| Summary: | Frequent plasma-desktop crashes [KCalCore::IncidenceBase::isReadOnly, CalendarSupport::Calendar::Private::dataChanged] | ||
|---|---|---|---|
| Product: | [Applications] kdepimlibs | Reporter: | Thomas Tanghus <thomas> |
| Component: | kcal | Assignee: | kdepim bugs <kdepim-bugs> |
| Status: | RESOLVED FIXED | ||
| Severity: | crash | CC: | andresbajotierra, asraniel, balcaen.john, bugs+kde, bugs.kde, dave.griffiths62, frederic.coiffier, johannes.schwall, kamikazow, kde, kdepim-bugs, keplicz, mfraz74+kde, osterfeld, p92, plasma-bugs, richardfeciskanin, sebsauer, smartins, Stefan.Borggraefe, thomas, winter, wstephenson |
| Priority: | NOR | ||
| Version: | 4.6 | ||
| Target Milestone: | --- | ||
| Platform: | unspecified | ||
| OS: | All | ||
| Latest Commit: | Version Fixed In: | 4.6.3 | |
| Sentry Crash Report: | |||
| Attachments: |
New crash information added by DrKonqi
New crash information added by DrKonqi New crash information added by DrKonqi New crash information added by DrKonqi New crash information added by DrKonqi New crash information added by DrKonqi New crash information added by DrKonqi New crash information added by DrKonqi New crash information added by DrKonqi New crash information added by DrKonqi |
||
|
Description
Thomas Tanghus
2010-12-30 09:20:34 UTC
Created attachment 55390 [details]
New crash information added by DrKonqi
plasma-desktop (0.4) on KDE Platform 4.5.90 (4.6 RC1) using Qt 4.7.0
- What I was doing when the application crashed:
I had just logged in. Plasma was non-responsive/sluggish sp I tabbed between session. Crash.
-- Backtrace (Reduced):
#7 KCalCore::IncidenceBase::isReadOnly (this=0xace7e70) at ../../kcalcore/incidencebase.cpp:286
#8 0x07ff8f80 in CalendarSupport::Calendar::Private::dataChanged (this=0xace7e70, topLeft=..., bottomRight=...) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:157
#9 0x07fd7394 in CalendarSupport::Calendar::Private::qt_metacall (this=0xace7e70, _c=QMetaObject::InvokeMetaMethod, _id=15, _a=0xbfc2b234) at moc_calendar_p.cpp:155
[...]
#12 0x08e48559 in QAbstractItemModel::dataChanged (this=0xace3a70, _t1=..., _t2=...) at .moc/release-shared/moc_qabstractitemmodel.cpp:148
#13 0x079523f0 in Akonadi::EntityTreeModelPrivate::dataChanged (this=0xace71f0, top=..., bottom=...) at ../../akonadi/entitytreemodel_p.cpp:1402
*** Bug 261759 has been marked as a duplicate of this bug. *** [Comment from a bug triager] Note that bug 261759 contains a reference to a null pointer on KCalCore::IncidenceBase::isReadOnly Created attachment 55513 [details]
New crash information added by DrKonqi
plasma-desktop (0.4) on KDE Platform 4.5.90 (4.6 RC1) using Qt 4.7.0
- What I was doing when the application crashed:
Same as last report. I had just logged in but this time just waited for the crash.
- Unusual behavior I noticed:
Events aren't shown when hovering over the clock/calendar applet as they where before upgrading to 4.6 RC1. My calendar is a classic KDE network resource running agains www.memotoo.com.
Is there anything I can disble or otherwise do to try to limit this to a specific component?
BTW: any typos is probably because I'm writing blidly becuase of font corruption.
-- Backtrace (Reduced):
#7 KCalCore::IncidenceBase::isReadOnly (this=0x9283390) at ../../kcalcore/incidencebase.cpp:286
#8 0x04f83f80 in CalendarSupport::Calendar::Private::dataChanged (this=0x9283390, topLeft=..., bottomRight=...) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:157
#9 0x04f62394 in CalendarSupport::Calendar::Private::qt_metacall (this=0x9283390, _c=QMetaObject::InvokeMetaMethod, _id=15, _a=0xbfc338b4) at moc_calendar_p.cpp:155
[...]
