(Solved)Firefox Causing Shutdowns?

[mdiemer]

I've been having problems with my computer shutting down. It only happens when I'm on Firefox, usually on a forum. I click on a forum topic, and my computer shuts down. It does not happen on Windows, where I also use Firefox. Is there a problem with the Linux version of Firefox? I had thought the shutdowns were a problem with my computer, but since it does not happen on Windows, I am now not so sure. In the meantime, I have installed and am using Opera. I tried installing Palemoon but was not successful, despite great help from Karl. Either I'm an idiot, or my computer is weird. Maybe both. If I can't get these shutdowns to stop, I'll have to stop using Mint. so far, the only Linux system it has not happened on is Zorin 9 32 bit.

[kukamuumuka]

Disable hardware acceleration from Firefox.

[Pierre]

will have to try that, as this article:
http://www.bbc.co.uk/news/resources/idt ... 2a22bd1ebc
re-booted this Mint 17x - TWICE - in as many minutes.


yet the Mageia 5 version of FF esr was fine & could read the entire article.

seen that on Windows Systems, where some oddbod issue caused that,
but, rarely on a Linux System.
so - it must be a FF issue.

Btw: the windows version of Palemoon runs well in ./wine.
https://www.codeweavers.com/compatibili ... e-moon-25x

EDIT: LM is always been very stable, so simply staying away from that link,
and the system is fine.

[Cosmo.]

If disabling hardware acceleration does not help, the problem is most likely caused by the firefox-profile. To find this out do the following:
Close FF.
Open your file-manager and make hidden files visible.
Rename the folder .mozilla to .mozilla.bck
Start FF, this will create automatically a new FF-profile.
Check.

To get back to your old profile:
Close FF.
Delete the new folder .mozilla
Rename the folder .mozila.bck back to .mozilla
Start FF. Done.

[Pierre]

back up your bookmarks, first though.


but anyway, that made no difference - that article still re-boots Mint

[karlchen]

will have to try that, as this article:
http://www.bbc.co.uk/news/resources/idt ... 2a22bd1ebc
re-booted this Mint 17x - TWICE - in as many minutes.

Was more lucky with FF 42.0 on Mint 13 64-bit Cinnamon. Being very patient I managed to close the Firefox tab where this webpage had been loaded. Everything was back to normal within 2 seconds then.
But while the BBC page was loaded it was clearly noticeable that there must be something in that page that makes FF 42.0 start hogging all kinds of resources. In particular mouse and keyboard input became really slow. Memory usage climbed up, but not dramatically. Hm. 500 MB for just one webpage might be considered dramatical though ?!
I wonder how long I would have had to wait till my system would have become 100% unresponsive in the end.

[kukamuumuka]

New features, like Sync, have spoiled Firefox.

[karlchen]

Btw: the windows version of Palemoon runs well in ./wine.
https://www.codeweavers.com/compatibili ... e-moon-25x

No need to resort to Wine and the Windows version of Pale Moon. The Linux edition can be installed and run just fine.
You only have to follow the installation instruction carefully and precisely.

[Flemur]

will have to try that, as this article:
http://www.bbc.co.uk/news/resources/idt ... 2a22bd1ebc

That's a funky web page; whenever I'm displaying it, FF and Xorg are each using 20-30% CPU.
(Slightly less if you're at the top of the page).
No crash or hang, though. Hardware accel was off, now I turned it on - I'll post if that changes anything.
Edit: HW accel ON = the same or slightly more CPU use. That's just looking at the page...flashblock blocking the videos.

Code: Select all

yet the Mageia 5 version of FF esr was fine & could read the entire article.

Perhaps use that working version in Mint ... ?

[Dave B]

Hi mdiemer,

Considering you're using older hardware, there are three things you can try.

• 1) Check Driver Manager for an legacy NVIDIA driver, and see if there are any recommended drivers available.
  2) "Firefox" -> "Add-ons" -> "Plugins" -> "Shockwave Flash", set to either "Ask to Activate" or "Never Activate".
  3) Try Firefox add-on's "Addblock Plus" and possibly "No Script".

David

[karlchen]

Hi, guys.

