Summary: | Ampache Crash | ||
---|---|---|---|
Product: | [Plasma] Oxygen | Reporter: | Malar Kannan <malarkannan.p> |
Component: | general | Assignee: | Hugo Pereira Da Costa <hugo.pereira.da.costa> |
Status: | RESOLVED WAITINGFORINFO | ||
Severity: | crash | CC: | hugo.pereira.da.costa, mchugh19 |
Priority: | NOR | ||
Version: | 4.0 | ||
Target Milestone: | --- | ||
Platform: | Fedora RPMs | ||
OS: | Linux | ||
Latest Commit: | Version Fixed In: | ||
Sentry Crash Report: | |||
Attachments: |
New crash information added by DrKonqi
New crash information added by DrKonqi |
Description
Malar Kannan
2011-02-11 16:53:11 UTC
The problem is in Oxygen, not in Amarok, reassigning. *** Bug 266853 has been marked as a duplicate of this bug. *** Confirmed by duplicates. Both crash don't actually occur at the exact same line in the code. (one at line 542 the other at line 551) Both try to either cast a widget, or access a method of the cast widget. This probably indicates that a dangling pointer to an invalid widget is passed to the incriminated method (::CanDrag). I cant really tell more unless I suceed at reproducing the bug. How do I 'enable' ampache in amarok at the first place (I don't seem to have such a thing here) ? Hugo (In reply to comment #4) > How do I 'enable' ampache in amarok at the first place (I don't seem to have > such a thing here) ? In the ampache settings for internet plugins, I have an option to add an amarok server. Standard distro packages come with it enabled, otherwise it's a compile time option. I can give you a login to my server if you want to send me a private email. Thanks! ok. I now have ampache running, and a brand new large playlist, but still can't crash the application. Questions: - does the crash happens everytime ? - is there a chance you can post a screenshot of amarok's state *before* the crash (I'm not sure I do the right things to trigger it). Thx, Hugo (In reply to comment #6) > Questions: > - does the crash happens everytime ? > - is there a chance you can post a screenshot of amarok's state *before* the > crash Often when opening amarok to play from ampache, I'll open the app, click internet, click ampache and wait a little bit, but the ampache listing never populates. In these cases I try entering some text into the search filter, wait a second, then delete it. Sometimes that seems to help, most of the time not. Then I'll go into settings and disable the ampache plugin, apply, then enable it again, and click ok. I might also go to edit and hit ok. This bug seemed to be triggered by disabling/enabling the ampache plugin. I don't know if it happens every time, but I'd bet pretty reliably. When I get to work I'll see if I can repeat the crash. Created attachment 57466 [details]
New crash information added by DrKonqi
amarok (2.4.0) on KDE Platform 4.6.00 (4.6.0) using Qt 4.7.0
- What I was doing when the application crashed:
This is a crash from running amarok and clicking enable/disable and apply on the ampache resource.