#12 0x03bd9559 in QAbstractItemModel::dataChanged (this=0x92852b0, _t1=..., _t2=...) at .moc/release-shared/moc_qabstractitemmodel.cpp:148
#13 0x07b403f0 in Akonadi::EntityTreeModelPrivate::dataChanged (this=0x9281cd8, top=..., bottom=...) at ../../akonadi/entitytreemodel_p.cpp:1402
This seems to be fixed in RC2 but I've only just logged into it. Created attachment 55725 [details] New crash information added by DrKonqi plasma-desktop (0.4) on KDE Platform 4.5.95 (4.6 RC2) using Qt 4.7.0 - What I was doing when the application crashed: Came home a few hours after logging into 4.6.0 RC2 to find this crash again. Originally reported as bug #261759 which was marked as a duplicate of this. -- Backtrace (Reduced): #6 KCalCore::IncidenceBase::isReadOnly (this=0x0) at ../../kcalcore/incidencebase.cpp:286 #7 0x00007f3738d05f23 in CalendarSupport::Calendar::Private::updateItem (this=0x1422a70, item=<value optimized out>, mode=<value optimized out>) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:222 #8 0x00007f3738d08e9b in CalendarSupport::Calendar::Private::dataChanged (this=<value optimized out>, topLeft=<value optimized out>, bottomRight=<value optimized out>) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:157 #9 0x00007f3738ce9f0c in CalendarSupport::Calendar::Private::qt_metacall (this=0x1422a70, _c=QMetaObject::InvokeMetaMethod, _id=<value optimized out>, _a=0x7fffaa2318b0) at moc_calendar_p.cpp:155 [...] #11 0x00007f375a51e2c7 in QAbstractItemModel::dataChanged (this=0x0, _t1=<value optimized out>, _t2=<value optimized out>) at .moc/release-shared/moc_qabstractitemmodel.cpp:148 I got hit by it too. It ran stable for ~12 hrs. then today when I had to reboot it crashed when I logged in again. [KCrash Handler] #7 KCalCore::IncidenceBase::isReadOnly (this=0x9531918) at ../../kcalcore/incidencebase.cpp:286 #8 0x05a01f80 in CalendarSupport::Calendar::Private::dataChanged (this=0x9531918, topLeft=..., bottomRight=...) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:157 #9 0x059e02f4 in CalendarSupport::Calendar::Private::qt_metacall (this=0x9531918, _c=QMetaObject::InvokeMetaMethod, _id=15, _a=0xbfcb58b4) at moc_calendar_p.cpp:155 #10 0x05faa8ca in QMetaObject::metacall (object=0x9531918, cl=2871242497, idx=15, argv=0xbfcb58b4) at kernel/qmetaobject.cpp:237 #11 0x05fbd6ad in QMetaObject::activate (sender=0x9533708, m=0x60d7198, local_signal_index=0, argv=0x0) at kernel/qobject.cpp:3280 #12 0x0600e559 in QAbstractItemModel::dataChanged (this=0x9533708, _t1=..., _t2=...) at .moc/release-shared/moc_qabstractitemmodel.cpp:148 #13 0xab0f33e0 in Akonadi::EntityTreeModelPrivate::dataChanged (this=0x952f498, top=..., bottom=...) at ../../akonadi/entitytreemodel_p.cpp:1402 #14 0xab0f86b2 in Akonadi::EntityTreeModelPrivate::monitoredItemChanged (this=0x952f498, item=...) at ../../akonadi/entitytreemodel_p.cpp:940 #15 0xab0eb03f in Akonadi::EntityTreeModel::qt_metacall (this=0x9533708, _c=QMetaObject::InvokeMetaMethod, _id=48, _a=0xbfcb5b74) at ./entitytreemodel.moc:156 #16 0x059e04d2 in CalendarSupport::CalendarModel::qt_metacall (this=0x9533708, _c=QMetaObject::InvokeMetaMethod, _id=48, _a=0xbfcb5b74) at moc_calendarmodel.cpp:66 #17 0x05faa8ca in QMetaObject::metacall (object=0x9533708, cl=2871242497, idx=48, argv=0xbfcb5b74) at