Things are getting more confusing and more complicated than a simple "it is this stupid Firefox 42.0 on Linux that causes all the problems"

• I have never experienced that Firefox directly or indirectly caused any of my Linux Mint systems to reboot like reported by mdiemer. - This does not mean that I doubt that exactly this happens on mdiemer's machine. It only means it has not happened here.
• I have seen that the BBC webpage that Pierre had linked to and that made his system reboot causes my "power machine", a quad core with 4 GB of RAM and an NVidia GeForce graphics card and Nvidia drivers software, slow down to a crawl.
• So far all this might confirm that all of you are right who put the blame on Firefox 42.0 on Linux as the culprit.
• But right before writing this post I have opened the same webpage in Firefox 42.0 on a really weak Acer Aspire One D260 netbook, running Mint 17.2 xfce 32-bit, Firefox 42.0, Flash 11.2.202.554 (always enabled).
  And what happened?
  The page got loaded completely. No problem.
  The system remained responsive. No noticeable slowdown. No problem.
  I even clicked to play one of the embedded Flash movies. The Flash movie played. The 2 Atom cores worked at almost 100%. No problem. No slowdown. Ram usage did not exceed 600 MB (of 990 MB total).
• How does this fit the assumption Firefox 42.0 for Linux were the (only) culprit?
  I am afraid that there is not just one culprit, it might be the combination of several factors, several software components that co-operate badly. - In my specific case I would not be amazed if the proprietary NVidia driver were one of the culprits, too.

I strongly assume that in all cases where a webpage, displayed in Firefox 42.0, seems to be able to trigger a system reboot, we will have to take a closer look at the system hardware. In particular at the graphics card and at the video driver software in use.

Regards,
Karl
--
In all test cases mentioned below, Flash 11.2.202.554 was enabled, java scripts were enabled.
--
inxi report of Medion QuadCore Desktop with nVidia GeForce 8500 GT and nVidia driver
This is the machine brought down to a crawl when the BBC page had been loaded. - But so far it only happened once. Could not reproduce the problem on the same machine later on the same day (cf. my next post), though no configuration item had been changed.
Maybe the Medion desktop inxi report will help find common hardware / driver details nonetheless which can be found on other machine where the problem occurs, too, regularly or sporadically.

Code: Select all

$ inxi -Fxz 
System:    Host: voyager Kernel: 3.2.0-95-generic x86_64 (64 bit gcc: 4.6.3)
           Desktop: Cinnamon 2.0.14 (Gtk 3.4.2-0ubuntu0.8) Distro: Linux Mint 13 Maya
Machine:   System: MEDIONPC product: MS-7358
           Mobo: MICRO-STAR model: MS-7358 v: Fab D Bios: Phoenix v: 6.00 PG date: 07/24/2007
CPU:       Quad core Intel Core2 Quad (-MCP-) cache: 4096 KB flags: (lm nx sse sse2 sse3 ssse3 vmx) bmips: 19141 
           clock speeds: max: 2400 MHz 1: 1600 MHz 2: 1600 MHz 3: 1600 MHz 4: 1600 MHz
Graphics:  Card: NVIDIA G86 [GeForce 8500 GT] bus-ID: 01:00.0
           Display Server: X.Org 1.11.3 drivers: nvidia (unloaded: nouveau,vesa,fbdev)
           Resolution: 1280x1024@60.0hz
           GLX Renderer: GeForce 8500 GT/PCIe/SSE2 GLX Version: 3.3.0 NVIDIA 319.32 Direct Rendering: Yes
Audio:     Card Intel 82801I (ICH9 Family) HD Audio Controller driver: snd_hda_intel bus-ID: 00:1b.0
           Sound: Advanced Linux Sound Architecture v: 1.0.24
Network:   Card: Realtek RTL8169 PCI Gigabit Ethernet Controller
           driver: r8169 v: 2.3LK-NAPI port: ac00 bus-ID: 04:00.0
           IF: eth0 state: up speed: 1000 Mbps duplex: full mac: <filter>
Drives:    HDD Total Size: 500.1GB (3.5% used) ID-1: /dev/sda model: ST3500830AS size: 500.1GB temp: 38C
Partition: ID-1: / size: 46G used: 9.0G (21%) fs: ext4 dev: /dev/sda6
           ID-2: swap-1 size: 8.22GB used: 0.00GB (0%) fs: swap dev: /dev/sda8
RAID:      No RAID devices: /proc/mdstat, md_mod kernel module present
Sensors:   System Temperatures: cpu: 21.8C mobo: N/A gpu: 0.0:47C
           Fan Speeds (in rpm): cpu: N/A
Info:      Processes: 164 Uptime: 5 min Memory: 467.6/3261.8MB Init: Upstart runlevel: 2 Gcc sys: 4.6.3
           Client: Shell (bash 4.2.251) inxi: 2.2.16 

inxi report of Acer Aspire One D260
(problem not reproduced)

