Firefox freezes Tumbleweed

As the title says, fresh new installation of Tumbleweed but when I open firefox and go for example to reddit.com or even facebook.com the whole machine freezes and I have to hard-reboot manually. :frowning:

Why is this happening?

For some reason the newest Firefox seems to need swap and lots of it even it the OS is not swapping - make sure you have as much swap as you have ram and then some.

I have 8gb of ram and Firefox got hung, if you are lucky and have another machine you can ssh in and kill Firefox. After I added 9 gb of swap - Firefox rarely hangs for long - it will kill the runaway and continue. It seems that the new Firefox spawns a new instance for every site it connects to and Facebook pictures and videos each may need it’s own spawn.

this is what was running to connect to forums.opensuse.org and google

> ps -ef | grep fire
llrainey 2752 1647 9 Aug10 ? 02:39:08 /usr/lib64/firefox/firefox
llrainey 2827 2752 0 Aug10 ? 00:05:50 /usr/lib64/firefox/firefox -contentproc -childID 1 -isForBrowser -boolPrefs 36:1|183:1|300:0|310:0| -stringPrefs 286:36;471c7c62-0ebe-46bf-8ea8-b0e9fd12ecbd| -schedulerPrefs 0001,2 -greomni /usr/lib64/firefox/omni.ja -appomni /usr/lib64/firefox/browser/omni.ja -appdir /usr/lib64/firefox/browser 2752 true tab
llrainey 2911 2752 3 Aug10 ? 01:01:31 /usr/lib64/firefox/firefox -contentproc -childID 3 -isForBrowser -boolPrefs 36:1|183:1|300:0|310:0| -stringPrefs 286:36;471c7c62-0ebe-46bf-8ea8-b0e9fd12ecbd| -schedulerPrefs 0001,2 -greomni /usr/lib64/firefox/omni.ja -appomni /usr/lib64/firefox/browser/omni.ja -appdir /usr/lib64/firefox/browser 2752 true tab
llrainey 3601 2752 0 Aug10 ? 00:11:16 /usr/lib64/firefox/firefox -contentproc -childID 6 -isForBrowser -boolPrefs 36:1|183:1|300:0|310:0| -stringPrefs 286:36;471c7c62-0ebe-46bf-8ea8-b0e9fd12ecbd| -schedulerPrefs 0001,2 -greomni /usr/lib64/firefox/omni.ja -appomni /usr/lib64/firefox/browser/omni.ja -appdir /usr/lib64/firefox/browser 2752 true tab
llrainey 18923 2752 0 12:19 ? 00:00:06 /usr/lib64/firefox/firefox -contentproc -childID 41 -isForBrowser -boolPrefs 36:1|183:1|300:0|310:0| -stringPrefs 286:36;471c7c62-0ebe-46bf-8ea8-b0e9fd12ecbd| -schedulerPrefs 0001,2 -greomni /usr/lib64/firefox/omni.ja -appomni /usr/lib64/firefox/browser/omni.ja -appdir /usr/lib64/firefox/browser 2752 true tab
llrainey 20415 19178 0 13:09 pts/12 00:00:00 grep --color=auto fire

I am not experiencing this, but I am also not running the newest series of Firefox, instead I have the latest Firefox ESR 52 series, 52.9.0.

If you can have it confirmed by other TW users that this is a common problem, perhaps you might try installing the ESR version.

what’s your graphic card and or driver?

sudo lspci | grep VGA

This is the result:


00:01.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Kaveri [Radeon R7 Graphics]
01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Baffin [Radeon RX 460/560D / Pro 450/455/460/555/560] (rev cf)

Two discreet AMD PCIe gfxcards?

Once inxi is installed, what is its output run in an Xterm or Konsole?

inxi -Gxx

If xf86-video-amdgpu is installed/in use and xf86-video-ati is not, what happens if you swap them? What happens if you remove both?

