You are not logged in.
What date were you trying to downgrade to when that message was produced? Please post the command and its full output.
Offline
Given that the system is malfunctioning, why has a solution not been implemented? I require the camera for my job...
Can you access the camera with
$ ffplay /dev/video0
?
What is output of
$ udevadm info /dev/video0
?
ls -l /dev/video0
crw-rw-rw- 1 root root 81, 0 12-20 21:28 /dev/video0
Systemd uses video group but permissions in your system differ from default. Do you have some additional udev rules for v4l devices installed?
Offline
https://bbs.archlinux.org/viewtopic.php … 2#p2215682
You're running into the egl split from nvidia-utils a month ago, you'll have to manually remove the egl-* packages.
pacman -Qo /usr/lib/libnvidia-egl-xcb.so
will tell you what currently owns that file.
I already said the kernel isn't the problem.
We actually don't 100% know that because the LTS kernel could have picked up the bogus commit, but if and since the LTS kernel was also affected it actually seems more likely the systemd 257 update, which can be downgraded w/o having to deal w/ the nvidia packages.
Until we figured whom to blame, this is your problem, your responsibility and PSA:
You'll have to keep moving your ass to do something about that - and stomping your foot real hard and yelling demands into the internet generally won't cut it in life.
Offline
ffplay /dev/video0
[video4linux2,v4l2 @ 0x7dfbcc000c80] Cannot open video device /dev/video0: No such file or directory
/dev/video0: No such file or directory
udevadm info /dev/video0
Unknown device "/dev/video0": No such device
What date were you trying to downgrade to when that message was produced? Please post the command and its full output.
Server=https://archive.archlinux.org/repos/2024/12/01/$repo/os/$arch
błąd: nie udało się dokonać transakcji (konfliktujące pliki)
nvidia-utils: /usr/lib/libnvidia-egl-xcb.so występuje w systemie plików (właścicielem jest egl-x11)
nvidia-utils: /usr/lib/libnvidia-egl-xcb.so.1 występuje w systemie plików (właścicielem jest egl-x11)
nvidia-utils: /usr/lib/libnvidia-egl-xcb.so.1.0.0 występuje w systemie plików (właścicielem jest egl-x11)
nvidia-utils: /usr/lib/libnvidia-egl-xlib.so występuje w systemie plików (właścicielem jest egl-x11)
nvidia-utils: /usr/lib/libnvidia-egl-xlib.so.1 występuje w systemie plików (właścicielem jest egl-x11)
nvidia-utils: /usr/lib/libnvidia-egl-xlib.so.1.0.0 występuje w systemie plików (właścicielem jest egl-x11)
nvidia-utils: /usr/share/egl/egl_external_platform.d/20_nvidia_xcb.json występuje w systemie plików (właścicielem jest egl-x11)
nvidia-utils: /usr/share/egl/egl_external_platform.d/20_nvidia_xlib.json występuje w systemie plików (właścicielem jest egl-x11)
Wystąpiły błędy, nie zaktualizowano żadnego pakietu.
We actually don't 100% know that because the LTS kernel could have picked up the bogus commit, but if and since the LTS kernel was also affected it actually seems more likely the systemd 257 update, which can be downgraded w/o having to deal w/ the nvidia packages.
sudo downgrade systemd-libs systemd
wczytywanie pakietów…
ostrzeżenie: dezaktualizowanie pakietu systemd (257.1-1 => 256.9-1)
ostrzeżenie: dezaktualizowanie pakietu systemd-libs (257.1-1 => 256.9-1)
rozwiązywanie zależności…
szukanie sprzecznych pakietów...
błąd: nie udało się przygotować transakcji (nie udało się rozwiązać zależności)
Until we figured whom to blame, this is your problem, your responsibility and PSA:
You'll have to keep moving your ass to do something about that - and stomping your foot real hard and yelling demands into the internet generally won't cut it in life.
I think this is not just my problem but an issue for thousands of other Arch users. I’m not just sitting idly by—I’m reporting the problem and helping to solve it as much as I can. On the other hand, this is not my test machine but a system I work on every day, so I can’t turn it into a testing ground either.
Offline
Remove from /etc/pacman.conf and IgnorePkg or NoUpgrade entries added by `downgrade` then:
# pacman -Rdd egl-x11
# pacman -Syyuu
Offline
The system has been restored to the date 01.12.2024, and the camera is functioning correctly. Should I investigate further and change the date to another one? If so, to which date?
Offline
Try 09.12.2024 which should be the day before systemd was updated to 257.
Offline
Update to 09.12.2024 - the camera is working. Next update to 10.12.2024, and the camera stops working. However, interestingly, there was no systemd update here. Details below:
[2025-01-01T15:42:25+0100] [ALPM] upgraded appstream (1.0.3-1 -> 1.0.4-1)
[2025-01-01T15:42:25+0100] [ALPM] upgraded mesa (1:24.3.1-2 -> 1:24.3.1-3)
[2025-01-01T15:42:25+0100] [ALPM] upgraded appstream-qt (1.0.3-1 -> 1.0.4-1)
[2025-01-01T15:42:25+0100] [ALPM] upgraded json-glib (1.10.0-1 -> 1.10.6-1)
[2025-01-01T15:42:25+0100] [ALPM] upgraded lib32-libelf (0.192-1 -> 0.192-2)
[2025-01-01T15:42:25+0100] [ALPM] upgraded lib32-mesa (1:24.3.1-2 -> 1:24.3.1-3)
[2025-01-01T15:42:26+0100] [ALPM] upgraded linux (6.12.3.arch1-1 -> 6.12.4.arch1-1)
[2025-01-01T15:42:27+0100] [ALPM] upgraded linux-lts (6.6.63-1 -> 6.6.64-1)
[2025-01-01T15:42:27+0100] [ALPM] upgraded nvidia (565.77-4 -> 565.77-5)
[2025-01-01T15:42:27+0100] [ALPM] upgraded nvidia-lts (1:565.77-2 -> 1:565.77-3)
[2025-01-01T15:42:27+0100] [ALPM] upgraded python-numpy (2.1.3-1 -> 2.2.0-1)
[2025-01-01T15:42:27+0100] [ALPM] upgraded python-threadpoolctl (3.4.0-2 -> 3.5.0-1)
pacman -Ss systemd
core/systemd 256.9-1 [zainstalowano]
EDIT:
I'm currently testing up-to-date system from today, 01.01.2025, with kernel 6.6.63-1-lts. The camera is functioning correctly. Therefore, we can confidently conclude that the change in the kernel from December 10.12.2024, is responsible for the malfunctioning camera.
