Using the gtk 2.12.10 available for download on GTK+ for windows site  I see
this printed out

(23:54:51) stun: got public ip 208.120.218.196
(23:55:08) Gdk: gdkgc-win32.c:823: SaveDC failed: Not enough storage is
available to process this command.
(23:55:08) Gdk: gdkdrawable-win32.c:1534: BitBlt failed: The operation
completed successfully.
(23:55:08) Gdk: gdkgc-win32.c:963: RestoreDC failed: The parameter is
incorrect.
(23:55:08) Gdk: gdkgc-win32.c:823: SaveDC failed: Not enough storage is
available to process this command.
(23:55:08) Gdk: gdkgc-win32.c:963: RestoreDC failed: The parameter is
incorrect.
(23:55:08) Gdk: gdkpixmap-win32.c:302: CreateDIBSection failed: The parameter
is incorrect.
(23:55:08) Gdk: gdkpixmap-win32.c:114: DeleteObject failed: The operation
completed successfully.
(23:55:08) Gdk: _gdk_drawable_ref_cairo_surface: assertion `GDK_IS_DRAWABLE
(drawable)' failed

_gdk_drawable_ref_cairo_surface() can return NULL, but
gdk_window_begin_paint_region() does not check if it did so before passing the
result to cairo_surface_set_device_offset(), which also doesn't check if it's
NULL before dereferencing it. 

Is it possible my error is a NULL pointer dereference?
-------------------------------------------------------
Excerpt from gdkwindow.c line 970-1026 
...
void          
gdk_window_begin_paint_region (GdkWindow *window,
                               GdkRegion *region)
{
...
  paint->surface = _gdk_drawable_ref_cairo_surface (paint->pixmap);
  cairo_surface_set_device_offset (paint->surface,
                                   - paint->x_offset, - paint->y_offset);
...
}
...
-------------------------------------------------------
Excerpt from  gdkdraw.c lines 1257-1275
...
/**
 * _gdk_drawable_ref_cairo_surface:
 * @drawable: a #GdkDrawable
 * 
 * Obtains a #cairo_surface_t for the given drawable. If a
 * #cairo_surface_t for the drawable already exists, it will be
 * referenced, otherwise a new surface will be created.
 * 
 * Return value: a newly referenced #cairo_surface_t that points
 *  to @drawable. Unref with cairo_surface_destroy()
 **/
cairo_surface_t *
_gdk_drawable_ref_cairo_surface (GdkDrawable *drawable)
{
  g_return_val_if_fail (GDK_IS_DRAWABLE (drawable), NULL);

  return GDK_DRAWABLE_GET_CLASS (drawable)->ref_cairo_surface (drawable);
}
....

This comes in kind of a roundabout way from the reporter of this bug via the pidgin ticket.

Based on tml's suggestion in #gtk+, the output immediately prior to the crash with G_DEBUG=events:misc is :

679	(16:37:32) Gdk: gdkwindow-win32.c:2075: CopyCursor failed: The operation completed successfully.
680	(16:37:32) Gdk: gdkpixmap-win32.c:302: CreateDIBSection failed: The parameter is incorrect.
681	(16:37:32) Gdk: gdkpixmap-win32.c:114: DeleteObject failed: The operation completed successfully.
682	(16:37:32) Gdk: _gdk_drawable_ref_cairo_surface: assertion `GDK_IS_DRAWABLE (drawable)' failed 

(This seems like not a lot of output, should there be more?)

I'm sure that Dave will be happy to provide any additional output.

yes the output looked similar to that from my comment #1 but please let me know if I can provide any more information.  

May I ask why we are investigating this? 

It looked clear to me that a NULL pointer was being passed to a cairo function and according to the cairo people (Carl Worth) here  [http://bugs.freedesktop.org/show_bug.cgi?id=15761#c11] that is not allowed and should be guarded against in gdk_window_begin_paint_region().

why  _gdk_drawable_ref_cairo_surface() is returning NULL might be an important thing to figure out, but we should stop the crashing caused gdk_window_begin_paint_region() by since _gdk_drawable_ref_cairo_surface() is documented as possibly returning a null pointer. Because the cast to GDK_IS_DRAWABLE failed.

please see cairo bug comment# 13

http://bugs.freedesktop.org/show_bug.cgi?id=15761#c13

  ------- Comment  #13 From Chris Wilson  2008-10-10 11:30:08 PST  -------

The crash is caused by GTK+ passing a NULL surface to cairo, something which
should not happen. In particular, it looks like GTK+ made an error when trying
to create the win32 drawable and that should be fixed at source.

I get this crash daily.

reminds me on what I saw with bug #562574 - a GDI leak will crash a gtk+ application sooner or later and all the debug spew looks like one. 
You can check yourself with task manager (gdi objects) if the value is near 10000, which apparently is the limit on XP. To be a duplicate of bug #562574 Pidgin needs to create and (try to) destroy a lot of Pango(Cairo)Font though.

pidgin is def. leaking gdi.

every time I double click on a user to send an im, and then close that. there is always a net increase of at least 2 GDI objects in use by pidgin.

in the case of the steps used to cause this particular crash , opening the log viewer and clicking around. on different logs.  pidgin leaks 10-100 GDI per click.

how do we pinpoint whether pidgin is not using gtk correctly and thus pidgins fault, or that pidgin is using gtk correctly but gtk is not managing gdi correctly, as was found to be the case in the bug Hans mentioned.

also, just when opening the log viewer, 200 GDI are leaked each time.

downloaded Pango  	1.22.4 from here  http://www.gtk.org/download-windows.html 

this appears to have fixed the worst gdi leaks.

previous pango dlls were dated before the patch in bug #562574 was accepted (2008-11-28)


01/25/2009  01:23 AM    <DIR>          .
01/25/2009  01:23 AM    <DIR>          ..
11/04/2008  10:54 AM           320,299 libpango-1.0-0.dll
11/04/2008  10:54 AM            77,792 libpangocairo-1.0-0.dll
11/04/2008  10:54 AM           272,802 libpangoft2-1.0-0.dll
11/04/2008  10:54 AM            99,648 libpangowin32-1.0-0.dll
11/04/2008  10:54 AM            25,925 pango-querymodules.exe

resolving as duplicate

*** This bug has been marked as a duplicate of bug 562574 ***