The entire system crashes when using the seekbar of an YouTube video in Firefox

Ask Fedora
, , , ,
1

Did someone ever experience this issue before? When I’m watching an YouTube video and use the seek bar, the entire system freezes. Sometimes the display turns black and other times I can still see the display contents, but I can’t interact with anything; not even switching ttys works. The crash doesn’t happen consistently though. Also, this seems to happen only when I’m using a bluetooth headset.

I have no idea where to find the proper logs to find more about what’s going on. Gnome-shell logs don’t show anything; the ABRT app also doesn’t show anything; same thing for the logs entries in GNOME logs.

The crash happened again today 10 minutes ago… :frowning:

I’m using:

  • Firefox 124.0 (Flatpak from Flathub with the MOZ_ENABLE_WAYLAND env variable)
  • Fedora 39 Workstation
  • GNOME 45.4
  • Wayland
  • Shell extensions: GSConnect (the issue already happened without extensions IIRC)

Edit: sometimes when using the seekbar, the bluetooth codec profile switches “by itself” and I need to go to Settings > Sound and change the codec profile again, otherwise the audio keeps playing in very low quality. Other times, one side of the headset stops working and I have to reboot the system, because repairing the device doesn’t fix this issue. It seems that the root cause of the system crash might be related to the bluetooth stack.

2

I’d try launching firefox from a terminal window and see if any interesting notices are displayed when the error occurs.

3

Pretty much any YouTube activity causes multiple failures now using Xorg or Wayland constant tab crashes in Firefox or Chrome. I have many gnome crash messages as well.

My system has become almost un-usable.

4

Without a meaningful error message, we can only guess as to what is wrong. I would guess that it is due to a video codec and/or a video driver, but that is just a guess.

5

Will follow-up with logs but I’m out on the road at the moment

6

A quick test that can help to narrow things down would be to create a new user account on your system, sign in with that, and see if the problem persists. If not, the problem is likely something in your home directory. It is possible to install things like python libraries under your home directory in a way that other system utilities will pick them up and cause severe failures like what you are describing.

If you can leave journalctl -f running in a terminal (perhaps on a secondary monitor or in a SSH session connected from a separate computer), that would be a good way to “monitor” your logs as an event occurs.

1 Like
7

just going to post here the problems as they arise.

Apr 03 08:12:21 marks-linux-desktop abrt-server[8577]: Unsupported container technology
Apr 03 08:12:21 marks-linux-desktop abrt-server[8577]: Lock file ‘.lock’ was locked by process 8580, but it crashed?
Apr 03 08:12:21 marks-linux-desktop abrt-server[8577]: Error: No segments found in coredump ‘./coredump’
Apr 03 08:12:21 marks-linux-desktop abrt-server[8577]: Can’t open file ‘core_backtrace’ for reading: No such file or directory
Apr 03 08:12:21 marks-linux-desktop abrt-applet[3820]: g_app_info_should_show: assertion ‘G_IS_APP_INFO (appinfo)’ failed
Apr 03 08:12:21 marks-linux-desktop abrt-notification[8648]: [🡕] Process 8307 (postman) crashed in ??()

8

Thanks! The issue happened again today at 16h10, so I browsed the logs from the last boot and the last entry is from 15h50:

Aug 05 15:50:40 fedora kernel: perf: interrupt took too long (3298 > 3267), lowering kernel.perf_event_max_sample_rate to 60000

Also, now that I’m using Silverblue, the screen no longer blacks out. Instead, the UI freezes and I see a lot of red artifacts across the screen:

20240805_160440
20240805_1604401366×768 190 KB

9

I don’t think “interrupt took too long” is necessarily a serious error. It just means the system is bogged down a bit. Though if the kernel is being starved of resources, that might be an indication of a bug in a driver (probably your video driver since you are seeing artifacts there). What video card and driver are you using?

10

What video card and driver are you using?

