Description: (See stack trace) Steps to Reproduce: 1.Launch Writer 2.Launch Orca Actual Results: Writer sometimes crashes. Expected Results: Writer never crashes. Reproducible: Sometimes User Profile Reset: No Additional Info: Thread 1 "soffice.bin" received signal SIGSEGV, Segmentation fault. 0x00007f86d5c0d0cc in atk_object_get_index_in_parent (accessible=0x560eafda7fe0) at ../atk/atkobject.c:1065 Downloading source file /usr/src/debug/at-spi2-core-2.48.2-1.fc38.x86_64/redhat-linux-build/../atk/atkobject.c 1065 if (klass->get_index_in_parent) (gdb) bt #0 0x00007f86d5c0d0cc in atk_object_get_index_in_parent (accessible=0x560eafda7fe0) at ../atk/atkobject.c:1065 #1 0x00007f86c82e5433 in wrapper_get_index_in_parent(AtkObject*) (atk_obj=0x560eafda9210) at /usr/src/debug/libreoffice-7.5.3.2-2.fc38.x86_64/vcl/unx/gtk3/a11y/atkwrapper.cxx:524 #2 0x00007f86c79d6d2c in append_cache_item (obj=0x560eafda9210, data=0x7ffeda863e50) at ../atk-adaptor/adaptors/cache-adaptor.c:183 #3 0x00007f86d7ab65eb in g_hash_table_foreach (hash_table=0x560ea9b279d0 = {...}, func=0x7f86c79d6ef0 <append_accessible_hf>, user_data=0x7ffeda863e50) at ../glib/ghash.c:2175 #4 0x00007f86c79d3bfd in spi_cache_foreach (cache=<optimized out>, func=<optimized out>, data=<optimized out>) at ../atk-adaptor/accessible-cache.c:423 #5 0x00007f86c79dc089 in impl_GetItems (bus=<optimized out>, message=<optimized out>, user_data=<optimized out>) at ../atk-adaptor/adaptors/cache-adaptor.c:328 #6 0x00007f86c79df8b0 in handle_other (pathstr=0x560eaf5b7788 "/org/a11y/atspi/cache", member=0x560eaf5b77d8 "GetItems", iface=<optimized out>, path=0x560ea962da40, message=0x560eaf5ad0a0, bus=0x560eafdcf760) at ../droute/droute.c:558 #7 handle_message (bus=0x560eafdcf760, message=message@entry=0x560eaf5ad0a0, user_data=user_data@entry=0x560ea962da40) at ../droute/droute.c:605 #8 0x00007f86deaa22f4 in _dbus_object_tree_dispatch_and_unlock (found_object=<synthetic pointer>, message=<optimized out>, tree=0x560eafd02bf0) at ../../dbus/dbus-object-tree.c:1021 #9 dbus_connection_dispatch (connection=0x560eafdcf760) at ../../dbus/dbus-connection.c:4749 #10 dbus_connection_dispatch (connection=connection@entry=0x560eafdcf760) at ../../dbus/dbus-connection.c:4577 #11 0x00007f86c78336c1 in message_queue_dispatch (source=<optimized out>, callback=<optimized out>, user_data=<optimized out>) at ../atspi/atspi-gmain.c:89 #12 0x00007f86d7ad439c in g_main_dispatch (context=0x560ea896e7a0) at ../glib/gmain.c:3460 #13 g_main_context_dispatch (context=0x560ea896e7a0) at ../glib/gmain.c:4200 #14 0x00007f86d7b32438 in g_main_context_iterate.isra.0 (context=0x560ea896e7a0, block=1, dispatch=1, self=<optimized out>) at ../glib/gmain.c:4276 #15 0x00007f86d7ad1a23 in g_main_context_iteration (context=0x560ea896e7a0, may_block=1) at ../glib/gmain.c:4343 #16 0x00007f86c82ffcec in GtkSalData::Yield(bool, bool) (bHandleAllCurrentEvents=<optimized out>, bWait=<optimized out>, this=0x560ea883c670) at /usr/src/debug/libreoffice-7.5.3.2-2.fc38.x86_64/vcl/unx/gtk3/gtkdata.cxx:405 #17 GtkInstance::DoYield(bool, bool) (this=<optimized out>, bWait=true, bHandleAllCurrentEvents=<optimized out>) at /usr/src/debug/libreoffice-7.5.3.2-2.fc38.x86_64/vcl/unx/gtk3/gtkinst.cxx:431 #18 0x00007f86db528dfc in ImplYield(bool, bool) (i_bWait=true, i_bAllEvents=false) at /usr/src/debug/libreoffice-7.5.3.2-2.fc38.x86_64/vcl/source/app/svapp.cxx:475 #19 0x00007f86db52fccd in Application::Execute() () at /usr/src/debug/libreoffice-7.5.3.2-2.fc38.x86_64/vcl/source/app/svapp.cxx:453 #20 0x00007f86ded5e46a in desktop::Desktop::Main() (this=0x7ffeda8646b0) at /usr/src/debug/libreoffice-7.5.3.2-2.fc38.x86_64/desktop/source/app/app.cxx:1604 #21 0x00007f86db53bf9a in ImplSVMain() () at /usr/src/debug/libreoffice-7.5.3.2-2.fc38.x86_64/vcl/source/app/svmain.cxx:203 #22 0x00007f86db53c019 in SVMain() () at /usr/src/debug/libreoffice-7.5.3.2-2.fc38.x86_64/vcl/source/app/svmain.cxx:235 #23 0x00007f86ded771b8 in soffice_main() () at /usr/src/debug/libreoffice-7.5.3.2-2.fc38.x86_64/desktop/source/app/sofficemain.cxx:94 #24 0x0000560ea6fe30c7 in sal_main () at /usr/src/debug/libreoffice-7.5.3.2-2.fc38.x86_64/desktop/source/app/main.c:51 #25 main (argc=<optimized out>, argv=<optimized out>) at /usr/src/debug/libreoffice-7.5.3.2-2.fc38.x86_64/desktop/source/app/main.c:49