Code: Select all

karl@paulchen ~ $ inxi -Fxz
System:    Host: paulchen Kernel: 3.13.0-71-generic i686 (32 bit gcc: 4.8.2) Desktop: Xfce 4.12.2 (Gtk 2.24.23)
           Distro: Linux Mint 17.2 Rafaela
Machine:   System: Acer (portable) product: AOD260 v: V1.08_Vodafone
           Mobo: Acer model: AOD260 v: V1.08_Vodafone Bios: Acer v: V1.08_Vodafone date: 06/29/2010
CPU:       Single core Intel Atom N450 (-HT-) cache: 512 KB flags: (lm nx pae sse sse2 sse3 ssse3) bmips: 3325 
           clock speeds: max: 1666 MHz 1: 1333 MHz 2: 1666 MHz
Graphics:  Card: Intel Atom Processor D4xx/D5xx/N4xx/N5xx Integrated Graphics Controller bus-ID: 00:02.0
           Display Server: X.Org 1.15.1 drivers: intel (unloaded: fbdev,vesa) Resolution: 1024x600@60.0hz
           GLX Renderer: Mesa DRI Intel IGD x86/MMX/SSE2 GLX Version: 1.4 Mesa 10.1.3 Direct Rendering: Yes
Audio:     Card Intel NM10/ICH7 Family High Definition Audio Controller driver: snd_hda_intel bus-ID: 00:1b.0
           Sound: Advanced Linux Sound Architecture v: k3.13.0-71-generic
Network:   Card-1: Qualcomm Atheros AR8132 Fast Ethernet driver: atl1c v: 1.0.1.1-NAPI port: 5000 bus-ID: 01:00.0
           IF: eth0 state: down mac: <filter>
           Card-2: Broadcom BCM4313 802.11bgn Wireless Network Adapter driver: wl bus-ID: 02:00.0
           IF: wlan0 state: up mac: <filter>
Drives:    HDD Total Size: 250.1GB (3.4% used) ID-1: /dev/sda model: TOSHIBA_MK2565GS size: 250.1GB
Partition: ID-1: / size: 70G used: 6.2G (10%) fs: ext4 dev: /dev/sda5
           ID-2: swap-1 size: 2.15GB used: 0.00GB (0%) fs: swap dev: /dev/sda3
RAID:      No RAID devices: /proc/mdstat, md_mod kernel module present
Sensors:   System Temperatures: cpu: 45.0C mobo: N/A
           Fan Speeds (in rpm): cpu: N/A
Info:      Processes: 167 Uptime: 1:00 Memory: 443.2/991.4MB Init: Upstart runlevel: 2 Gcc sys: 4.8.4
           Client: Shell (bash 4.3.111) inxi: 2.2.16 
karl@paulchen ~ $

<added Monday, December 14th, 2015/>
inxi report of Zotac with nVidia ION GPU chip and nouveau driver
(problem not reproduced, either, neither using Firefox 42.0, nor using Palemoon 25.8.1)

Code: Select all

karl@unimatrix0 ~ $ inxi -Fxz
System:    Host: unimatrix0 Kernel: 3.13.0-71-generic i686 (32 bit gcc: 4.8.2) Desktop: Xfce 4.11.8 (Gtk 2.24.23)
           Distro: Linux Mint 17.1 Rebecca
Machine:   Mobo: N/A model: N/A Bios: American Megatrends v: 080015 date: 10/09/2009
CPU:       Dual core Intel Atom 330 (-HT-MCP-) cache: 512 KB flags: (lm nx pae sse sse2 sse3 ssse3) bmips: 6399 
           clock speeds: max: 1599 MHz 1: 1599 MHz 2: 1599 MHz 3: 1599 MHz 4: 1599 MHz
Graphics:  Card: NVIDIA ION VGA bus-ID: 03:00.0
           Display Server: X.Org 1.15.1 drivers: nouveau (unloaded: fbdev,vesa) Resolution: 1280x1024@60.0hz
           GLX Renderer: Gallium 0.4 on NVAC GLX Version: 3.0 Mesa 10.1.3 Direct Rendering: Yes
Audio:     Card NVIDIA MCP79 High Definition Audio driver: snd_hda_intel bus-ID: 00:08.0
           Sound: Advanced Linux Sound Architecture v: k3.13.0-71-generic
Network:   Card-1: NVIDIA MCP79 Ethernet driver: forcedeth port: d080 bus-ID: 00:0a.0
           IF: eth0 state: down mac: <filter>
           Card-2: Qualcomm Atheros AR9285 Wireless Network Adapter (PCI-Express) driver: ath9k bus-ID: 04:00.0
           IF: wlan0 state: up mac: <filter>
Drives:    HDD Total Size: 160.0GB (12.9% used) ID-1: /dev/sda model: SAMSUNG_HM160HI size: 160.0GB temp: 46C
Partition: ID-1: / size: 48G used: 15G (33%) fs: ext4 dev: /dev/sda6
           ID-2: swap-1 size: 5.42GB used: 0.00GB (0%) fs: swap dev: /dev/sda5
RAID:      No RAID devices: /proc/mdstat, md_mod kernel module present
Sensors:   System Temperatures: cpu: 84.0C mobo: N/A gpu: 60.0
           Fan Speeds (in rpm): cpu: N/A
Info:      Processes: 184 Uptime: 25 min Memory: 421.3/1760.2MB Init: Upstart runlevel: 2 Gcc sys: 4.8.4
           Client: Shell (bash 4.3.111) inxi: 2.2.16 
karl@unimatrix0 ~ $ 

