GNOME Bugzilla – Bug 377079
crash in at-spi (fixed in version 1.7.14)
Last modified: 2007-12-09 12:08:51 UTC
Version: 2.16.1 What were you doing when the application crashed? Ran the help from Power Information Dialog and clicked on Usage. It is crashing independent of where I click inside the window. Distribution: Ubuntu 7.04 (feisty) Gnome Release: 2.17.2 2006-11-14 (Ubuntu) BugBuddy Version: 2.16.0 Memory status: size: 124575744 vsize: 0 resident: 124575744 share: 0 rss: 49344512 rss_rlim: 0 CPU usage: start_time: 1163959485 rtime: 0 utime: 119 stime: 0 cutime:108 cstime: 0 timeout: 11 it_real_value: 0 frequency: 2 Backtrace was generated from '/usr/bin/yelp' Using host libthread_db library "/lib/tls/i686/cmov/libthread_db.so.1". [Thread debugging using libthread_db enabled] [New Thread -1209809200 (LWP 16632)] [New Thread -1275348080 (LWP 16636)] [New Thread -1266852976 (LWP 16635)] [New Thread -1228252272 (LWP 16634)] 0xffffe410 in __kernel_vsyscall ()
+ Trace 87753
Thread 1 (Thread -1209809200 (LWP 16632))
Thanks for taking the time to report this bug. Unfortunately, that stack trace is missing some elements that will help a lot to solve the problem, so it will be hard for the developers to fix that crash. Can you get us a stack trace with debugging symbols? Please see http://live.gnome.org/GettingTraces for more information on how to do so. Thanks in advance!
Distribution: Ubuntu 7.04 (feisty) Gnome Release: 2.17.2 2006-11-14 (Ubuntu) BugBuddy Version: 2.16.0 Memory status: size: 124235776 vsize: 0 resident: 124235776 share: 0 rss: 49381376 rss_rlim: 0 CPU usage: start_time: 1164035449 rtime: 0 utime: 140 stime: 0 cutime:129 cstime: 0 timeout: 11 it_real_value: 0 frequency: 3 Backtrace was generated from '/usr/bin/yelp' Using host libthread_db library "/lib/tls/i686/cmov/libthread_db.so.1". [Thread debugging using libthread_db enabled] [New Thread -1209702704 (LWP 7888)] [New Thread -1275241584 (LWP 7894)] [New Thread -1266746480 (LWP 7893)] [New Thread -1228145776 (LWP 7892)] 0xffffe410 in __kernel_vsyscall ()
+ Trace 87990
Thread 1 (Thread -1209702704 (LWP 7888))
*** Bug 377086 has been marked as a duplicate of this bug. ***
*** Bug 377106 has been marked as a duplicate of this bug. ***
*** Bug 377226 has been marked as a duplicate of this bug. ***
*** Bug 376407 has been marked as a duplicate of this bug. ***
*** Bug 375793 has been marked as a duplicate of this bug. ***
*** Bug 377487 has been marked as a duplicate of this bug. ***
*** Bug 375586 has been marked as a duplicate of this bug. ***
*** Bug 375435 has been marked as a duplicate of this bug. ***
*** Bug 377557 has been marked as a duplicate of this bug. ***
*** Bug 377844 has been marked as a duplicate of this bug. ***
*** Bug 377690 has been marked as a duplicate of this bug. ***
*** Bug 377671 has been marked as a duplicate of this bug. ***
*** Bug 377625 has been marked as a duplicate of this bug. ***
*** Bug 377325 has been marked as a duplicate of this bug. ***
*** Bug 377294 has been marked as a duplicate of this bug. ***
*** Bug 377267 has been marked as a duplicate of this bug. ***
*** Bug 377193 has been marked as a duplicate of this bug. ***
*** Bug 377174 has been marked as a duplicate of this bug. ***
*** Bug 377113 has been marked as a duplicate of this bug. ***
*** Bug 377082 has been marked as a duplicate of this bug. ***
*** Bug 376976 has been marked as a duplicate of this bug. ***
*** Bug 376567 has been marked as a duplicate of this bug. ***
*** Bug 376436 has been marked as a duplicate of this bug. ***
*** Bug 376293 has been marked as a duplicate of this bug. ***
*** Bug 376264 has been marked as a duplicate of this bug. ***
*** Bug 376260 has been marked as a duplicate of this bug. ***
*** Bug 376179 has been marked as a duplicate of this bug. ***
*** Bug 376080 has been marked as a duplicate of this bug. ***
*** Bug 376068 has been marked as a duplicate of this bug. ***
*** Bug 376066 has been marked as a duplicate of this bug. ***
*** Bug 376035 has been marked as a duplicate of this bug. ***
*** Bug 375843 has been marked as a duplicate of this bug. ***
*** Bug 375820 has been marked as a duplicate of this bug. ***
*** Bug 375803 has been marked as a duplicate of this bug. ***
*** Bug 375802 has been marked as a duplicate of this bug. ***
*** Bug 375768 has been marked as a duplicate of this bug. ***
*** Bug 375699 has been marked as a duplicate of this bug. ***
*** Bug 375616 has been marked as a duplicate of this bug. ***
*** Bug 375607 has been marked as a duplicate of this bug. ***
*** Bug 375468 has been marked as a duplicate of this bug. ***
*** Bug 375439 has been marked as a duplicate of this bug. ***
*** Bug 375422 has been marked as a duplicate of this bug. ***
*** Bug 375942 has been marked as a duplicate of this bug. ***
CCing Daniel Holbach. Daniel is this an Ubuntu / downstream problem (note it only happens in Feisty)?
