Red Hat Bugzilla – Bug 1252172
[abrt] evolution-data-server: _gdata_service_actually_send_message(): evolution-addressbook-factory-subprocess killed by SIGABRT
Last modified: 2015-09-09 06:54:39 EDT
Version-Release number of selected component:
cmdline: /usr/libexec/evolution-addressbook-factory-subprocess --factory google --bus-name org.gnome.evolution.dataserver.Subprocess.Backend.AddressBookx2651x3 --own-path /org/gnome/evolution/dataserver/Subprocess/Backend/AddressBook/2651/3
runlevel: N 5
Thread no. 1 (9 frames)
#4 _gdata_service_actually_send_message at gdata/gdata-service.c:688
#5 _gdata_service_send_message at gdata/gdata-service.c:721
#6 _gdata_service_query at gdata/gdata-service.c:939
#7 __gdata_service_query at gdata/gdata-service.c:979
#8 query_thread at gdata/gdata-service.c:816
#9 run_in_thread at gsimpleasyncresult.c:898
#10 io_job_thread at gioscheduler.c:85
#11 g_task_thread_pool_thread at gtask.c:1283
#13 g_thread_proxy at gthread.c:764
Created attachment 1061253 [details]
Created attachment 1061254 [details]
Created attachment 1061255 [details]
Created attachment 1061256 [details]
Created attachment 1061257 [details]
Created attachment 1061258 [details]
Created attachment 1061259 [details]
Created attachment 1061260 [details]
Created attachment 1061261 [details]
Created attachment 1061262 [details]
Created attachment 1061263 [details]
Thanks for a bug report. This looks like a crash deep in libgdata. Your version of it is libgdata-0.17.2-1.fc23.x86_64. I'm not sure how this works, I do not know libgdata internals, and I neither found any similar bug report upstream. The backtrace shows that this crashed due to:
assertion failed: (message->status_code != SOUP_STATUS_NONE)
Philip, do you think it's an issue in evolution-data-server (like some sort of use-after-free), or this could happen for some other reason?
(In reply to Milan Crha from comment #12)
> Philip, do you think it's an issue in evolution-data-server (like some sort
> of use-after-free), or this could happen for some other reason?
Very hard to tell. Since it’s an assertion failure I doubt it’s caused by memory corruption (I would expect that to cause a crash). I suspect it’s a race condition somewhere in libgdata or its use of libsoup — this area of code has changed a lot in libsoup recently, which means that libgdata is now using backwards-compatibility code in libsoup.
One thing which would provide more information is to reproduce this failure under valgrind’s helgrind or drd tools, which would highlight threading problems (much the same way memcheck highlights memory problems). But I doubt it’s reproducible.
Unless that’s possible, I think this will have to be closed as INSUFFICIENT_DATA, and let’s hope that it’s fixed by https://bugzilla.gnome.org/show_bug.cgi?id=754769.