| Summary: | krunner and plasma crash opening System Settings | ||
|---|---|---|---|
| Product: | [Plasma] plasmashell | Reporter: | Ron Willhoite <ronwilhoite> |
| Component: | general | Assignee: | David Edmundson <kde> |
| Status: | RESOLVED UPSTREAM | ||
| Severity: | crash | CC: | abyss.7, agrar.pfurtz, ales, alex, andrey.a.pavlenko, antonis+kdebugs, arjunak234, arngrimur.bjarnason, benjamin.slabbert, bernie, bhush94, bjoern.ricks, bonifacioluiz, bwduncan, christian_bay, dabster, danofsatx, diane, dima, dmatej, dode, drdrnewman, dutchslab2000, editor.buzzfeed, edoubrayrie, exa.atto1818, gtwilliams, jamb.t, javiersmail, jc, jirded, johannes.hirte, jona.juarez, kde, kde, leh, leonard, m.wege, matthias, mrami, notmart, Ondrej.Bozek, paulo.fidalgo.pt, plasma-bugs, rick.beldin, ronwilhoite, Samuele.Kaplun, sandro, silwanyx, svalentine, transitpassenger31, wodencafe |
| Priority: | NOR | Keywords: | drkonqi |
| Version: | 5.2.2 | ||
| Target Milestone: | 1.0 | ||
| Platform: | Ubuntu | ||
| OS: | Linux | ||
| Latest Commit: | Version Fixed In: | ||
| Sentry Crash Report: | |||
| Attachments: |
New crash information added by DrKonqi
New crash information added by DrKonqi New crash information added by DrKonqi |
||
|
Description
Ron Willhoite
2015-03-26 11:46:50 UTC
Created attachment 91765 [details]
New crash information added by DrKonqi
plasmashell (5.2.2) using Qt 5.4.1
- What I was doing when the application crashed:
Kmenu > Konsole
Krunner crashed, then Plasma.
Plasma restarted itself, and Konsole eventually opened.
-- Backtrace (Reduced):
#6 0x00007f291e1fd267 in __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:55
#7 0x00007f291e1feeca in __GI_abort () at abort.c:89
[...]
#12 0x00007f291eebc96a in QPlatformScreen::physicalSize() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5
#13 0x00007f291eef67b2 in QScreen::physicalSize() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5
#14 0x00007f291eef6869 in QScreen::physicalDotsPerInch() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5
From 344336: FWIW, following up on Vishesh's comment about multiple monitors, I left my laptop lid open for a couple of days and didn't experience any crashes. This morning, before opening the laptop lid, I clicked on the update notifier icon in the panel and the got the krunner crash followed by the plasma crash. The crash reporter said the backtrace indicated a duplicate of 345544. I'll post this comment there as well. *** Bug 345768 has been marked as a duplicate of this bug. *** *** Bug 345791 has been marked as a duplicate of this bug. *** *** Bug 345801 has been marked as a duplicate of this bug. *** *** Bug 345845 has been marked as a duplicate of this bug. *** *** Bug 345955 has been marked as a duplicate of this bug. *** is same as bug 343216 and probably upstream. d->platformScreen->physicalSize(); from QScreen::physicalSize() gets probably called when platformScreen is either not valid anymore or not valid yet, causing the pure vitual exception due to invoking a virtual from a dangling pointer Created attachment 92084 [details]
New crash information added by DrKonqi
plasmashell (5.2.2) using Qt 5.4.1
- What I was doing when the application crashed:
This crash is reproducible by disconnecting my laptop from its docking station which has a VGA monitor attached. Plasma restarts and all is well.
-- Backtrace (Reduced):
#6 0x000000323a035267 in __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:55
#7 0x000000323a036eca in __GI_abort () at abort.c:89
[...]
