Tumbleweed+Lenovo Legion 7 : Black screen at boot after upgrading to kernel 5.14+

Good day.

I have OSTW running on my kaptop. I don’t have proprietary drivers installed.

Here is inxi -F info.

After I upgraded to kernel 5.14+ using sudo zypper refresh && sudo zypper dup , the screen now goes black at boot, I can’t switch to terminal. alt+ctrl+del is working.

When I boot using the previous kernel everything works fine.

What could possibly go wrong? What logs should I check? Has anyone experienced anything like this?

Regards.

There are more things that can go wrong when there are multple GPUs. I suggest you run sudo zypper al kernel-default before doing any additional upgrades. I looked for an existing bug report about this and found none, but there are several bugs about other things that no longer work in 5.14. I’ve seen 5.14 complaints elsewhere too. Just keep using 5.13 until the cause is found and corrected. The lock will prevent removal of 5.13. You can remove it with zypper rl kernel-default when it’s no longer needed.

Note on TW you should only use zypper dup.

I doubt that zypper up will hurt as long as you do dup second but you never know…

I always use zypper dup :wink:

Anyways, the latest updats haven’t fixed anything. Still black screen…

We should have answered this before now. If booted to 5.14 and you can login on a vtty (Ctrl-Alt-F3), then cat /var/log/Xorg.0.log | susepaste. If you can’t and must boot 5.13, then cat /var/log/Xorg.0.log.old | susepaste from booting immediately into 5.13 after last booting 5.14. IOW, we need to see a log from a 5.14 boot. Also from a 5.13 boot immediately after a 5.14 boot: sudo journalctl -b -1 | grep aile. If the output is modest or less, paste it here. Otherwise, susepaste it.

cat /var/log/Xorg.0.log | susepaste

I’m getting “cat: /var/log/Xorg.0.log.old: No such file or directory”

sudo journalctl -b -1 | grep aile

https://paste.opensuse.org/36377939

As the susepaste log will expire soon, let me post what I think is relevant:

сен 25 00:47:22 wtf kernel: tsc: Marking TSC unstable due to check_tsc_sync_source failed
сен 25 00:47:34 wtf kernel: nouveau 0000:01:00.0: disp: init failed, -16
сен 25 00:47:34 wtf kernel: nouveau 0000:01:00.0: init failed with -16
сен 25 00:47:34 wtf kernel: nouveau: systemd-udevd[445]:00000000:00000080: init failed with -16
сен 25 00:47:34 wtf kernel: nouveau: DRM-master:00000000:00000000: init failed with -16
сен 25 00:47:34 wtf kernel: nouveau: DRM-master:00000000:00000000: init failed with -16
сен 25 00:47:34 wtf kernel: nouveau 0000:01:00.0: DRM: Client resume failed with error: -16
сен 25 00:47:34 wtf kernel: nouveau 0000:01:00.0: DRM: resume failed with: -16
сен 25 00:47:45 wtf systemd[1]: haveged.service: Failed with result 'exit-code'.
сен 25 00:47:45 wtf systemd-modules-load[973]: Failed to find module 'bbswitch'
сен 25 00:47:46 wtf gnome-shell[1787]: Failed to open gpu '/dev/dri/card0': GDBus.Error:org.freedesktop.DBus.Error.InvalidArgs: Invalid argument
сен 25 00:47:46 wtf kernel: thermal thermal_zone0: failed to read out thermal zone (-61)
сен 25 00:47:47 wtf kernel: ACPI: \: failed to evaluate _DSM (0x1001)
сен 25 00:47:47 wtf kernel: ACPI: \: failed to evaluate _DSM (0x1001)
сен 25 00:47:47 wtf kernel: ACPI: \: failed to evaluate _DSM (0x1001)
сен 25 00:47:47 wtf kernel: ACPI: \: failed to evaluate _DSM (0x1001)

So not only problems with the graphics but also problems reading out ACPI.

I have secure boot disabled. Is this relevant in any way?

I also use encryption.

PS I experimented with other distros and here is my experience so far

  1. Ubuntu 20.04 can’t install mainline kernel 5.14 - requires new dependency

  2. Ubuntu 21.04 can’t install mainline kernel 5.14 - requires even newer dependency

I could install it via https://raw.githubuser…m/k3dar/kumk/main/kumk script it uses some workarounds but after reboot- black screen

  • I suppose compiling from source is another option but this was my first experience doing it and I kept getting debian canonical cert error despite tweaking my config file
  1. OpenSUSE TW works fine on kernel 5.13 but after upgrading to 5.14 boots into a black screen

The latest offline installer boots into a black screen. ;(

  1. Manjaro (I have never installed pure Arch so I decided to go the easy way coz I didn’t have much time) - it updated to 5.14 and worked without any issues. I didn’t see any errors in the logs.

You’re using NVIDIA GA104M with nouveau driver + AMD Cezanne with amdgpu driver.
Your Nvidia card (chip) is too new for nouveau drivers, I recommend using Nvidia proprietary drivers.
Also use X11, not Wayland.

For some unknown purpose Nvidia card is 1st, AMD is 2nd (Lenovo engineers detected). So ILL kernel tries to use Nvidia for output and fails with it. BIOS update may help.
Try to load Linux with AMD as 1st. IDK how to do this.
Or wait for kernel update to get newer nouveau driver. Or use modesetting driver.

Try to boot into command prompt using kernel parameter “3”.

Message: Advanced graphics data unavailable in console for root.

Use inxi as user in DE.

With

inxi -aFz

you’ll get more info.

It’s possible Xorg.0.log is in ~/.local/share/xorg/ instead. Please provide output here from the following:

ls -l /var/log/Xorg*	~/.local/share/xorg/Xorg*

Sorry I took so long to reply, but I was unexpectedly in hospital 4 nights.