<added Wednesday, December 16th, 2015/>
inxi report of Zotac with nVidia ION GPU chip and nVidia driver
(problem not reproduced, neither using Firefox 43.0, nor using Firefox 38.5.0esr)

Code: Select all

karl@unimatrix0 ~ $ inxi -Fxz
System:    Host: unimatrix0 Kernel: 3.2.0-95-generic i686 (32 bit, gcc: 4.6.3) 
           Desktop: Xfce 4.10.0 (Gtk 2.24.10) Distro: Linux Mint 13 Maya
Machine:   Mobo: N/A model: N/A Bios: American Megatrends version: 080015 date: 10/09/2009
CPU:       Dual core Intel Atom CPU 330 (-HT-MCP-) cache: 512 KB flags: (lm nx sse sse2 sse3 ssse3) bmips: 6399.76 
           Clock Speeds: 1: 1599.940 MHz 2: 1599.940 MHz 3: 1599.940 MHz 4: 1599.940 MHz
Graphics:  Card: NVIDIA ION VGA bus-ID: 03:00.0 
           X.Org: 1.11.3 drivers: nvidia (unloaded: fbdev,vesa,nouveau) Resolution: 1280x1024@60.0hz 
           GLX Renderer: ION/integrated/SSE2 GLX Version: 3.3.0 NVIDIA 319.32 Direct Rendering: Yes
Audio:     Card: NVIDIA MCP79 High Definition Audio driver: snd_hda_intel bus-ID: 00:08.0 Sound: ALSA ver: 1.0.24
Network:   Card-1: Qualcomm Atheros AR9285 Wireless Network Adapter (PCI-Express) driver: ath9k bus-ID: 04:00.0
           IF: wlan0 state: up mac: <filter>
           Card-2: NVIDIA MCP79 Ethernet driver: forcedeth port: d080 bus-ID: 00:0a.0
           IF: eth0 state: down mac: <filter>
Drives:    HDD Total Size: 167.8GB (7.9% used) 1: /dev/sda SAMSUNG_HM160HI 160.0GB 
           2: USB /dev/sdb Premium 7.8GB 
Partition: ID: / size: 48G used: 6.9G (16%) fs: ext4 ID: swap-1 size: 5.42GB used: 0.00GB (0%) fs: swap 
Sensors:   System Temperatures: cpu: 67.6C mobo: 51.0C gpu: 0.0:58C 
           Fan Speeds (in rpm): cpu: N/A 
Info:      Processes: 174 Uptime: 1:20 Memory: 453.8/1759.2MB Runlevel: 2 Gcc sys: 4.6.3 Client: Shell inxi: 1.7.33 
karl@unimatrix0 ~ $ 

[Cosmo.]

Here LM 17.3 Cinnamon with FF 42 from the repo in a virtual machine, 1 CPU assigned: No problem with the page, the vm runs stable, cpu usage jumps between 1 and at max 10 %, firefox process occupies about 250 MB RAM. No add-ons in ff added, no settings changed from default. Scrolling the page works smoothly.

If the verdict about the nvidia driver should be true, selecting the Intel prime profile in the nvidia driver might probably help and possibly give a hint.