kernel/qmetaobject.cpp:237 #18 0x05fbd6ad in QMetaObject::activate (sender=0x95310d8, m=0xab238834, local_signal_index=0, argv=0x0) at kernel/qobject.cpp:3280 #19 0xab144a69 in Akonadi::Monitor::itemChanged (this=0x95310d8, _t1=..., _t2=...) at ./monitor.moc:169 #20 0xab147dfc in Akonadi::MonitorPrivate::emitItemNotification (this=0x9531108, msg=..., item=..., collection=..., collectionDest=...) at ../../akonadi/monitor_p.cpp:432 #21 0xab148fc7 in Akonadi::MonitorPrivate::emitNotification (this=0x9531108, msg=...) at ../../akonadi/monitor_p.cpp:267 #22 0xab0a0f3c in Akonadi::ChangeRecorderPrivate::emitNotification(Akonadi::NotificationMessage const&) () from /usr/lib/libakonadi-kde.so.4 #23 0xab148c5c in Akonadi::MonitorPrivate::dataAvailable (this=0x9531108) at ../../akonadi/monitor_p.cpp:281 #24 0xab144b01 in Akonadi::Monitor::qt_metacall (this=0x95310d8, _c=QMetaObject::InvokeMetaMethod, _id=27, _a=0xbfcb5edc) at ./monitor.moc:156 #25 0xab0a0b6a in Akonadi::ChangeRecorder::qt_metacall (this=0x95310d8, _c=QMetaObject::InvokeMetaMethod, _id=27, _a=0xbfcb5edc) at ./changerecorder.moc:72 #26 0x05faa8ca in QMetaObject::metacall (object=0x95310d8, cl=2871242497, idx=27, argv=0xbfcb5edc) at kernel/qmetaobject.cpp:237 #27 0x05fbd6ad in QMetaObject::activate (sender=0x953114c, m=0xab237388, local_signal_index=0, argv=0x0) at kernel/qobject.cpp:3280 #28 0xab0e34c7 in Akonadi::EntityCacheBase::dataAvailable (this=0x953114c) at ./entitycache_p.moc:88 #29 0xab0a1238 in Akonadi::EntityCache<Akonadi::Item, Akonadi::ItemFetchJob, Akonadi::ItemFetchScope>::fetchResult (this=0x953114c, job=0xa1b7db0) at ../../akonadi/entitycache_p.h:189 #30 0xab0e3554 in Akonadi::EntityCacheBase::qt_metacall (this=0x953114c, _c=QMetaObject::InvokeMetaMethod, _id=5, _a=0xbfcb60b8) at ./entitycache_p.moc:77 #31 0x05faa8ca in QMetaObject::metacall (object=0x953114c, cl=2871242497, idx=5, argv=0xbfcb60b8) at kernel/qmetaobject.cpp:237 #32 0x05fbd6ad in QMetaObject::activate (sender=0xa1b7db0, m=0x420f548, local_signal_index=3, argv=0x0) at kernel/qobject.cpp:3280 #33 0x0408dd83 in KJob::result (this=0xa1b7db0, _t1=0xa1b7db0) at ./kjob.moc:194 #34 0x0408e0e8 in KJob::emitResult (this=0xa1b7db0) at ../../kdecore/jobs/kjob.cpp:312 #35 0xab13a6f9 in delayedEmitResult (this=0xa1b7db0, _c=QMetaObject::InvokeMetaMethod, _id=34, _a=0xa2e8878) at ../../akonadi/job.cpp:144 #36 Akonadi::Job::qt_metacall (this=0xa1b7db0, _c=QMetaObject::InvokeMetaMethod, _id=34, _a=0xa2e8878) at ./job.moc:91 #37 0xab125f8a in Akonadi::ItemFetchJob::qt_metacall (this=0xa1b7db0, _c=QMetaObject::InvokeMetaMethod, _id=34, _a=0xa2e8878) at ./itemfetchjob.moc:73 #38 0x05faa8ca in QMetaObject::metacall (object=0xa1b7db0, cl=2871242497, idx=34, argv=0xa2e8878) at kernel/qmetaobject.cpp:237 #39 0x05fb5df6 in QMetaCallEvent::placeMetaCall (this=0x9bc88b0, object=0xa1b7db0) at kernel/qobject.cpp:534 #40 0x05fb76a2 in QObject::event (this=0xa1b7db0, e=0xbfcb56b8) at kernel/qobject.cpp:1219 #41 0x01735fdc in QApplicationPrivate::notify_helper (this=0x84ac9f8, receiver=0xa1b7db0, e=0x9bc88b0) at kernel/qapplication.cpp:4396 #42 0x0173c04e in QApplication::notify (this=0x84a1658, receiver=0xa1b7db0, e=0x9bc88b0) at kernel/qapplication.cpp:3798 #43 0x00e96f8a in KApplication::notify (this=0x84a1658, receiver=0xa1b7db0, event=0x9bc88b0) at ../../kdeui/kernel/kapplication.cpp:311 #44 0x05fa4b3b in QCoreApplication::notifyInternal (this=0x84a1658, receiver=0xa1b7db0, event=0x9bc88b0) at kernel/qcoreapplication.cpp:732 #45 0x05fa7d8b in sendEvent (receiver=0x0, event_type=0, data=0x846b638) at ../../include/QtCore/../../src/corelib/kernel/qcoreapplication.h:215 #46 QCoreApplicationPrivate::sendPostedEvents (receiver=0x0, event_type=0, data=0x846b638) at kernel/qcoreapplication.cpp:1373 #47 0x05fa7f4d in QCoreApplication::sendPostedEvents (receiver=0x0, event_type=0) at kernel/qcoreapplication.cpp:1266 #48 0x05fd3a74 in sendPostedEvents (s=0x84aed20) at ../../include/QtCore/../../src/corelib/kernel/qcoreapplication.h:220 #49 postEventSourceDispatch (s=0x84aed20) at kernel/qeventdispatcher_glib.cpp:277 #50 0x021bd855 in g_main_context_dispatch () from /lib/libglib-2.0.so.0 #51 0x021c1668 in ?? () from /lib/libglib-2.0.so.0 #52 0x021c1848 in g_main_context_iteration () from /lib/libglib-2.0.so.0 #53 0x05fd3565 in QEventDispatcherGlib::processEvents (this=0x849f198, flags=...) at kernel/qeventdispatcher_glib.cpp:415 #54 0x017f7be5 in QGuiEventDispatcherGlib::processEvents (this=0x849f198, flags=...) at kernel/qguieventdispatcher_glib.cpp:204 #55 0x05fa3609 in QEventLoop::processEvents (this=0xbfcb6964, flags=) at kernel/qeventloop.cpp:149 #56 0x05fa3a8a in QEventLoop::exec (this=0xbfcb6964, flags=...) at kernel/qeventloop.cpp:201 #57 0x05fa800f in QCoreApplication::exec () at kernel/qcoreapplication.cpp:1009 #58 0x01734e07 in QApplication::exec () at kernel/qapplication.cpp:3672 #59 0x007fafb7 in kdemain (argc=1, argv=0xbfcb6b74) at ../../../../plasma/desktop/shell/main.cpp:118 #60 0x080485bb in main (argc=1, argv=0xbfcb6b74) at plasma-desktop_dummy.cpp:3 Should this be moved to another product? I looked at the code in http://websvn.kde.org/branches/KDE/4.6/kdebase/workspace/plasma/generic/dataengines/calendar/akonadi/calendar.cpp?revision=1211754&view=markup but saw nothing obvious (mainly due to my very rusty C++ knowledge). It looks like the crash is in KCalCore::IncidenceBase but I have no idea where that lives in the svn tree. [Comment from a bug triager] Indeed, the crash is related to the KDEPIM classes; and the KDEPIM developers should be already notified as their mailing list is added to the CC list of this bug report. Regards *** Bug 263651 has been marked as a duplicate of this bug. *** *** Bug 264352 has been marked as a duplicate of this bug. *** I'm getting this every time I log in now with KDE 4.6 on Kubuntu 10.10. Created attachment 56451 [details]
New crash information added by DrKonqi
plasma-desktop (0.4) on KDE Platform 4.6.00 (4.6.0) using Qt 4.7.0
- What I was doing when the application crashed:
Logging in. Also happens in KDE SC 4.6 release.
-- Backtrace (Reduced):
#7 KCalCore::IncidenceBase::isReadOnly (this=0xae3c280) at ../../kcalcore/incidencebase.cpp:286
#8 0x04320f80 in CalendarSupport::Calendar::Private::dataChanged (this=0xae3c280, topLeft=..., bottomRight=...) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:157
#9 0x042ff054 in CalendarSupport::Calendar::Private::qt_metacall (this=0xae3c280, _c=QMetaObject::InvokeMetaMethod, _id=15, _a=0xbfa14404) at moc_calendar_p.cpp:155
[...]