-- Backtrace (Reduced):
#6 Oxygen::WindowManager::canDrag (this=<value optimized out>, widget=0x36fd880, child=<value optimized out>, position=...) at ../../../kstyles/oxygen/oxygenwindowmanager.cpp:551
#7 0x00007f82b2fc37cf in Oxygen::WindowManager::mousePressEvent (this=0x2212470, object=0x36fd880, event=<value optimized out>) at ../../../kstyles/oxygen/oxygenwindowmanager.cpp:241
#8 0x00007f82c63d7507 in QCoreApplicationPrivate::sendThroughObjectEventFilters (this=<value optimized out>, receiver=0x36fd880, event=0x7fff4402ff20) at kernel/qcoreapplication.cpp:847
#9 0x00007f82c6db9fac in QApplicationPrivate::notify_helper (this=0x21bb7c0, receiver=0x36fd880, e=0x7fff4402ff20) at kernel/qapplication.cpp:4392
#10 0x00007f82c6dc030e in QApplication::notify (this=0x7fff44030be0, receiver=0x36fd880, e=0x7fff4402ff20) at kernel/qapplication.cpp:3959
(In reply to comment #8) > Created an attachment (id=57466) [details] Hey all. Want to watch a video of me fumbling to make amarok crash? Well, you're in luck! http://www.youtube.com/watch?v=IQJwOJJXf0M Made a youtube account just for this bug report. Yea ha @Christian Nice video ! Well, I must say that it does not seem *that* easy to make it crash ;) (if only you did not have this issue that your ampache list would not update). My belief is that indeed the crash is triggered by a widget being accessed while being deleted, due to some race condition, because basically all your "playing around" (no offense meant), results in creating destroying quite some widgets, until you hit the conflict. Don't get me wrong, I'm not saying that there is no bug. I'm just saying it will be hard to fix. Will keep you posted, and thanks again for taking the effort of recording this video. Created attachment 57472 [details]
New crash information added by DrKonqi
amarok (2.4.0) on KDE Platform 4.6.00 (4.6.0) using Qt 4.7.0
- What I was doing when the application crashed:
Thanks much Hugo. I understand this might be a tricky one to track down. The first time I had the problem it was from going into the settings once, and it immediatly crashed. As with all things technological, it suddently becomes more difficult once you have someone to show :)
I think this crash might be related. In trying to just play music (not even trying to replicate this bug) I removed the configuration for my ampache server and added it back with a slightly different name. Since my ampache list was not populating anymore I was hoping to clear out a cache or something. Anyway, while the ampache list may have been loading I bumped the mouse and right clicked on the blank ampache browser and amarok crashed.
Thanks again
-- Backtrace (Reduced):
#6 Oxygen::WindowManager::canDrag (this=<value optimized out>, widget=0x282c680, child=<value optimized out>, position=...) at ../../../kstyles/oxygen/oxygenwindowmanager.cpp:551
#7 0x00007f4caa3927cf in Oxygen::WindowManager::mousePressEvent (this=0x1345980, object=0x282c680, event=<value optimized out>) at ../../../kstyles/oxygen/oxygenwindowmanager.cpp:241
#8 0x00007f4cbd7a6507 in QCoreApplicationPrivate::sendThroughObjectEventFilters (this=<value optimized out>, receiver=0x282c680, event=0x7fffd9537510) at kernel/qcoreapplication.cpp:847
#9 0x00007f4cbe188fac in QApplicationPrivate::notify_helper (this=0x12fe7c0, receiver=0x282c680, e=0x7fffd9537510) at kernel/qapplication.cpp:4392
#10 0x00007f4cbe18f30e in QApplication::notify (this=0x7fffd95381d0, receiver=0x282c680, e=0x7fffd9537510) at kernel/qapplication.cpp:3959
aha With the latest description, I could reproduce ! Now we have something to work with :) Git commit 631f5dd7ef5c68a1735985a96b9f34df944ec0e9 by Hugo Pereira Da Costa. Committed on 23/02/2011 at 17:42. Pushed by hpereiradacosta into branch 'master'. added protection when accessing listview model. CCBUG: 266087 M +2 -2 kstyles/oxygen/oxygenwindowmanager.cpp http://commits.kde.org/kde-workspace/631f5dd7ef5c68a1735985a96b9f34df944ec0e9 Git commit 1021789c698efc4f0598c57eb3e9efcf304c77c1 by Hugo Pereira Da Costa. Committed on 23/02/2011 at 17:42. Pushed by hpereiradacosta into branch 'KDE/4.6'. added protection when accessing listview model. CCBUG: 266087 M +2 -2 kstyles/oxygen/oxygenwindowmanager.cpp http://commits.kde.org/kde-workspace/1021789c698efc4f0598c57eb3e9efcf304c77c1 @Christian I think it fixes it. (And was easier than I thought). For the record: you can reproduce the crash in 100% of the cases. Just set an invalid ampache server in the config. Now: everytime you click on the (naturally empty) ampache music list, it crashes. Thats what my buxfix address. Will keep the bug open for a while, untill someone can confirm (possibly when kde4.6.1 is out, or if you compile KDE from sources), that the crash is fixed. Hugo |