[mdiemer]

Thanks for all the interesting comments. As I said in my post, right now I am using Opera, and in fact will install it on all Linux systems (Mint and Zorin). If I like it, and so far I do, and I don't get shutdowns, I will just stay with Opera. much simpler than figuring out exactly what settings re causing the problem.

Re: video drivers, I routinely update to the recommended Nvidia driver when installing an OS. It does help, the shutdowns occur less often, but they still occur. I agree, it is probably an interaction of browser settings with my hardware.

Opera seems more streamlined than Firefox. Also, it has DuckDuck as the default search engine. Another plus. I'll see how it goes, and let you all know.

[karlchen]

Hi, mdiemer.

Personally I perfectly understand your decision to try out Opera and to find out whether the problem will simply not happen when using Opera. I am curious to learn how Opera will perform.

Cheers,
Karl

[karlchen]

Hi, guys.

Although I do understand that having a usable browser is more important than analyzing problems experienced when using Firefox, I was, of course, curious to find out whether
+ the problems would be reproducible on my Medion Desktop when using Firefox 42.0 (where they had occurred)
+ the problems would be reproducible on my Medion Desktop when using Firefox 38.4 ESR (where I had not tried, yet.)
I was even more intrigued to waste some more time on the problems, because the weakest machine that I have had beaten my beloved Medion Desktop. (see my previous post).
What can I say? To my amazement, tonight I could not reproduce the problem at all
+ neither using Firefox 42.0
+ nor using Firefox 38.4 ESR
No lags, mouse responsive, keyboard responsive. Yet, I could see that the overall CPU usage was higher by roundabout 10 to 20 % when using Firefox 42.0 to view the BBC page and in particular when watching some of the imbedded Flash movies.
Xorg climbed up to 80% CPU usage in both cases, but Firefox and Flash plugin used clearly more CPU when using Firefox 42.0, than when using Firefox 38.4 ESR.
I might feel like repeating this BBC webpage test on the same and on other machines in order to find out whether the problems experienced this afternoon are reproducible regularly, sporadically or not at all.
Do not worry, I will not bore you stiff by writing a lengthy post each time.

Cheers,
Karl

[mdiemer]

Well, I tried Opera. Unfortunately, it does not have PRNG, and I have email that requires it. Sooo, I downloaded Chromium. Trying that out now. It does have PRNG, so all good there. It also imported my bookmarks really fast. So far it seems like a fast and sleek browser. Very intuitive. The test, of course, is whether the shutdowns will occur using it. That will take a week or so to figure out, as I try it on my various GNU/Linux systems (MInt and Zorin). I really do want to get to the bottom of this, but I'd like to do so by using other browsers. If they don't have the problem, that tells me it is Firefox, or Firefox on my system. I'm not as concerned with figuring out exactly what in Firefox is the culprit. I just want a browser that works on my computer.

[karlchen]

I just want a browser that works on my computer.

I do understand this wish and this expectation.
Yet, I do not share the underlying assumption that it is the browser that causes the shutdown, not in your case and not in Pierre's case.
Shutting down the system requires root privileges, but the browser runs with user privileges under your account. No browser offers any function to shutdown or reboot a system.
Even though my tests are not scientific, they should be sufficient to raise doubts whether the assumption can be right that Firefox 42.0 triggers system shutdowns.
For a lot of users the browser is merely the application which they use 90% of the time while they use their computers. This is why the browser will be up and running when something goes wrong. And users will tend to put the blame on the browser while in reality some lower level system components cause the malfunctions. Only they may not have a visible user interface.
Inspecting the system logfile /var/log/syslog for error messages generated around the time when the system shuts down unexpectedly might shed some light on the mystery.

[mdiemer]

Could certain settings, like accelerators - as has been suggested - be triggering video card malfunction, which then causes the shutdowns? If so, I would think that using a browser that is less bloated might avoid this. In any case, if I do not experience shutdowns with Chromium, but I do with Firefox, it does seem like a logical conclusion that Firefox, on this particular system, just doesn't work well. Again, time will tell. By no means have I concluded anything. In fact, I expect the problem will return, even with Chromium. I can only try it and hope.

[Dave B]

May I also recommend running "Memory Test (memtest86+)" found on the GRUB boot screen, to help rule out potential hardware defects, test takes around 20-30 minutes to complete.

[mdiemer]

May I also recommend running "Memory Test (memtest86+)" found on the GRUB boot screen, to help rule out potential hardware defects, test takes around 20-30 minutes to complete.

Good advice, I will do that.