Re: Debugging propagated events

2016-11-29 Thread zahlenmeer
I added some logic and after that I got a warning that there is a
mismatch between cairo_save and cairo_restore. So I fixed that and I got
it working. The cairo logic was faulty all the time but only some new
lines of code triggered that warning.
So I guess I was just lucky that I stopped debugging for a moment and
added those few lines.

On 11/29/16 03:13, Jim Charlton wrote:
> In Gtkmm (C++) you can overload the on_draw() function with your own. 
> This might give you a foothold in the on_draw() routine that is called
> after a redraw event occurs.  I guess in GTK you would have to connect a
> callback function to the "draw" signal.  Might help.??
> 
> jim...
> 
> 
> On 2016-11-28 04:50 PM, zahlenm...@gmx.de wrote:
>> This was one of the first things I did. Debug compilation option, debug
>> CFLAGS, no stripping - without a change. The backtrace only shows 3
>> lines, my draw function and two from libc.
>> The faulty address is not null. But the segfault happens after the
>> return where all local variables are no longer in scope, and the
>> addresses of the widget and the cairo context are valid.
>>
>> On 11/29/16 00:44, Alfonso Arbona Gimeno wrote:
>>> Hi,
>>>
>>> Why don't you link against the debug version of the library so that it
>>> shows you the complete stack?
>>> It will give you some extra clues on why the segfault is happening.
>>>
>>> Regards,
>>> Alfonso Arbona
>>>
>>> On November 29, 2016 6:17:05 AM GMT+09:00, zahlenm...@gmx.de wrote:
>>>
>>>  For a custom widget I have a draw function. I do some drawing
>>> there and
>>>  when I am finished I return FALSE to let the parent draw.
>>> Immediately
>>>  after the return statement I get a segfault. gdb's backtrace
>>> only shows
>>>  my draw function and an address I cannot associate to any object.
>>>  Are there some general tricks how to debug such errors?
>>>
>>>  Kind regards
>>> 
>>> 
>>>
>>>  gtk-app-devel-list mailing list
>>>  gtk-app-devel-list@gnome.org
>>>  https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list
>>>
>> ___
>> gtk-app-devel-list mailing list
>> gtk-app-devel-list@gnome.org
>> https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list
> 
> 
> ___
> gtk-app-devel-list mailing list
> gtk-app-devel-list@gnome.org
> https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list
___
gtk-app-devel-list mailing list
gtk-app-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list


Re: Debugging propagated events

2016-11-28 Thread Jim Charlton
In Gtkmm (C++) you can overload the on_draw() function with your own.  
This might give you a foothold in the on_draw() routine that is called 
after a redraw event occurs.  I guess in GTK you would have to connect a 
callback function to the "draw" signal.  Might help.??


jim...


On 2016-11-28 04:50 PM, zahlenm...@gmx.de wrote:

This was one of the first things I did. Debug compilation option, debug
CFLAGS, no stripping - without a change. The backtrace only shows 3
lines, my draw function and two from libc.
The faulty address is not null. But the segfault happens after the
return where all local variables are no longer in scope, and the
addresses of the widget and the cairo context are valid.

On 11/29/16 00:44, Alfonso Arbona Gimeno wrote:

Hi,

Why don't you link against the debug version of the library so that it
shows you the complete stack?
It will give you some extra clues on why the segfault is happening.

Regards,
Alfonso Arbona

On November 29, 2016 6:17:05 AM GMT+09:00, zahlenm...@gmx.de wrote:

 For a custom widget I have a draw function. I do some drawing there and
 when I am finished I return FALSE to let the parent draw. Immediately
 after the return statement I get a segfault. gdb's backtrace only shows
 my draw function and an address I cannot associate to any object.
 Are there some general tricks how to debug such errors?

 Kind regards
 

 gtk-app-devel-list mailing list
 gtk-app-devel-list@gnome.org
 https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list


___
gtk-app-devel-list mailing list
gtk-app-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list



___
gtk-app-devel-list mailing list
gtk-app-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list


Re: Debugging propagated events

2016-11-28 Thread Alfonso Arbona Gimeno
Hi,

Why don't you link against the debug version of the library so that it shows 
you the complete stack?
It will give you some extra clues on why the segfault is happening.

Regards,
Alfonso Arbona

On November 29, 2016 6:17:05 AM GMT+09:00, zahlenm...@gmx.de wrote:
>For a custom widget I have a draw function. I do some drawing there and
>when I am finished I return FALSE to let the parent draw. Immediately
>after the return statement I get a segfault. gdb's backtrace only shows
>my draw function and an address I cannot associate to any object.
>Are there some general tricks how to debug such errors?
>
>Kind regards
>___
>gtk-app-devel-list mailing list
>gtk-app-devel-list@gnome.org
>https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list
___
gtk-app-devel-list mailing list
gtk-app-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list


Re: Debugging propagated events

2016-11-28 Thread zahlenmeer
This was one of the first things I did. Debug compilation option, debug
CFLAGS, no stripping - without a change. The backtrace only shows 3
lines, my draw function and two from libc.
The faulty address is not null. But the segfault happens after the
return where all local variables are no longer in scope, and the
addresses of the widget and the cairo context are valid.

On 11/29/16 00:44, Alfonso Arbona Gimeno wrote:
> Hi,
> 
> Why don't you link against the debug version of the library so that it
> shows you the complete stack?
> It will give you some extra clues on why the segfault is happening.
> 
> Regards,
> Alfonso Arbona
> 
> On November 29, 2016 6:17:05 AM GMT+09:00, zahlenm...@gmx.de wrote:
> 
> For a custom widget I have a draw function. I do some drawing there and
> when I am finished I return FALSE to let the parent draw. Immediately
> after the return statement I get a segfault. gdb's backtrace only shows
> my draw function and an address I cannot associate to any object.
> Are there some general tricks how to debug such errors?
> 
> Kind regards
> 
> 
> gtk-app-devel-list mailing list
> gtk-app-devel-list@gnome.org
> https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list
> 
___
gtk-app-devel-list mailing list
gtk-app-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list


Debugging propagated events

2016-11-28 Thread zahlenmeer
For a custom widget I have a draw function. I do some drawing there and
when I am finished I return FALSE to let the parent draw. Immediately
after the return statement I get a segfault. gdb's backtrace only shows
my draw function and an address I cannot associate to any object.
Are there some general tricks how to debug such errors?

Kind regards
___
gtk-app-devel-list mailing list
gtk-app-devel-list@gnome.org
https://mail.gnome.org/mailman/listinfo/gtk-app-devel-list