#12 0x0000003b1c8d896a in QPlatformScreen::physicalSize() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5
#13 0x0000003b1c9127b2 in QScreen::physicalSize() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5
#14 0x0000003b1c912869 in QScreen::physicalDotsPerInch() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5
This is the reproducible krunner backtrace, only for the crashing thread: Thread 1 (Thread 0x7f7bd2d307c0 (LWP 2385)): [KCrash Handler] #6 0x000000323a035267 in __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:55 #7 0x000000323a036eca in __GI_abort () at abort.c:89 #8 0x0000003b1b66106d in __gnu_cxx::__verbose_terminate_handler() () from /usr/lib/x86_64-linux-gnu/libstdc++.so.6 #9 0x0000003b1b65eee6 in ?? () from /usr/lib/x86_64-linux-gnu/libstdc++.so.6 #10 0x0000003b1b65ef31 in std::terminate() () from /usr/lib/x86_64-linux-gnu/libstdc++.so.6 #11 0x0000003b1b65fa7f in __cxa_pure_virtual () from /usr/lib/x86_64-linux-gnu/libstdc++.so.6 #12 0x0000003b1c8d896a in QPlatformScreen::physicalSize() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5 #13 0x0000003b1c9127b2 in QScreen::physicalSize() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5 #14 0x0000003b1c912869 in QScreen::physicalDotsPerInch() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5 #15 0x0000003b28086c49 in ?? () from /usr/lib/x86_64-linux-gnu/libQt5Quick.so.5 #16 0x0000003b1c1129c9 in QMetaObject::activate(QObject*, int, int, void**) () from /usr/lib/x86_64-linux-gnu/libQt5Core.so.5 #17 0x0000003b1c8efd4f in QWindow::screenChanged(QScreen*) () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5 #18 0x0000003b1c8f05a5 in QWindowPrivate::emitScreenChangedRecursion(QScreen*) () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5 #19 0x0000003b1c8f3cc3 in QWindowPrivate::setTopLevelScreen(QScreen*, bool) () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5 #20 0x00000000004077c0 in operator() (screen=0x19a4c20, __closure=<synthetic pointer>) at ../../krunner/view.cpp:97 #21 call (arg=<optimized out>, f=...) at /usr/include/x86_64-linux-gnu/qt5/QtCore/qobjectdefs_impl.h:494 #22 call<QtPrivate::List<QScreen*>, void> (arg=<optimized out>, f=...) at /usr/include/x86_64-linux-gnu/qt5/QtCore/qobjectdefs_impl.h:551 #23 QtPrivate::QFunctorSlotObject<View::View(QWindow*)::<lambda(QScreen*)>, 1, QtPrivate::List<QScreen*>, void>::impl(int, QtPrivate::QSlotObjectBase *, QObject *, void **, bool *) (which=<optimized out>, this_=<optimized out>, r=<optimized out>, a=<optimized out>, ret=<optimized out>) at /usr/include/x86_64-linux-gnu/qt5/QtCore/qobject_impl.h:192 #24 0x0000003b1c11235a in QMetaObject::activate(QObject*, int, int, void**) () from /usr/lib/x86_64-linux-gnu/libQt5Core.so.5 #25 0x0000003b1c8e1cb2 in QGuiApplication::screenRemoved(QScreen*) () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5 #26 0x0000003b1c912659 in QScreen::~QScreen() () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5 #27 0x0000003b1c912679 in QScreen::~QScreen() () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5 #28 0x0000003b1c8d8f2a in QPlatformScreen::~QPlatformScreen() () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5 #29 0x00007f7bd29e3769 in ?? () from /usr/lib/x86_64-linux-gnu/qt5/plugins/platforms/libqxcb.so #30 0x00007f7bd29d65ab in ?? () from /usr/lib/x86_64-linux-gnu/qt5/plugins/platforms/libqxcb.so #31 0x00007f7bd29d727d in ?? () from /usr/lib/x86_64-linux-gnu/qt5/plugins/platforms/libqxcb.so #32 0x00007f7bd29d84bb in ?? () from /usr/lib/x86_64-linux-gnu/qt5/plugins/platforms/libqxcb.so #33 0x0000003b1c11373a in QObject::event(QEvent*) () from /usr/lib/x86_64-linux-gnu/libQt5Core.so.5 #34 0x0000003b1cf5bb2c in QApplicationPrivate::notify_helper(QObject*, QEvent*) () from /usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5 #35 0x0000003b1cf61000 in QApplication::notify(QObject*, QEvent*) () from /usr/lib/x86_64-linux-gnu/libQt5Widgets.so.5 #36 0x0000003b1c0e1c2b in QCoreApplication::notifyInternal(QObject*, QEvent*) () from /usr/lib/x86_64-linux-gnu/libQt5Core.so.5 #37 0x0000003b1c0e3c9b in QCoreApplicationPrivate::sendPostedEvents(QObject*, int, QThreadData*) () from /usr/lib/x86_64-linux-gnu/libQt5Core.so.5 #38 0x0000003b1c13a843 in ?? () from /usr/lib/x86_64-linux-gnu/libQt5Core.so.5 #39 0x000000323b849c3d in g_main_context_dispatch () from /lib/x86_64-linux-gnu/libglib-2.0.so.0 #40 0x000000323b849f20 in ?? () from /lib/x86_64-linux-gnu/libglib-2.0.so.0 #41 0x000000323b849fcc in g_main_context_iteration () from /lib/x86_64-linux-gnu/libglib-2.0.so.0 #42 0x0000003b1c13ac57 in QEventDispatcherGlib::processEvents(QFlags<QEventLoop::ProcessEventsFlag>) () from /usr/lib/x86_64-linux-gnu/libQt5Core.so.5 #43 0x0000003b1c0df3e2 in QEventLoop::exec(QFlags<QEventLoop::ProcessEventsFlag>) () from /usr/lib/x86_64-linux-gnu/libQt5Core.so.5 #44 0x0000003b1c0e702c in QCoreApplication::exec() () from /usr/lib/x86_64-linux-gnu/libQt5Core.so.5 #45 0x0000000000407447 in main (argc=1, argv=<optimized out>) at ../../krunner/main.cpp:71 Created attachment 92165 [details]
New crash information added by DrKonqi
plasmashell (5.2.2) using Qt 5.4.1
- What I was doing when the application crashed:
Disconnected external screen VGA1. This causes the primary display to move from external screen VGA1 to the laptop screen LVDS1.
-- Backtrace (Reduced):
#6 0x00007f30ed3e2267 in __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:55
#7 0x00007f30ed3e3eca in __GI_abort () at abort.c:89
[...]
#12 0x00007f30ee0a196a in QPlatformScreen::physicalSize() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5
#13 0x00007f30ee0db7b2 in QScreen::physicalSize() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5
#14 0x00007f30ee0db869 in QScreen::physicalDotsPerInch() const () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5
*** Bug 346479 has been marked as a duplicate of this bug. *** Seems it's all in Qt code; I haven't seen this ever, so maybe it's fixed in what will be 5.4.2? All of these are missing the one symbol that matters: #15 0x0000003b28086c49 in ?? () from /usr/lib/x86_64-linux-gnu/libQt5Quick.so.5 it's where it's actually doing the wrong thing on reaction to a screen change, accessing some invalid object. Could someone who can reproduce this install -dbg for qt-declarative? Edit, it's going to be: QQuickScreenAttached::screenChanged(QScreen *screen) Which mostly doesn't do anything wrong. It checks if a value has changed from the new screen compared to the old screen. if (!oldScreen || screen->physicalDotsPerInch() != oldScreen->physicalDotsPerInch()) However this is called from the destructor of the old screen: #27 0x0000003b1c912679 in QScreen::~QScreen() () from /usr/lib/x86_64-linux-gnu/libQt5Gui.so.5 calling methods on an object before we delete it is normally fine, we're not actually deleted till this method ends. but in this case we're calling a virtual method from a destructor. This is sometimes safe, sometimes not safe depending on black magic: http://stackoverflow.com/questions/12092933/calling-virtual-function-from-destructor Hi David, Thanks very much for looking into this. I've installed the qtdeclarative5-dbg package, which pulled in a few others. I can reproduce this by docking and undocking my laptop, which I will do first thing tomorrow when I get to work. Cheers, Bruce bwduncan@gmail.com, there's no need now. I know exactly what's wrong; I'm not super sure I can get a workaround into 5.4. I'll try. I'll open a bug upstream tomomorrow and put a link to there. Please follow updates here https://bugreports.qt.io/browse/QTBUG-45753 It's fixed in Qt 5.5, but it's probably a long time till distros get that. I've opened a review request working round the problem for Qt5.4.2 which