Joan Marie, thank you for the bug report. Just for clarification: Writer crashes immediately after launching Orca, correct? No further steps are needed? Could you please retest with a clean user profile? Thank you => NEEDINFO
Similar report is bug 154353.
*When* the crashes happen(ed?), yes it was immediate. I just removed my profile and also tried safe mode. And from a few tries, I've not yet been able to reproduce the crash. Whether or not that means it's fixed <insert giant shrug here>. Aside from the fact that the crashes only happen(ed?) sometimes, my environment has changed (updated Fedora, updated AT-SPI2, updated Orca, etc.) I'm somewhat curious as to how a profile, even a corrupted one, would cause the crash with the associated stack trace. Is there a short and simple explanation? Regardless, feel free to close this. The next time I'm actively working on Orca's support for LO (which was the case when I encountered the crashes), I can re-open or re-file is the problem persists.
Turns out this still happens.
Setting to NEW, I've at least seen this in the past.
*** Bug 154343 has been marked as a duplicate of this bug. ***
Michael: While working on the spellcheck support, LO was super crashy. Is this something you could look into? Thanks in advance!
(In reply to Joanmarie Diggs from comment #7) > Michael: While working on the spellcheck support, LO was super crashy. Is > this something you could look into? Thanks in advance! I think I've seen that in the past and the spell dialog crash is a separate issue. Will add to my list of things to look into as I find time.
(In reply to Michael Weghorn from comment #8) > I think I've seen that in the past and the spell dialog crash is a separate > issue. Will add to my list of things to look into as I find time. On further analysis, this actually seems to be the same issue, at least the backtrace is very similar to the one seen in the original bug description here. As far as I can tell, https://gerrit.libreoffice.org/c/core/+/176787 fixes this. (While it's hard to tell that for sure with an issue that's not 100% reproducible, at least I didn't see any more crashes when starting Orca or pyatspi scripts multiple times while the spell check dialog was shown, while it triggered crashes every once in a while without that fix in place.)
Michael Weghorn committed a patch related to this issue. It has been pushed to "master": https://git.libreoffice.org/core/commit/357f040d36c4bdebb59c9ada1a2ee20423ca2155 tdf#155449 gtk3 a11y: Hold reference to the original AtkObject It will be available in 25.2.0. The patch should be included in the daily builds available at https://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More information about daily builds can be found at: https://wiki.documentfoundation.org/Testing_Daily_Builds Affected users are encouraged to test the fix and report feedback.
(In reply to Michael Weghorn from comment #9) > As far as I can tell, https://gerrit.libreoffice.org/c/core/+/176787 fixes > this. Merged to master now. Backport for 24.8 pending in Gerrit: https://gerrit.libreoffice.org/c/core/+/176811 If you still see crashes with a version that has that fix, please let me know (ideally with a new backtrace).
Michael Weghorn committed a patch related to this issue. It has been pushed to "libreoffice-24-8": https://git.libreoffice.org/core/commit/957aea0cf9d21909a8488c80ebb272bd91fb30cd tdf#155449 gtk3 a11y: Hold reference to the original AtkObject It will be available in 24.8.4. The patch should be included in the daily builds available at https://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More information about daily builds can be found at: https://wiki.documentfoundation.org/Testing_Daily_Builds Affected users are encouraged to test the fix and report feedback.