#12 0x05fcd559 in QAbstractItemModel::dataChanged (this=0xae3d9a8, _t1=..., _t2=...) at .moc/release-shared/moc_qabstractitemmodel.cpp:148
#13 0xabb623f0 in Akonadi::EntityTreeModelPrivate::dataChanged (this=0xae3e988, top=..., bottom=...) at ../../akonadi/entitytreemodel_p.cpp:1402
Could this bug be reassigned to kdepimlibs? [Comment from a bug triager] KDEPIM developers are already notified of this report, but OK, let's move it to kdepimlibs... Created attachment 56645 [details]
New crash information added by DrKonqi
plasma-desktop (0.4) on KDE Platform 4.6.00 (4.6.0) using Qt 4.8.0
- What I was doing when the application crashed:
Sending an email using KMail. Other that that I did nothing.
-- Backtrace (Reduced):
#7 0xa7dcadc4 in KCalCore::IncidenceBase::isReadOnly (this=0x0) at /usr/src/debug/kdepimlibs-4.6.0/kcalcore/incidencebase.cpp:286
#8 0xa80a9466 in CalendarSupport::Calendar::Private::updateItem (this=0x89a6268, item=..., mode=CalendarSupport::Calendar::Private::AssertExists)
at /usr/src/debug/kdebase-workspace-4.6.0/plasma/generic/dataengines/calendar/akonadi/calendar.cpp:227
#9 0xa80ac7ac in CalendarSupport::Calendar::Private::dataChanged (this=0x89a6268, topLeft=..., bottomRight=...)
at /usr/src/debug/kdebase-workspace-4.6.0/plasma/generic/dataengines/calendar/akonadi/calendar.cpp:157
#10 0xa808b05a in CalendarSupport::Calendar::Private::qt_metacall (this=0x89a6268, _c=QMetaObject::InvokeMetaMethod, _id=<value optimized out>, _a=0xbf80c3e4)
at /usr/src/debug/kdebase-workspace-4.6.0/build/plasma/generic/dataengines/calendar/moc_calendar_p.cpp:155
[...]
#13 0xb662d5bd in QAbstractItemModel::headerDataChanged (this=0x89a5828, _t1=150571000, _t2=-1082080128, _t3=7) at .moc/release-shared/moc_qabstractitemmodel.cpp:155
Created attachment 56662 [details]
New crash information added by DrKonqi
plasma-desktop (0.4) on KDE Platform 4.6.00 (4.6.0) using Qt 4.7.0
- What I was doing when the application crashed:
Simply logging into the desktop causes plasma to crash. After the first crash, the second restart of plasma-desktop is successful.
-- Backtrace (Reduced):
#6 KCalCore::IncidenceBase::isReadOnly (this=0x0) at ../../kcalcore/incidencebase.cpp:286
#7 0x00007f958c4eef03 in CalendarSupport::Calendar::Private::updateItem (this=0x2511c80, item=<value optimized out>, mode=<value optimized out>) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:225
#8 0x00007f958c4f1e7b in CalendarSupport::Calendar::Private::dataChanged (this=<value optimized out>, topLeft=<value optimized out>, bottomRight=<value optimized out>) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:157
#9 0x00007f958c4d2d1c in CalendarSupport::Calendar::Private::qt_metacall (this=0x2511c80, _c=QMetaObject::InvokeMetaMethod, _id=<value optimized out>, _a=0x7fff7af1a6f0) at moc_calendar_p.cpp:155
[...]
#11 0x00007f95af60d2c7 in QAbstractItemModel::dataChanged (this=0x0, _t1=<value optimized out>, _t2=<value optimized out>) at .moc/release-shared/moc_qabstractitemmodel.cpp:148
Created attachment 56699 [details]
New crash information added by DrKonqi
plasma-desktop (0.4) on KDE Platform 4.6.00 (4.6.0) using Qt 4.7.0
- What I was doing when the application crashed:
plasma crashed after I changed twice the birthdate of one of my contacts in kontact (address book on akonadi)
- Unusual behavior I noticed:
-- Backtrace (Reduced):
#6 KCalCore::IncidenceBase::isReadOnly (this=0x0) at ../../kcalcore/incidencebase.cpp:286
#7 0x00007f69536b1f03 in CalendarSupport::Calendar::Private::updateItem (this=0x198ee20, item=<value optimized out>, mode=<value optimized out>) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:225
#8 0x00007f69536b4e7b in CalendarSupport::Calendar::Private::dataChanged (this=<value optimized out>, topLeft=<value optimized out>, bottomRight=<value optimized out>) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:157
#9 0x00007f6953695d1c in CalendarSupport::Calendar::Private::qt_metacall (this=0x198ee20, _c=QMetaObject::InvokeMetaMethod, _id=<value optimized out>, _a=0x7fffe89bc890) at moc_calendar_p.cpp:155
[...]
