Hi -- Anyone seen a problem with dia where it segfaults after many
thousand nested renderer_gdk_copy_to_window() calls?
A stack trace looks like this (with about 30000 lines cut):
#30380 0x8067ef4 in renderer_gdk_copy_to_window ()
#30381 0x807c6a6 in shape_info_print ()
#30382 0x805df57 in ddisplay_flush ()
#30383 0x806d25f in layer_render ()
#30384 0x806d1e2 in data_render ()
#30385 0x805e004 in ddisplay_render_pixmap ()
#30386 0x805deb6 in ddisplay_flush ()
#30387 0x8055dc7 in diagram_flush ()
#30388 0x8065827 in defaults_show ()
#30389 0x805b5d1 in ddisplay_canvas_events ()
#30390 0x4012a79d in gtk_marshal_BOOL__POINTER ()
#30391 0x400f2037 in gtk_handlers_run ()
#30392 0x400f152f in gtk_signal_real_emit ()
#30393 0x400ef800 in gtk_signal_emit ()
#30394 0x40122338 in gtk_widget_event ()
#30395 0x400c71a2 in gtk_propagate_event ()
#30396 0x400c64da in gtk_main_do_event ()
#30397 0x4016aab2 in gdk_event_dispatch ()
#30398 0x401972c6 in g_main_dispatch ()
#30399 0x40197801 in g_main_iterate ()
#30400 0x40197979 in g_main_run ()
#30401 0x400c5f3a in gtk_main ()
#30402 0x8063c03 in main ()
#30403 0x403c7cb3 in __libc_start_main (main=0x8063bf0 <main>, argc=1,
argv=0xbffff8e4, init=0x8053284 <_init>, fini=0x807f34c <_fini>,
rtld_fini=0x4000a350 <_dl_fini>, stack_end=0xbffff8dc)
at ../sysdeps/generic/libc-start.c:78
I get this *all the time*. I was using gtk 1.2.1, libxml 1.0.0, dia 0.80.
I moved to dia 0.82 & then libxml 1.4.0, and this behavior is preserved.
It's very unpredictable & can depend on really weird things (like whether
I use the scrollbar buttons at a particular moment).
b.c.
