Closed
Bug 840378
Opened 12 years ago
Closed 10 years ago
Right-click menu and extensions menu do not open after a while
Categories
(Firefox :: Untriaged, defect)
Tracking
()
RESOLVED
INVALID
People
(Reporter: stephan.diestelhorst, Unassigned)
Details
User Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:20.0) Gecko/20130210 Firefox/20.0
Build ID: 20130210122734
Steps to reproduce:
Browse the web for a while (a few hours). Try to open the right-click context menu, or the menu of an extension.
Things tried:
* started Firefox without hardware accel -> could reproduce
* started window manager (KDE 4.10 / kwin) without hardware accel -> could reproduce
* started Firefox with a clean profile -> could reproduce
Actual results:
Eventually, the context menu for right-clicking into a web site does not appear anymore. On click, only the border of the menu flashes briefly, but the content is not properly painted. Also cannot click into the (non-existent) menu to invoke any actions.
The same is true for the menus from the extensions in my extension bar.
Expected results:
The menu should properly be displayed.
Reporter | ||
Comment 1•12 years ago
|
||
Note that this has been a recent development, so this either broke with the Aurora versions since a few days (less than a week, so 7.Feb.2013), or with an update in KDE (from 4.9.x to 4.10). I am happy to cross-report with KDE.
In contrast to https://bugzilla.mozilla.org/show_bug.cgi?id=633955, the standard menus "File", etc. do still work and the issue is not related to the monitor this is running on.
Do you know a "fast" method to reproduce it without browsing the web a few hours?
If yes, try to use the tool mozregression to find a regression range (maybe you could leave open each test builds during 1 or 2 hours before trying to reproduce the issue).
See http://harthur.github.com/mozregression/
Flags: needinfo?(stephan.diestelhorst)
Keywords: regressionwindow-wanted
Reporter | ||
Comment 3•12 years ago
|
||
Not yet :( I just restarted Firefox, will look into specifying a more precise regression range with mozregression. Is there any way to inspect state when this has gone bad? I tried strace-ing, but without anything to look for, this is likely hard. I also got GDB installed, I could attach to a session and inspect with breakpoints, etc.
Flags: needinfo?(stephan.diestelhorst)
Comment 4•12 years ago
|
||
stephan, is you video card driver updated to the latest version?
Reporter | ||
Comment 5•12 years ago
|
||
I was on a fairly recent one (fglrx 9.002 from Ubuntu) and just upgraded (to fglrx 9.012 from AMD website). Let's see if it still crashes. Had a bit of a run yesterday, but it was still unclear what I did to trigger. Any pointers how to inspect the process, once I have the issue?
Reporter | ||
Comment 6•12 years ago
|
||
It is quite likely not the video driver (or not a specific one). This happened again on my other machine:
lspci -v
00:02.0 VGA compatible controller: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07) (prog-if 00 [VGA conntroller])
Subsystem: Lenovo Device 20e4
Flags: bus master, fast devsel, latency 0, IRQ 45
Memory at f2000000 (64-bit, non-prefetchable) [size=4M]
Memory at d0000000 (64-bit, prefetchable) [size=256M]
I/O ports at 1800 [size=8]
Expansion ROM at <unassigned> [disabled]
Capabilities: [90] MSI: Enable+ Count=1/1 Maskable- 64bit-
Capabilities: [d0] Power Management version 3
Kernel driver in use: i915
Kernel modules: i915
00:02.1 Display controller: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07)
Subsystem: Lenovo Device 20e4
Flags: bus master, fast devsel, latency 0
Memory at f2400000 (64-bit, non-prefetchable) [size=1M]
Capabilities: [d0] Power Management version 3
And the original machine is
lspci -v
00:01.0 VGA compatible controller: Advanced Micro Devices [AMD] nee ATI Wrestler [Radeon HD 6310] (prog-if 00 [VGA controller])
Subsystem: Lenovo Device 21df
Flags: bus master, fast devsel, latency 0, IRQ 45
Memory at e0000000 (32-bit, prefetchable) [size=256M]
I/O ports at 4000 [size=256]
Memory at f0200000 (32-bit, non-prefetchable) [size=256K]
Expansion ROM at <unassigned> [disabled]
Capabilities: [50] Power Management version 3
Capabilities: [58] Express Root Complex Integrated Endpoint, MSI 00
Capabilities: [a0] MSI: Enable+ Count=1/1 Maskable- 64bit+
Capabilities: [100] Vendor Specific Information: ID=0001 Rev=1 Len=010 <?>
Kernel driver in use: fglrx_pci
Kernel modules: fglrx, radeon
Both machines run the same KDE and Aurora versions. When I found the error again, I tried switching the window manager (from Kwin to IceWM and LWM), but that didn't bring back the menu. Will proceed with trying the regression browsers.
