Cedric Gustin wrote:

Dave Aitel wrote:


Hopefully this is something you'll see and say "We fixed that already"

Replicating the wacky thread condition that causes this bug is somewhat
difficult in a small program. I can replicate it consistantly in CANVAS,
however. Here's a stack trace from what I believe is the offending
thread. The last few frames (USER32, etc) are going to be Python.exe
throwing up a dialog box to "debug" it. The actual bug is a double-free
bug, I believe. (Dies in Free() due to accessing invalid memory,
different every time, but I've seen a lot of double free bugs in my day,
and it's that, or something similar)

Platform:
GTK+ Development 2.4 rc22
Python 2.4
Python 2.4pygtk-2.4.1
Windows XP SP2
(program: CANVAS 5.5)



A simple question : did you try with more recent builds (GTK+ Developement 2.4 rc26 (from gladewin32)) or with GTK+ 2.6 (gtk-win32-devel-2.6-rc5.exe and pygtk-2.6.1) ?

Cedric



I gave it a shot with 2.6.4 rc1 (latest version I could find) and pygtk 2.6.1 and that crashed as well. Is there a "has all the symbols" way of debugging this? I wonder if cygwin's gdb likes threads more than normal gdb. :>

I can push into some Ollydbg time later this week and see if I can track it down.

-dave


_______________________________________________ pygtk mailing list [email protected] http://www.daa.com.au/mailman/listinfo/pygtk Read the PyGTK FAQ: http://www.async.com.br/faq/pygtk/

Reply via email to