is in a few days. https://codereview.qt-project.org/#/c/111100/ *** Bug 346513 has been marked as a duplicate of this bug. *** *** Bug 346514 has been marked as a duplicate of this bug. *** *** Bug 346512 has been marked as a duplicate of this bug. *** Update for people who watch this stuff. Qt fixed it properly for Qt 5.5 which won't be out for a while. I submitted a workaround for Qt 5.4.2 which is now submitted, and most distros (Fedora, OpenSUSE) are going to apply that patch early. Cool, thanks David! Thanks David, that's fantastic! Do you happen to know if Ubuntu is one of the distros who are going to patch this early or should I submit a bug with them too? Ta, Bruce *** Bug 346886 has been marked as a duplicate of this bug. *** *** Bug 346996 has been marked as a duplicate of this bug. *** *** Bug 346751 has been marked as a duplicate of this bug. *** *** Bug 347072 has been marked as a duplicate of this bug. *** Fixed for Fedora with qt5-qtdeclarative-5.4.1-4.fc22 (https://admin.fedoraproject.org/updates/qt5-qtdeclarative). I can't reproduce the crash any more. *** Bug 347106 has been marked as a duplicate of this bug. *** *** Bug 347156 has been marked as a duplicate of this bug. *** *** Bug 347196 has been marked as a duplicate of this bug. *** Will be addressed in an SRU to Kubuntu 15.04: https://bugs.launchpad.net/ubuntu/+source/qtdeclarative-opensource-src/+bug/1450137 *** Bug 347494 has been marked as a duplicate of this bug. *** *** Bug 347133 has been marked as a duplicate of this bug. *** *** Bug 347696 has been marked as a duplicate of this bug. *** *** Bug 347908 has been marked as a duplicate of this bug. *** *** Bug 347932 has been marked as a duplicate of this bug. *** *** Bug 347958 has been marked as a duplicate of this bug. *** *** Bug 348111 has been marked as a duplicate of this bug. *** *** Bug 348359 has been marked as a duplicate of this bug. *** *** Bug 348520 has been marked as a duplicate of this bug. *** *** Bug 348576 has been marked as a duplicate of this bug. *** (In reply to David Edmundson from comment #22) > Update for people who watch this stuff. > > Qt fixed it properly for Qt 5.5 which won't be out for a while. > > I submitted a workaround for Qt 5.4.2 which is now submitted, and most > distros (Fedora, OpenSUSE) are going to apply that patch early. Okay, so Qt 5.4.2 has been released, but the patch was not included (as checked against the tar balls)? How could that have happened? Ok so I compiled Qt 5.4.1 with the patch and also Arch Linux just provided 5.4.2 including this patch. While the overall stability is indeed improved - I still see crashes when changing monitor settings and most prominently when undocking my laptop. Are there any other outstanding monitor related bugfixes in 5.4.2? Also see https://bugs.archlinux.org/task/43986 where another user has reported remaining problems with a patch Qt 5.4.2. *** Bug 348804 has been marked as a duplicate of this bug. *** *** Bug 348947 has been marked as a duplicate of this bug. *** *** Bug 348961 has been marked as a duplicate of this bug. *** *** Bug 348936 has been marked as a duplicate of this bug. *** *** Bug 348995 has been marked as a duplicate of this bug. *** *** Bug 349049 has been marked as a duplicate of this bug. *** Update for people watching; it got merged into the 5.4 branch but for some reason it's not in 5.4.2. I don't know why :( *** Bug 348925 has been marked as a duplicate of this bug. *** *** Bug 348956 has been marked as a duplicate of this bug. *** @David Edmundson: Are you sure that the Qt crashes are thoroughly resolved by your Qt 5.4.2 patch? I'm on 5.4.2 with applied patch and still get crashes when connecting and removing monitors. I have seen krunner, systemsettings5 and yakuake crash despite the patch. Of all the duplicates closed recently, might not be many of these already