#11 0x00007f69760562c7 in QAbstractItemModel::dataChanged (this=0x0, _t1=<value optimized out>, _t2=<value optimized out>) at .moc/release-shared/moc_qabstractitemmodel.cpp:148
Created attachment 56817 [details]
New crash information added by DrKonqi
plasma-desktop (0.4) on KDE Platform 4.6.00 (4.6.0) using Qt 4.7.0
- What I was doing when the application crashed:
Not sure, this crash appears randomly about four times per work day for me. But I am a heavy Kontact user.
-- Backtrace (Reduced):
#7 KCalCore::IncidenceBase::isReadOnly (this=0x9c26c90) at ../../kcalcore/incidencebase.cpp:286
#8 0x04f7bf80 in CalendarSupport::Calendar::Private::dataChanged (this=0x9c26c90, topLeft=..., bottomRight=...) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:157
#9 0x04f5a054 in CalendarSupport::Calendar::Private::qt_metacall (this=0x9c26c90, _c=QMetaObject::InvokeMetaMethod, _id=15, _a=0xbfae5334) at moc_calendar_p.cpp:155
[...]
#12 0x05446559 in QAbstractItemModel::dataChanged (this=0x9c3b158, _t1=..., _t2=...) at .moc/release-shared/moc_qabstractitemmodel.cpp:148
#13 0x05b8a3f0 in Akonadi::EntityTreeModelPrivate::dataChanged (this=0x9c19968, top=..., bottom=...) at ../../akonadi/entitytreemodel_p.cpp:1402
*** Bug 265087 has been marked as a duplicate of this bug. *** Please, is there any workaround for this bug while it is not fixed? Something that can be disabled or changed to avoid the code path that is generating these crashes? It is driving me crazy. I'm experiencing more than 50 plasma-desktop crashes every single day, all while using Kopete and/or Kontact. It is seriously, negatively impacting my productivity, and I can't wait for an official patch or updated release anymore. Try removing the Calendars and Clocks from your Desktop/Panel. Regards For some reason it almost works for me after I've deleted a couple of activities. The only thing I can think of which could have caused it is the Micro Blog Plasmoid which was on one of the activities. Now it only crashes on login and I guess it's because of the Digital Clock Plasmoid. Thanks Dario. Removing the clock from the panel eliminated plasma-desktop crashes, for now. Can someone CONFIRM this bug, and set it to block 4.6.1? I think it is a really major issue. I'm changing the UNCONFIRMED part to NEW but as I'm not a member of the PIM team, I won't set it as blocker unless I'm asked to do so by a team member. Has this bug been fixed? In 4.6.0 plasma used to crash on every login, no such case ever since I upgraded to 4.6.1. [Comment from a bug triager] As noted in bug 265287, there was a workaround added for this crash in Plasma @ KDE SC 4.6.1; but I think the root issue is still there (I may be wrong) Regards Created attachment 58660 [details]
New crash information added by DrKonqi
plasma-desktop (0.4) on KDE Platform 4.6.2 (4.6.2) using Qt 4.7.2
- What I was doing when the application crashed:
I simply logging to my session (on KDE SC 4.6.2)
If it can help i'm using akonadi-googledata & i have also several ical file over http pulled by akonadi.
Regards
-- Backtrace (Reduced):
#6 KCalCore::IncidenceBase::isReadOnly (this=0x0) at /usr/src/debug/kdepimlibs-4.6.2/kcalcore/incidencebase.cpp:286
#7 0x00007fd60d3ce9a5 in CalendarSupport::Calendar::Private::updateItem (this=0x2714e80, item=..., mode=<value optimized out>)
at /usr/src/debug/kdebase-workspace-4.6.2/plasma/generic/dataengines/calendar/akonadi/calendar.cpp:323
#8 0x00007fd60d3d0703 in CalendarSupport::Calendar::Private::dataChanged (this=0x2714e80, topLeft=..., bottomRight=<value optimized out>)
at /usr/src/debug/kdebase-workspace-4.6.2/plasma/generic/dataengines/calendar/akonadi/calendar.cpp:228
#9 0x00007fd60d3b23c4 in CalendarSupport::Calendar::Private::qt_metacall (this=0x2714e80, _c=QMetaObject::InvokeMetaMethod, _id=<value optimized out>, _a=0x7ffff1783530)
at /usr/src/debug/kdebase-workspace-4.6.2/build/plasma/generic/dataengines/calendar/moc_calendar_p.cpp:158
[...]
