purposes.
> >         >
> >         > The core algorithms are already running fine as a console
> >         application in a
> >         > very stable version. I am working in adding some graphical
> >         interface to the
> >         > program using GTK+. So, I built a little Window that asks
> >         for the different
> >         > parameters and makes a call to the already defined functions
> >         in the console
> >         > application, which are indeed stable. However, when making
> >         the call to the
> >         > algorithm with GTK+ running, some of the pointers in my data
> >         structures get
> >         > lost, and I just get a Segmentation Fault.
> >         >
> >         > While trying to debug this weird behavior, I have narrowed
> >         the problem to
> >         > the fact of having a window running.
> >         >
> >         > Right now, I just have a main that creates a GtkWindow with
> >         a GtkButton, and
> >         > the callback of the button calls the stable routine of the
> >         console
> >         > application using some "hard-wired" parameters. I have made
> >         tests and the
> >         > console application gives the correct answer for those
> >         parameters. However,
> >         > when calling the routine from the callback, Segmentation
> >         Fault is the only
> >         > result I have been able to get.
> >         >
> >         > I have tried with different versions of GTK+ and the outcome
> >         is persistent.
> >         >
> >         > The data structure of the console application looks
> >         something like this:
> >         >
> >         > typedef struct{
> >         >     int length;
> >         >     char* seq;
> >         > }sequence_t;
> >         >
> >         > typedef struct{
> >         >     int n;
> >         >     int m;
> >         >     int** table;
> >         > }table_t;
> >         >
> >         > typedef struct{
> >         >     sequence_t* vp;
> >         >     sequence_t* wp;
> >         >     table_t* table;
> >         >     int scoring;
> >         >     int result;
> >         > }alignment_t;
> >         >
> >         > Using gdb i have found that the Segmentation Fault occurs at
> >         some point when
> >         > accesing the table from the alignment.
> >         >
> >         > I believe i have made no memory errors in my console
> >         application, since it
> >         > runs ok every time. However, i haven't seen the correct
> >         output from my GTK
> >         > application.
> >         >
> >         > Is there any hint that you could give me in order to correct
> >         the error?
> >         >
> >
> >         How did you add the GTK component to the existing console app;
> >         are they
> >         one process or two?
> >
> >         Your segfault suggest that they are two separate processes.
> >          In which
> >         case attempting to share memory elements will cause a segfault
> >         every
> >         time no matter how correct and hard-coded the structures.
> >
> >         You may need to consider designing in and using some type of
> >         IPC
> >         (Inter-Process Communications) like a named-pipe,
> >         shared-memory, socket
> >         communications, memory queue, etc.  Something designed to
> >         communicate
> >         between two active processes.
> >
> >         James,
> >
> >
> >         > Thanks
> >         > - Diego Rivera -
> >         > _______________________________________________
> >         > gtk-app-devel-list mailing list
> >         > gtk-app-devel-list@gnome.org
> >         > http://mail.gnome.org/mailman/listinfo/gtk-app-devel-list
> >
> >
> >
>
>
_______________________________________________
gtk-app-devel-list mailing list
gtk-app-devel-list@gnome.org
http://mail.gnome.org/mailman/listinfo/gtk-app-devel-list

Reply via email to