You are not logged in.

#101 2014-01-21 19:02:21

Cotton
Member
From: Cornwall, UK
Registered: 2004-09-17
Posts: 568

Re: [SOLVED] Firefox 26 Crashing - Assertion Failure

I concur.
No failures here with the latest nvidia, X and firefox packages.

Offline

#102 2014-01-23 00:05:26

AndreLDM
Member
From: Brazil
Registered: 2013-07-25
Posts: 8

Re: [SOLVED] Firefox 26 Crashing - Assertion Failure

Seems fixed after the last update.

Offline

#103 2014-01-31 01:12:03

slakam
Member
Registered: 2014-01-31
Posts: 3

Re: [SOLVED] Firefox 26 Crashing - Assertion Failure

Firefox started going crazy again for me a couple of days ago. I'm talking constant crashes. I'm running it on Virtualbox. Swapping to the linux-ck kernel in both the host ans the guest seems to have solved the issue.

Offline

#104 2014-02-03 17:17:31

bartonc
Member
From: Minden, Nevada, USA
Registered: 2012-02-27
Posts: 25

Re: [SOLVED] Firefox 26 Crashing - Assertion Failure

Freezing X: Yes
Also crashing in Virtualbox Guest (64-bit Windows): Yes
I'll disable Sync and see if that makes a difference.

EDIT: Only had Sync on in the vm guest, not the 64-bit host where ff crashed, but did not freeze X.
My 32-bit HTPC had Sync enable - X has been freezing there.

Last edited by bartonc (2014-02-03 17:28:46)

Offline

#105 2014-02-12 15:14:24

JKAbrams
Member
From: Sweden
Registered: 2011-11-07
Posts: 81

Re: [SOLVED] Firefox 26 Crashing - Assertion Failure

The issue is not solved while running nuoveau (nouveau-dri 10.0.3-1, lib32-nouveau-dri 10.0.3-1, xf86-video-nouveau 1.0.10-2, mesa 10.0.3-1 mesa-libgl 10.0.3-1) running KDE 4.12.2-1.
I've been having this issue for half a year, hoping for every update that maybe it's fixed now, but crashing is now a fact of life with Arch for me, it's depressing because I really like Arch.
It's the same deal, random crashes but heavy sites, or sites using flash or the built in PDF-viewer seem to have a higher risk of crashing.

I've also been experiencing X-crashes the last couple of months, there have been three four different kinds:
-Total screen freeze, not responding to anything, these seem to have been replaced by the following kind since an update a few weeks ago:
-Being thrown out to shell, mouse cursor still visible.
-Gradual escalation of "acting up", system becomes slow to respond on user input, but if I catch this early enough and manage to start the task manager to kill (in order of success-rate) flashplayer, firefox, audacious, vlc, okular I might get the system back.
This last one is somewhat like an out of memory crash, but the actual memory (according to task manager) is not full (usually about 2.5/5gb usage). (Would a full graphics memory give the same behavior?)
Edit: just found another kind:
-Automatic restart (no useful information in logs)

The reason I mention these other crashes is that a hunch (from the experience of the crashes) tells me they are related (I always run Firefox and most of the time when the crashes occur that was the application I was working in) but I might be mistaken. It seems likely to me that these are graphics related.

What I've tried:
-All settings I could find relating to hardware acceleration (firefox, flashplugin, KDE): enabling them makes the problems much worse, but disabling them still gives lots of problems
-Clean profile, no addons, safe-mode, reinstalling (firefox, nouveau, X, KDE), clean KDE-profile (new user)
-Running from the command line
-Firefox 26-27
-I don't run firefox sync
-I do use linux-ck
-I do use two screens

For about every 5 firefox crashes there is a full system crash.

Last edited by JKAbrams (2014-02-12 18:56:39)

Offline

#106 2014-04-16 00:43:26

JKAbrams
Member
From: Sweden
Registered: 2011-11-07
Posts: 81

Re: [SOLVED] Firefox 26 Crashing - Assertion Failure

Still crashing like a destruction derby car...
Now on FF 28, with nvidia closed drivers (no difference).
Maybe one day It'll make me be sad enough that I teach myself to create proper traceback logs...

For some time I thought the issue was this: bug 833117 (turns out it wasn't)
Some Mozilla guys suggests perhaps this one: bug 920200

Edit: Oh the irony - it crashed while posting a edit to this message (fixing a typo).....

Last edited by JKAbrams (2014-04-16 00:45:49)

Offline

#107 2014-05-14 09:07:51

JKAbrams
Member
From: Sweden
Registered: 2011-11-07
Posts: 81

Re: [SOLVED] Firefox 26 Crashing - Assertion Failure

Just found this bug which is very much related:
https://bugs.archlinux.org/task/39806

One person suggested disabling systemd application crash dump journaling as a way to combat the system crashes:
https://wiki.archlinux.org/index.php/Sy … journaling

I'm gonna give it a try.

Offline

#108 2014-05-14 09:15:35

JKAbrams
Member
From: Sweden
Registered: 2011-11-07
Posts: 81

Re: [SOLVED] Firefox 26 Crashing - Assertion Failure

When posting a comment to the bug tracker I got this error:

Completely unexpected exception: Address in mailbox given [] does not comply with RFC 2822, 3.6.2.
This should never happend, please inform Flyspray Developers

While trying to post the last message to the forum I got this:

An error was encountered
Error: Unable to delete from online list.

When trying to post this message I get:

An error was encountered
Error: Unable to update user visit data.

Alright Arch, I surrender.

Offline

#109 2015-05-01 23:30:05

osirisgothra
Member
Registered: 2013-09-24
Posts: 4

Re: [SOLVED] Firefox 26 Crashing - Assertion Failure

at the place where they actually are processing the bug, it is not a good sign because it is given 'normal' status, been reopened (as far as I can tell) and resolved a few times over, and yet, it keeps coming back on a regular basis like a bad cold. I think the folks at mozilla just issue patches to 'plug the leak' so to speak, and then whenever code shifts down the line (the 'bucket gets moved') there it is again. Since it isn't seen by people who don't use the terminal which are usually the biggest complainers, it isn't getting much focus. As for me, I dont know which is worse, having to mod the firefox startup script to hide the darn message, or, fixing the build every time I need a new binary with some iffy patch (or done myself, which is usually even iffier). And why are these messages not configurable to be shut off like others? We wont get answers, and its not important to anyone who could provide one.

Offline

Board footer

Powered by FluxBB