Last edited by gerwazy (2025-01-01 16:19:04)
Offline
So the changes would be in
[2025-01-01T15:42:26+0100] [ALPM] upgraded linux (6.12.3.arch1-1 -> 6.12.4.arch1-1)
[2025-01-01T15:42:27+0100] [ALPM] upgraded linux-lts (6.6.63-1 -> 6.6.64-1)
?
https://github.com/gregkh/linux/commit/ … 96f54feeba yells at me…
sudo journalctl -b | grep -E 'Found.*Unit.*with.*ID'
you can offset "… -b -1… " to check the journal of the previous boot.
Did you alread ypost a complete journal of an affected and/or good boot?
Offline
sudo journalctl -b -1 | grep -E 'Found.*Unit.*with.*ID'
sty 01 15:43:24 Linux kernel: usb 3-2: Found multiple Units with ID 5
sudo journalctl -b | grep -E 'Found.*Unit.*with.*ID'
Offline
Aaaaaand… we have a winner. Most likely.
I'll now yell into the internet and wish, nay, "demand!" somebody™ compile a kernel with that patch reverted.
Difference is, there's a good change loqs already has the compilation finished
In doubt you'll have to compile a kernel to secure the bet, so we can then yell at the patch author
Offline
I have already posted this error in one of my previous posts.
Offline
So?
Nobody here knows magically what commit introduced what error in which device - or whether it's actually relevant w/o any contrast.
Offline
Offline
hello guys, same problem also for me. Do this bug has been reported? Will it be fixed next arch update so? Im so noob, don't have competence to compile that and install that.
Entia non sunt multiplicanda praeter necessitatem.
Offline
You don't have to compile anything, but it would certainly helpful if you could install and test the kernel loqs provided in #39 to confirm the offending commit so this actually can be reported to the author.
Offline
Do this bug has been reported?
Does not appear to be.
Will it be fixed next arch update so?
Very unlikely.
Im so noob, don't have competence to compile that and install that.
Download linux-6.12.7.arch1-1.1-x86_64.pkg.tar.zst and linux-headers-6.12.7.arch1-1.1-x86_64.pkg.tar.zst then from the download location execute:
# pacman -U linux-6.12.7.arch1-1.1-x86_64.pkg.tar.zst linux-headers-6.12.7.arch1-1.1-x86_64.pkg.tar.zst
Last edited by loqs (2025-01-01 21:21:24)
Offline
I installed those pkg, loqs and the webcam is finally fixed in discord!!!! and I can get it back working also on obs!! thanks so much!
BTW, If i will run an -Syu in future and there will be a new kernel update those fixes will vanish so?
Entia non sunt multiplicanda praeter necessitatem.
Offline
BTW, If i will run an -Syu in future and there will be a new kernel update those fixes will vanish so?
Yes unless/until there is a fix from upstream or an alternative solution is found.
Offline
prove me wrong but If we don't report the problem it will never be implemented in new kernel update? p.s. thanks again for the files loqs, at first I was thinking of not trusting you, cuz ur not admin or so, and I was thinking you could inject malicious things xDxD but ur 18kpost had me trusting you xD lv u btw
Entia non sunt multiplicanda praeter necessitatem.
Offline
You can report it upstream using https://bugzilla.kernel.org possibly under Product v4l-dvb Component webcam or on the linux-media mailing list. See also Reporting issues/Reporting regressions.
Offline
Not working for me.
sty 01 23:21:26 Linux sddm[733]: Failed to read display number from pipe
sty 01 23:21:26 Linux sddm[733]: Display server stopping...
I will stay with the old working kernel version for now and wait for a fix.
Offline
@gerzway, you'll need an nvidia kernel package for 6.12, afaict you're using linux-lts and only have nvidia-lts installed?
You might even have to use nvidia-dkms, but you can confirm the behavior of the kernel w/o a GUI by ensuring that /dev/video0 gets automatically generated.
Offline
Of course, I have the NVIDIA package installed
pacman -Ss nvidia
extra/nvidia 565.77-9 (54.3 MiB 54.4 MiB) (Installed)
NVIDIA kernel modules
Working well with kernel linux 6.12.7.arch1-1 but not working with linux 6.12.7.arch1-1.1
Last edited by gerwazy (2025-01-02 11:58:19)
Offline
Ah, no - with the altered build number you'll have to use nvidia-dkms to get the nvidia modules for that kernel (you can also try to symlink the ones from 6.12.7.arch1-1 and force-load them)
However, as mentioned, the only critical information here is whether /dev/video0 appears again w/ the patched kernel. The nvidia situation is completely unrelated to that.
Offline