Intel® UHD Graphics 620 (WHL GT2) (Mesa 24.1.5)

I don’t think “interrupt took too long” is necessarily a serious error

Hmm. This entry didn’t appeared in the second time

By the way, the issue happened again at 16h39. The screen glitch was much more messy and the artifacts kept flickering a lot:

20240805_163930
20240805_1639301366×768 408 KB

11

If the behavior is inconsistent as you describe, then it is probably hardware problem such as bad memory or overheating. Have you tried updating your system’s firmware?

12

Have you tried updating your system’s firmware?

It’s updated already. I always update it as soon as it shows up in the Software app. The last updates was almost a month ago.

overheating

The fans keep spinning a lot in the crashes, but the device doesn’t get very hot.

13

The firmware/bios updates for your PC might not show up in the Software app. I would recommend going directly to the website for your PC’s manufacturer and searching there to see if there are any firmware updates available for your hardware.

14

Acc. to the support website, the last firmware update version is 1.30.0, the same that I have installed

15

What is the output of lspci -v? It should indicate what video driver you are using. I expect it would be i915, but go ahead and run that command and share the output here in a preformatted block so we can be sure.

16 
00:02.0 VGA compatible controller: Intel Corporation WhiskeyLake-U GT2 [UHD Graphics 620] (rev 02) (prog-if 00 [VGA controller])
	DeviceName: Onboard - Video
	Subsystem: Dell Device 08ca
	Flags: bus master, fast devsel, latency 0, IRQ 142
	Memory at c0000000 (64-bit, non-prefetchable) [size=16M]
	Memory at b0000000 (64-bit, prefetchable) [size=256M]
	I/O ports at 4000 [size=64]
	Expansion ROM at 000c0000 [virtual] [disabled] [size=128K]
	Capabilities: <access denied>
	Kernel driver in use: i915
	Kernel modules: i915

Also, when I click on the video seekbar, the audio gets glitched for a short time in my bluetooth headphones: sometimes I hear the sound in very low quality or just at one side.

17

I just did a little searching and it looks like there are known problems with the i915 driver and the UHD Graphics 620/630 chips (e.g. Bug #1958191 “[i915] Screen flickering in Ubuntu 22.04 (until i9...” : Bugs : linux package : Ubuntu). Maybe you could add the nomodeset kernel parameter as a workaround until the i915 driver gets fixed? That parameter will use your PC’s CPU instead of the GPU for a lot of the video processing. It will make your system run slower and you probably won’t be able to run anything that does a lot of complex rendering such as a modern video game. However, it might workaround the garbled video output.

18

Can’t use this parameter. I use virtual machines, for example, but thanks for the tip!

The issue happened for the third time again today. I see that Mesa was updated yesterday, so this must be the cause of these frequent crashes:

  mesa-dri-drivers 24.1.4-3.fc40 -> 24.1.5-2.fc40
  mesa-filesystem 24.1.4-3.fc40 -> 24.1.5-2.fc40
  mesa-libEGL 24.1.4-3.fc40 -> 24.1.5-2.fc40
  mesa-libGL 24.1.4-3.fc40 -> 24.1.5-2.fc40
  mesa-libgbm 24.1.4-3.fc40 -> 24.1.5-2.fc40
  mesa-libglapi 24.1.4-3.fc40 -> 24.1.5-2.fc40
  mesa-libxatracker 24.1.4-3.fc40 -> 24.1.5-2.fc40
  mesa-va-drivers 24.1.4-3.fc40 -> 24.1.5-2.fc40
  mesa-vulkan-drivers 24.1.4-3.fc40 -> 24.1.5-2.fc40
19

You can try downgrading packages with, e.g., dnf downgrade mesa* (and you’ll likely need to reboot immediately). Let us know if you find that a specific package version is causing problems.

20

Ah, thanks! I will likely have to boot into the previous image, as I use Silverblue; just don’t know how long will I have to wait with an outdated OS. Also, this issue was happening before this update, but it seems that is now much worse.

Another thing is that the issue is also happening with GNOME Boxes sometimes…