| Summary: | Editing new panel | ||
|---|---|---|---|
| Product: | [Plasma] plasmashell | Reporter: | trevor.d.ball |
| Component: | Task Manager and Icons-Only Task Manager | Assignee: | Eike Hein <hein> |
| Status: | RESOLVED DUPLICATE | ||
| Severity: | crash | CC: | bhush94, joe.christy, kde, plasma-bugs |
| Priority: | NOR | Keywords: | drkonqi |
| Version: | 5.3.1 | ||
| Target Milestone: | 1.0 | ||
| Platform: | Fedora RPMs | ||
| OS: | Linux | ||
| Latest Commit: | Version Fixed In: | ||
| Sentry Crash Report: | |||
| Attachments: |
fbscreen5.png
New crash information added by DrKonqi |
||
|
Description
trevor.d.ball
2015-06-12 13:33:18 UTC
Trevor, do you have the window list applet installed? Created attachment 93133 [details] fbscreen5.png Just the window list plasma widget. Icons only. I'll attach a screen shot. Also unrelated, super cool that you guys responded that quick. Just wanted to help file a bug! (that may or may not be my doing) On Fri, Jun 12, 2015 at 8:58 AM, David Edmundson <kde@davidedmundson.co.uk> wrote: > https://bugs.kde.org/show_bug.cgi?id=349062 > > David Edmundson <kde@davidedmundson.co.uk> changed: > > What |Removed |Added > > ---------------------------------------------------------------------------- > Resolution|--- |WAITINGFORINFO > Assignee|kde@davidedmundson.co.uk |hein@kde.org > Status|UNCONFIRMED |NEEDSINFO > Component|general |Task Manager > > --- Comment #1 from David Edmundson <kde@davidedmundson.co.uk> --- > Trevor, do you have the window list applet installed? > > -- > You are receiving this mail because: > You reported the bug. > While upgrading ~380 packages using dnf from the commandline in a konsole logged in as root (I was logged in to plasma as a normal user), I launched an emacs window, as root, from another konsole, to make a note. When I closed emacs using ctrl-X ctrl-C, plasmashell crashed. Contrary to what the Fedora abrt claimed, it did not attach my backtrace to this bug, nor did it report the bug as 349016 as it also claimed. Contrary to what kde-bugzilla believes, I do not believe that it is a a duplicate of either this bug or 349016. In my experience plasmashell crashes randomly several times a week, on both my F22 machines. Most of the time on my desktop/file-server/syslog-host, it crashes while the machine is dormant, but running so as to allow cron jobs to run and syslog to catch reports from the rest of my network, so I rarely see it at the time it crashes. To increase frustration, in those cases abrt doesn't catch the crash, so all I see is the lack of plasma and I only can tell that plasmashell has crashed with pgrep. At that point I restart plasmashell from the commandline with "plasmashell start --shut-up". Needless to say there's no backtrace to capture in those cases. On my laptop, it seems to crash, as I said, at random, mostly when I'm not actively doing anything memorable. Unlike this crash, I haven't filed a bug report since I have nothing to add to the many plasmashell crashing bug reports, both open and marked as duplicates. IMNSHO, plasmashell needs some extended, rigorous testing with a fuzzer, as well as some some user-accessible debugging/logging tools. Created attachment 93834 [details]
New crash information added by DrKonqi
plasmashell (5.3.2) using Qt 5.4.2
- What I was doing when the application crashed:
While upgrading ~380 packages using dnf from the commandline in a konsole logged in as root (I was logged in to plasma as a normal user), I launched an emacs window, as root, from another konsole, to make a note. When I closed emacs using ctrl-X ctrl-C, plasmashell crashed.
Contrary to what the Fedora abrt claimed, it did not attach my backtrace to this bug, nor did it report the bug as 349016 as it also claimed. Contrary to what kde-bugzilla believes, I do not believe that it is a a duplicate of either this bug or 349016.
In my experience plasmashell crashes randomly several times a week, on both my F22 machines. Most of the time on my desktop/file-server/syslog-host, it crashes while the machine is dormant, but running so as to allow cron jobs to run and syslog to catch reports from the rest of my network, so I rarely see it at the time it crashes. To increase frustration, in those cases abrt doesn't catch the crash, so all I see is the lack of plasma and I only can tell that plasmashell has crashed with pgrep. At that point I restart plasmashell from the commandline with "plasmashell start --shut-up". Needless to say there's no backtrace to capture in those cases.
On my laptop, it seems to crash, as I said, at random, mostly when I'm not actively doing anything memorable. Unlike this crash, I haven't filed a bug report since I have nothing to add to the many plasmashell crashing bug reports, both open and marked as duplicates.
IMNSHO, plasmashell needs some extended, rigorous testing with a fuzzer, as well as some some user-accessible debugging/logging tools.
-- Backtrace (Reduced):
#5 0x0000003cb38a6488 in VDMModelDelegateDataType::notify(QQmlAdaptorModel const&, QList<QQmlDelegateModelItem*> const&, int, int, QVector<int> const&) const () at /lib64/libQt5Qml.so.5
#6 0x0000003cb3b085ea in QQmlDelegateModel::_q_itemsChanged(int, int, QVector<int> const&) () at /lib64/libQt5Qml.so.5
#7 0x0000003cb3b2ebbc in QQmlDelegateModel::qt_static_metacall(QObject*, QMetaObject::Call, int, void**) () at /lib64/libQt5Qml.so.5
#8 0x0000003cb3b2ef95 in QQmlDelegateModel::qt_metacall(QMetaObject::Call, int, void**) () at /lib64/libQt5Qml.so.5
[...]
#10 0x0000003c9a126ffc in QAbstractItemModel::dataChanged(QModelIndex const&, QModelIndex const&, QVector<int> const&) () at /lib64/libQt5Core.so.5
If you don't have a trace, is there a particular reason why you chose this bug to comment on? >IMNSHO, plasmashell needs some extended, rigorous testing with a fuzzer, as well as some some user-accessible debugging/logging tools. It needs users to run Qt 5.5 more than anything else. *** This bug has been marked as a duplicate of bug 342763 *** Edit, you clearly do have a trace. I wrote that between you typing your second comment. You're basically seeing a form of this: https://bugreports.qt.io/browse/QTBUG-43827 Aha! Now I see that: a) I mistook DrKonqui for abrt and wasn't patient enough for it to attach my report to this bug. b) Scanning two backtraces, it looks to like this is a Qt5 bug. Know nothing much about the code for either beyond their comparative stabilities, it's too bad the more stable kde4 frameworks couldn't have been merged with the flaky Qt5 analogues somehow. Sigh. Don't be misled by the nvidia kernel module thread. My desktop, where plasmashell crashes considerably more frequently, has Intel integrated graphics. OK, how do I get Qt5.5? Do I need to prod the Fedora 22 maintainers? Rawhide has Qt5.5, but, alas, I'm too busy to set-up a Rawhide VM and properly report bugs. Sigh. |