#11 0x00007fd63b8f0367 in QAbstractItemModel::dataChanged (this=<value optimized out>, _t1=<value optimized out>, _t2=<value optimized out>) at .moc/release-shared/moc_qabstractitemmodel.cpp:148
Looked like this was fixed in KDE 4.6.1, but I'm seeing it again in KDE 4.6.2 with Kubuntu 10.10. Same thing on Gentoo Git commit b2a322448e644cdf55b9990c90920ce8c0ffcc7f by Sergio Martins. Committed on 08/04/2011 at 00:43. Pushed by smartins into branch 'master'. Don't segfault in release mode. No idea why it's 0, but we're going to delete this copy of calendar.cpp soon, so it's a waste of time to debugging it. PIM applications don't crash at all regarding this, in plasma the ETM is probably set incorrectly and has items with unsupported payloads. BUG: 261593 M +4 -1 plasma/generic/dataengines/calendar/akonadi/calendar.cpp http://commits.kde.org/kde-workspace/b2a322448e644cdf55b9990c90920ce8c0ffcc7f Please test master, or the next version. I doubt it will crash. *** Bug 270318 has been marked as a duplicate of this bug. *** (In reply to comment #32) > Please test master, or the next version. I doubt it will crash. It does not crash with this patch on 4.6.2 kdebase4-workspace, but maybe it should also be backport to the 4.6 branch ? Regards, Git commit 3dbdc68fa9fb84fbcc63dbb9acffdf43a1ebfdda by Sergio Martins. Committed on 08/04/2011 at 00:43. Pushed by smartins into branch 'KDE/4.6'. Don't segfault in release mode. No idea why it's 0, but we're going to delete this copy of calendar.cpp soon, so it's a waste of time to debugging it. PIM applications don't crash at all regarding this, in plasma the ETM is probably set incorrectly and has items with unsupported payloads. BUG: 261593 (cherry picked from commit b2a322448e644cdf55b9990c90920ce8c0ffcc7f) M +4 -1 plasma/generic/dataengines/calendar/akonadi/calendar.cpp http://commits.kde.org/kde-workspace/3dbdc68fa9fb84fbcc63dbb9acffdf43a1ebfdda What causes incidence to be 0 in that code? What can be done to avoid that code path? My distro just updated to 4.6.2 and this bug is back, and I'm trying to find a solution that doesn't involve manually hand-patching and compiling KDE or removing the clock from the panel. (In reply to comment #36) > My distro just updated to 4.6.2 and this bug is back, and I'm trying to find a > solution that doesn't involve manually hand-patching and compiling KDE or > removing the clock from the panel. the fix will be in 4.6.3. Ask your distribution to backport this commit to their 4.6.2 packages Also, it seems that the bug is still not completely fixed. Calendar::Private::~Private() is dereferencing incidence without testing for null.
Calendar::Private::~Private()
{
Q_FOREACH ( const Akonadi::Item &item, m_itemMap ) {
CalendarSupport::incidence( item )->unRegisterObserver( q );
}
...
Compare with the same destructor in https://projects.kde.org/projects/kde/kdebase/kde-workspace/repository/revisions/d6ca7745810173aa7639c1a9f76dfe258c484857/entry/plasma/generic/dataengines/calendar/akonadi/calendar.cpp (older fix).