Reporter | ||
Comment 7•12 years ago
|
||
I ran the mozregression between 2013-02-05 and 2013-02-11, but could not reproduce the issue (even though I also used my personal profile with all the extensions etc.) Is there any difference between versions in Nighty vs. Aurora?
Also I sometimes had issues with keyboard input stopped working in Nightly, is that a known issue or should I report?
NB: Using mozregression with your current profile modifies some settings in about:config, it's better to use a copy of your profile to do that. Think to reset the prefs.
Comment 9•12 years ago
|
||
(In reply to stephan.diestelhorst from comment #7)
> Also I sometimes had issues with keyboard input stopped working in Nightly,
> is that a known issue or should I report?
That's expected for Linux, unfortunately: bug 774892
To workaround: you can toggle the pref in about:config through copy-paste using a text editor and the context menu's paste option.
(In reply to stephan.diestelhorst from comment #7)
> Is there any difference between versions in Nighty vs.
> Aurora?
In theory, if you can reproduce something on Aurora, then it should have been present in Nightly as well. But with bugs ported to other branches/backed out/fixed you never know for sure.
There's a repo option for mozregression: you should be able to change the branch to Aurora through that (though I never used it, so I can't guarantee)
mozregression --repo=mozilla-1.9.2
Reporter | ||
Comment 10•12 years ago
|
||
Currently bisecting manually through Auroa builds.
![]() |
||
Updated•12 years ago
|
Flags: needinfo?(stephan.diestelhorst)
Reporter | ||
Comment 11•12 years ago
|
||
I have been bisecting through the Aurora builds manually.
Source:
http://ftp.mozilla.org/pub/mozilla.org/firefox/nightly/?C=M;O=D
*-aurora/firefox-20.0a2.en-US.linux-x86_64.tar.bz2
Bisection log:
2013-02-05
2013-02-06
2013-02-07
2013-02-08 Tested, no issue.
2013-02-09
2013-02-10 Tested, no issue.
2013-02-11 Tested, no issue.
This is very strange, since the build 2013-02-11 from the Aurora Ubuntu PPA has definitely been faulty. Is there any documentation on the difference between these two? One thing that is missing in the vanilla builds is the support to have the menu bar external to the application (in Unity / KDE). Apart from that, I tested with all extensions and my full profile.
I will switch back to the PPA builds and see if the error persists.
Flags: needinfo?(stephan.diestelhorst)
Reporter | ||
Comment 12•12 years ago
|
||
Switching back to the PPAs immediately toggled this bug again. I suspect the firefox-globalmenu plugin that changes the menu placement of Firefox.
http://packages.ubuntu.com/quantal/firefox-globalmenu
Will uninstall that and see if the issue persists.
Reporter | ||
Comment 13•12 years ago
|
||
Uninstalled this plugin, but the menu bar is still integrated into my global menu with the newest Ubuntu Aurora builds (22.0a2 (2013-04-04)) and this still fails. Reporting this as Ubuntu-specific in their bugtracker: https://bugs.launchpad.net/ubuntu/+source/firefox/+bug/1164953
![]() |
||
Comment 14•10 years ago
|
||
On the same day, Chris Coulson wrote there: "This is a bug in dbusmenu-qt for sending bogus timestamps, and is a duplicate of an already existing bug".
Closing; please reopen if something in it turns out to be a Firefox bug. Thanks for filing.
Status: UNCONFIRMED → RESOLVED
Closed: 10 years ago
QA Whiteboard: [bugday-20150330]
Resolution: --- → INVALID
![]() |
||
Updated•10 years ago
|
Keywords: regressionwindow-wanted
You need to log in
before you can comment on or make changes to this bug.
Description
•