Which session type is running? You can get this information with following command:


echo $XDG_SESSION_TYPE

et al:

Yes. This is happening in my episode of TW GNOME . . . started yesterday after installing the latest suggested updates . . . FF boots but within a few seconds will not load tabs . . . then it crashes the GUI. I don’t know how to do the “-f install” gambit from ubuntu over in OpenSUSE . . . rebooted a couple times . . . tried to run “dup -l” to see if a new version of FF would show in the list–nothing to do, it said–no change.

Then I installed Chromium and had no problems . . . .

This morning a bunch of newer updates were “ready to install” . . . installed them, FF was not listed . . . after the install launched FF . . . system “froze the GUI” . . . no cursor. Yesterday I tried to get to a TTY shell . . . wouldn’t do that . . . using ubuntu style “control + alt + F1” or F2, or F3, etc.

I just saw the post about increasing swap, which I would have to admit I’m not up to snuff on swap since I added more RAM, now have 16 GB RAM, previously had 8 GB, so I have 8 GB swap . . . . I do a multi-boot deal, with a mix of ubuntu flavors, two OpenSUSE, a Geck rolling & GNOME TW, with three versions of OSX spread across two int drives . . . . These days in Lubuntu & Ubuntu they aren’t doing swap partitions anymore, they are doing “swap file” . . . so FF is doing something that isn’t “up to date”?

I’m now over in Gecko Rolling and in FF to type this out, updated a few days back, not sure if I run the latest updates if the same thing will happen over here in Gecko . . . have to try it out and see if it gets “crashy” . . . or not . . . other stuff to do right now; just wanted to add my name to the list of “sufferers of frozen TW” in the GNOME DE flavoring.

n_s

as that is a strange setup I’d say disable hardware acceleration
hit the three bars then Options in the Performance section uncheck use “Use recommended performance settings” then uncheck use hardware acceleration

So, for comparison, I ran ref/dup -l in my Gecko Rolling partition, 192 packages . . . FF was not mentioned in the list . . . and on reboot FF is still working, system is still working . . . whereas in TW it is not. Gecko kernel is 4.17.14-1.2x86_64.rpm . . . and this is with the same roughly 8GB of RAM that the TW system is using . . . .

n_s

This is the result of the given command:

Graphics:
  Card-1: AMD Kaveri [Radeon R7 Graphics] driver: radeon v: kernel 
  bus ID: 00:01.0 chip ID: 1002:1313 
  Card-2: AMD Baffin [Radeon RX 460/560D / Pro 450/455/460/555/560] 
  driver: amdgpu v: kernel bus ID: 01:00.0 chip ID: 1002:67ef 
  Display: wayland server: X.org 1.20.0 driver: amdgpu 
  resolution: <xdpyinfo missing> 
  OpenGL: renderer: AMD Radeon RX 460 Graphics (POLARIS11 DRM 3.25.0 
  4.17.9-1-default LLVM 6.0.1) 
  v: 4.5 Mesa 18.1.5 compat-v: 3.1 direct render: Yes 

I do have 2 ‘graphics’ gadgets since I have an AMD A8 CPU.

 echo $XDG_SESSION_TYPE
wayland



What I would try, since each gfxcard is now running on a different gfxchip-specific driver, to see if any driver is responsible or plays a part:
1-exit X session; systemctl isolate multi-user
2-zypper in xdpyinfo
3-zypper rm xf86-video-ati # runs newer gfx on gfx-specific, older gfx on generic
4-systemctl isolate graphical # any better?
5-exit X session; systemctl isolate multi-user
6-zypper rm xf86-video-amdgpu; zypper in xf86-video-ati # runs newer gfx on generic, older gfx on gfx-specific
7-systemctl isolate graphical # how about now?
8-exit X session; systemctl isolate multi-user
9-zypper rm xf86-video-ati
10-systemctl isolate graphical # and now? (runs both gfx on generic)