(In reply to comment #37) > the fix will be in 4.6.3. Ask your distribution to backport this commit to > their 4.6.2 packages Christophe, thanks for your suggestion. I don't want to sound ungrateful or harsh, but I asked what causes the incidence to be 0 (null) in the code. I'm well aware that I can ask the backport of this commit to my distribution, and it was not what I asked. I already went through this path before and from my experience this will take a few weeks to get processed at best (perhaps a month when 4.6.3 will be going out). And due to this very reason, they may just ignore or reject the request as unimportant, since this seems to only affect a few users. But this bug affects me now, and I'm willing to do something to fix my problem now. I want to know what is causing this situation on my configuration, because I presume there is something I can do to fix it on my side to avoid the code path that is causing the bug. It is certainly something that doesn't affect many users and has a very specific trigger, otherwise KDE developers themselves wouldn't release two versions (4.6.0 and 4.6.2) with so blatant crashes. I also verified with a new user (with a blank KDE profile) that the crash doesn't happen. So there is certainly something in my KDE configuration, something among the 56,280 files and directories residing inside my ~/.kde/ directory (that are accumulating since KDE 2.0), that triggers this for me (and a few other users who also reported this issue). I already straced plasma-desktop and the file accesses before the crash don't tell much. Instead of shooting in the dark I would like to ask the developers watching this report, who are much more familiar with KDE internals, what may be causing this situation. And take this opportunity to clean up some of the cruft accumulating in my configuration and in my ~/.kde/ . Juliano, you're using kdepim-runtime 4.4 and kdepimlibs 4.6, right ? Sergio, yes, kdepim-runtime 4.4.9 and kdepimlibs 4.6.2. KCalCore is a new library introduced with kdepimlibs 4.6. kde-workspace/../calendar.cpp uses KCalCore. kdepim-runtime 4.4 doesn't have the KCalCore serializer/deserializer, hence the invalid incidence. Possible solutions: - Make plasma calendar use KCal instead of KCalCore - Wait until kdepim-4.6 is released - Backport the serializer to kdepim-runtime 4.4 and release kdepim 4.4.11 ( !? ) - Move data serializers to kdepimlibs ( don't know if it's possible, will have to ask ) Sergio, as I explained in comment #39, if I create a new user in the same machine, with the same installation, running the very same version of KDE, but with an obviously empty ~/.kde/, a very pristine KDE profile, plasma doesn't crash. Also, very few users downloading KDE from the same repository (with stable kdepim-runtime 4.4) are reporting this crash. All other users are not experiencing the crash. That's just what I want. (In reply to comment #43) > Sergio, as I explained in comment #39, if I create a new user in the same > machine, with the same installation, running the very same version of KDE, but > with an obviously empty ~/.kde/, a very pristine KDE profile, plasma doesn't > crash. If it's pristine then it won't have any events in akonadi, so no crash. Even if you have events, it won't be easy to reproduce, by looking at the logs, several guards against invalid incidences have been made. kdepim-4.4.11 is going to be released and will fix this. It seems that I managed to fix my problem, but I don't know how valid and/or secure this procedure is. I stopped akonadi, wiped ~/.local/akonadi and ~/.config/akonadi and removed ~/.kde/share/apps/korganizer/std.ics . After that I logged out and back in and plasma didn't crash. I restored std.ics from the backup and restarted akonadi again. Relogged three more times and no more crashes. It was a shot in the dark. I have no idea if I did something wrong. Kontact forgot the address book, but I just readded std.vcf and it is back. It looks like everything is in place, but I could be missing something. Created attachment 60197 [details]
New crash information added by DrKonqi
plasma-desktop (0.4) on KDE Platform 4.6.2 (4.6.2) using Qt 4.7.0
- What I was doing when the application crashed:
changed an e-mail address in std.vcf in kontact, plasma crashed in background while kontact kept running without problem; the change seems to be saved
-- Backtrace (Reduced):
#6 KCalCore::IncidenceBase::isReadOnly (this=0x0) at ../../kcalcore/incidencebase.cpp:286
#7 0x00007fb394fa976f in CalendarSupport::Calendar::Private::updateItem (this=0x16f4040, item=<value optimized out>, mode=<value optimized out>) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:321
#8 0x00007fb394fadeab in CalendarSupport::Calendar::Private::dataChanged (this=<value optimized out>, topLeft=<value optimized out>, bottomRight=<value optimized out>) at ../../../../../plasma/generic/dataengines/calendar/akonadi/calendar.cpp:228
#9 0x00007fb394f8a714 in CalendarSupport::Calendar::Private::qt_metacall (this=0x16f4040, _c=QMetaObject::InvokeMetaMethod, _id=<value optimized out>, _a=0x7fff36a07e40) at moc_calendar_p.cpp:158
[...]
#11 0x00007fb3b91262c7 in QAbstractItemModel::dataChanged (this=0x0, _t1=<value optimized out>, _t2=<value optimized out>) at .moc/release-shared/moc_qabstractitemmodel.cpp:148
|