on 5.4.2 (+patch) with remaining issues? Is there even any distro out there shipping Qt 5.4.2 without the additional patch? For example the last marked duplicate bug 348956 is on Qt 5.4.2 and Fedora 22 which indeed already includes the patch: > http://pkgs.fedoraproject.org/cgit/qt5-qtdeclarative.git/commit/?h=f22&id=4741dae1b2aedabbc3e87cf8e3f3088386270860 *** Bug 349092 has been marked as a duplicate of this bug. *** *** Bug 349364 has been marked as a duplicate of this bug. *** I can confirm that bug 348961 (which is a crash reporter dupe of this one) is fixed for me with QT as in qt5-qt*-5.4.2-1.fc21 on Fedora 21. Thanks! *** Bug 350282 has been marked as a duplicate of this bug. *** *** Bug 350320 has been marked as a duplicate of this bug. *** *** Bug 350503 has been marked as a duplicate of this bug. *** Could I just ask how a regular kubuntu user can get this issue fixed? I have clean kubuntu 15.04 install with all regular updates, but get these crashes nearly every time I disconnect the external screen. Sometimes there is no taskbar at all so I need to restart, and even worse sometimes KDE crashes on startup returning to the login screen and only connecting the external screen solves this, a restart won't help. Currently I cannot use my laptop without external screen because of this. I don't know if this is really related to this bug, please forgive me if not, I'll create a new bug report then. Here the end of .xsession-errors, maybe helpful: kdeinit5: preparing to launch 'libkdeinit5_ksmserver' kscreen: Launcher finished with exit code 0 , status 1 kf5.kiconthemes: "Theme tree: (Breeze)" Communication problem with "kaccess" , it probably crashed. Error message was: "org.freedesktop.DBus.Error.NoReply" : " "Message did not receive a reply (timeout by message bus)" " kdeinit5: PID 1930 terminated. kdeinit5: PID 1936 terminated. startkde: Shutting down... kdeinit5: terminate KDE. klauncher: Exiting on signal 1 QDBusAbstractAdaptor: Cannot relay signal KDEDModule::moduleDeleted(KDEDModule*): Pointers are not supported: KDEDModule* QDBusAbstractAdaptor: Cannot relay signal KDEDModule::moduleDeleted(KDEDModule*): Pointers are not supported: KDEDModule* QDBusAbstractAdaptor: Cannot relay signal KDEDModule::moduleDeleted(KDEDModule*): Pointers are not supported: KDEDModule* startkde: Running shutdown scripts... startkde: Done. KCrash: Attempting to start /usr/bin/kdeinit5 from kdeinit Warning: connect() failed: : Datei oder Verzeichnis nicht gefunden KCrash: Attempting to start /usr/bin/kdeinit5 directly KCrash: crashing... crashRecursionCounter = 2 KCrash: Application Name = kdeinit5 path = /usr/bin pid = 1876 KCrash: Arguments: /usr/bin/kdeinit5 kdeinit5: preparing to launch 'libkdeinit5_klauncher' kdeinit5: Launched KLauncher, pid = 1999, result = 0 Invalid MIT-MAGIC-COOKIE-1 keyQXcbConnection: Could not connect to display :0 kdeinit5: Communication error with launcher. Exiting! *** Bug 343216 has been marked as a duplicate of this bug. *** (In reply to Torsten Römer from comment #62) > Could I just ask how a regular kubuntu user can get this issue fixed? > > I have clean kubuntu 15.04 install with all regular updates, but get these > crashes nearly every time I disconnect the external screen. > Sometimes there is no taskbar at all so I need to restart, and even worse > sometimes KDE crashes on startup returning to the login screen and only > connecting the external screen solves this, a restart won't help. Currently > I cannot use my laptop without external screen because of this. The status as far as I understand is, that 5.4.2 + a patch which was expected to be in there will solve most of the issues (I still had remaining issues). However qt 5.5 has restructured most of the code and the 5.4.2 solution does not apply anymore. On Arch Linux Qt 5.5 is already shipped. While the situation