*** Bug 379129 has been marked as a duplicate of this bug. ***
With so many reports, including mine, why is the status still 'UNCONFIRMED'?
I'm not sure what specific to Ubuntu could cause the bug. Interestingly enough the backtrace mentions /usr/lib/gtk-2.0/modules/libatk-bridge.so - this looks more like an at-spi problem.
*** Bug 379370 has been marked as a duplicate of this bug. ***
*** Bug 379222 has been marked as a duplicate of this bug. ***
Not Ubuntu-specific, bug 379222 is a dup from FC6
Bug 377568 has more information and I can confirm: the problem is not in at-spi 1.7.12. Some investigation showed that spi_atk_bridge_get_registry_ior() in at-spi/atk-bridge/bridge.c gives NULL, that's why g_warning ("IOR not set."); in spi_atk_bridge_get_registry() is issued and ultimately spi_atk_bridge_get_registry () == CORBA_OBJECT_NIL which results in g_error ("Could not locate registry"); Why XGetWindowProperty(bridge_display, XDefaultRootWindow (bridge_display), AT_SPI_IOR, 0L, (long)BUFSIZ, False, (Atom) 31, &actual_type, &actual_format, &nitems, &leftover, &data); results in data == NULL , I have no idea. This is all part of http://cvs.gnome.org/viewcvs/at-spi/atk-bridge/bridge.c?r1=1.96&r2=1.97 which tried to fix and implement bug 163132.
*** Bug 377568 has been marked as a duplicate of this bug. ***
*** Bug 379882 has been marked as a duplicate of this bug. ***
All the dups that involve failure to upgrade gnome-session and which were filed against 2.17, should be split off from this bug and possibly closed as WONTFIX. 2.16-based distros should revert to at-spi-1.7.12 until at-spi is (retroactively) branched for 2.16. The patch for bug 163132 should not be applied to 2.16.
Hum, this happens with gnome-session 2.17.2
52 duplicates in 15 days (but mostly because of don scorgie's mass duping). i'm happy to see that you guys try to find out the reason for this, because this one herre really becomes a blocker.
*** Bug 380698 has been marked as a duplicate of this bug. ***
*** Bug 381051 has been marked as a duplicate of this bug. ***
Can you please confirm that this bug does NOT occur when AT support is turned off? If it DOES occur with the assistive technology gconf key set to 'false', then either: 1) GTK_MODULES is set in the user shell, to load atk-bridge anyway (which would be a config error, workaround is to unset this variable); OR 2) yelp or gecko are loading the atk-bridge without checking the gconf key, which would be a yelp or more likely a gecko error. Also please ensure that the env variable GNOME_ACCESSIBILITY is unset. Thanks!
I can confirm bug does NOT occur when AT support is disabled (see bug 382622 )
*** Bug 383318 has been marked as a duplicate of this bug. ***
*** Bug 384473 has been marked as a duplicate of this bug. ***
Bug reporters, is gnome-session running when the crash happens? If not, I guess the problem is that user has enabled the a11y gconf key, but as gnome-session is not running, at-spi-registerd is not running. So when gnome apps start, they cannot find the registerd, hence the crash.
This can happen when the gnome conf key is unset, and I didn't set GTK_MODULES and GNOME_ACCESSIBILITY. And firefox does not have such problem. Maybe yelp's problem.
Actually it is a gecho problem. Gecho will try to load atkbridge when someone calls get_accessible. Firefox team will fix this on firefox3. And we will apply a patch to atkbridge to avoid crash.
the patch of 163132 can fix this issue. I have verified it on ubuntu and it works fine. Li, please commit it.
jeff, li, is bug 380147 a duplicate?
*** Bug 380147 has been marked as a duplicate of this bug. ***
andre: yes, the same bug.
We have commit patch from http://bugzilla.gnome.org/show_bug.cgi?id=163132 , which will fix the crash.
the bug will be fixed in at-spi 1.7.14
Do you have a bug number for Gecko bug (in Firefox) ? It would be helpful for distributors. Thanks in advance.
*** Bug 380038 has been marked as a duplicate of this bug. ***
*** Bug 381511 has been marked as a duplicate of this bug. ***
*** Bug 381526 has been marked as a duplicate of this bug. ***
*** Bug 382238 has been marked as a duplicate of this bug. ***
*** Bug 383056 has been marked as a duplicate of this bug. ***
*** Bug 383604 has been marked as a duplicate of this bug. ***
*** Bug 384901 has been marked as a duplicate of this bug. ***
*** Bug 384917 has been marked as a duplicate of this bug. ***
*** Bug 385236 has been marked as a duplicate of this bug. ***
*** Bug 385539 has been marked as a duplicate of this bug. ***
*** Bug 385736 has been marked as a duplicate of this bug. ***
*** Bug 385738 has been marked as a duplicate of this bug. ***
*** Bug 386914 has been marked as a duplicate of this bug. ***
*** Bug 383388 has been marked as a duplicate of this bug. ***
this has been on the auto-reject list for a long time, but we forgot about adding a comment here. +316 rejected reports by 2007-01-23, +347 rejected reports by 2007-02-24.
*** Bug 438159 has been marked as a duplicate of this bug. ***
no new rejected reports for more than a month now (we reject reports from GNOME <=2.16). removing from the auto-reject list. 387 rejected reports.