is better than in the 5.4.1 days, there are still issues: https://bugreports.qt.io/browse/QTBUG-44158 With the two patches applied in there the situation becomes bearable. Small issues remain but I don't think they have a bug tracker entry yet. To come back to your question: You'll have to get at least a patched Qt 5.4.2 version. Since vivid is still at 5.4.1, i doubt it will be easy. https://bugs.launchpad.net/ubuntu/+source/qtdeclarative-opensource-src/+bug/1450137 (In reply to Matthias Blaicher from comment #64) > To come back to your question: You'll have to get at least a patched Qt > 5.4.2 version. Since vivid is still at 5.4.1, i doubt it will be easy. > > https://bugs.launchpad.net/ubuntu/+source/qtdeclarative-opensource-src/+bug/ > 1450137 Thanks for the detailed information. I guess we'll have to wait for 15.10 then. The issue with KDE crashing during startup at certain display configurations was solved for me by deleting the screen configurations in ~/.local/share/kscreen. Since then there are also no crashes when I disconnect the external screen. The problem with the taskbar however remains, sometimes when the external screen is disconnected it is gone, and sometimes the "Primary display" setting and the actual position of the taskbar are inverted, i.e. I have to set the primary display to "Laptop" in order to have the taskbar on the external screen and vice-versa. *** Bug 351415 has been marked as a duplicate of this bug. *** (In reply to Matthias Blaicher from comment #64) > (In reply to Torsten Römer from comment #62) > > Could I just ask how a regular kubuntu user can get this issue fixed? > > > > I have clean kubuntu 15.04 install with all regular updates, but get these > > crashes nearly every time I disconnect the external screen. > > Sometimes there is no taskbar at all so I need to restart, and even worse > > sometimes KDE crashes on startup returning to the login screen and only > > connecting the external screen solves this, a restart won't help. Currently > > I cannot use my laptop without external screen because of this. > > The status as far as I understand is, that 5.4.2 + a patch which was > expected to be in there will solve most of the issues (I still had remaining > issues). > > However qt 5.5 has restructured most of the code and the 5.4.2 solution does > not apply anymore. On Arch Linux Qt 5.5 is already shipped. While the > situation is better than in the 5.4.1 days, there are still issues: > https://bugreports.qt.io/browse/QTBUG-44158 On Gentoo with Qt 5.5 I get this crash everytime when connecting/disconnecting an external monitor to my laptop. So I can't confirm that it's resolved upstream. As I said, even in Qt 5.5 there are still issues remaining. In particular you need to apply the patches merged after the 5.5.0 release: https://codereview.qt-project.org/#/c/114796/ https://codereview.qt-project.org/#/c/120700/ Even that will still not be bug free and there are still crashes coming up. See also the discussion in https://bugs.archlinux.org/task/43986 Best, Matthias *** Bug 354471 has been marked as a duplicate of this bug. *** *** Bug 354968 has been marked as a duplicate of this bug. *** *** Bug 355567 has been marked as a duplicate of this bug. *** *** Bug 355614 has been marked as a duplicate of this bug. *** *** Bug 360867 has been marked as a duplicate of this bug. *** *** Bug 361659 has been marked as a duplicate of this bug. *** Thanks for the detailed information. I guess we'll have to wait for 15.10 then. The issue with KDE crashing during startup at certain display configurations was solved for me by deleting the screen configurations in ~/.local/share/kscreen. Since then there are also no crashes when I disconnect the external screen. Thanks http://qwikfix.co.uk/sky-customer-services/ *** Bug 368431 has been marked as a duplicate of this bug. *** |