Upgrading to 15.3 - Kernel panic - no syncing - no working init found

I’ve been having most serious issues upgrading OpenSUSE 15.2 to 15.3 probably in whole 20+ years with SuSE.

After postponing upgrade several times finally it seemed that the installation is fixed to work on my machine and did not complain about anything prior to launching update. And not only on my primary machine, I got the same error message in the installer on my secondary machine (both AMD, one 8 core, one 4 core). However first attempt has frozen at 11%. The 15.2 working, but broken (no internet and many minor troubles). I had to rebuild RPM database. I relaunched it (newtwork install) and it went well. I had to leave at around 60% and when came back (about the time it supposed to finish) I had the error on my screen:

Failed to execute /init (error -2)
Kernel panic - not syncing: No working init found. Try passing init= option to kernel. See Linux Documentation admin.rst for guidance

Any advice appreciated.

Before upgrading I always try out a live version first. Does a live version work on your machine? http://download.opensuse.org/distribution/openSUSE-current/live/

Are you able to choose to boot a prior kernel from a Grub menu? No working init can result from an initrd that didn’t get built, or got defectively built.

Can you successfully use the installation media to boot the installed system?

Downloading to test.

Well maybe a good idea for the future :-). Usually the install and system is working better and better with minor updates, so I didn’t expect that. And I simply didn’t think of this option. Lesson learned.

Still, I’ll have to find a solution or downgrade to 15.2.

Thank you for trying to help. Here are requested details:

  • No, I believe I’m unable to go to grub menu (unless there’s some hidden option), it’s a dedicated Linux machine.
  • No, when I use installation media to boot the installed system it goes into the same error. (I believe I tried that, will double check).

The Live KDE OpenSUSE 15.3 seems to work fine! Booted from a USB flash drive.

Just confirming, “Boot from hardrive” option from the network installation started from an USB stick does result in the very same error.

I tried to reinstall it again, this time I succeeded to launch it from an external hardrive. But since the 15.3 in already installed, it only updated some 20 packages and I think it didn’t do anything to the system files. The error persists.

I don’t remember, since every PC here is multiboot, but I think to force the Grub menu and choose a prior kernel you can hit the shift key as POST completes. If you can’t make that work, try the ESC key.

You were quite right about the option:

> GRUB_HIDDEN_TIMEOUT
Waits the specified number of seconds for the user to press a key. During the period no menu is shown unless the user presses a key. If no key is pressed during the time specified, the control is passed to GRUB_TIMEOUT. GRUB_HIDDEN_TIMEOUT=0 first checks whether Shift is pressed and shows the boot menu if yes, otherwise immediately boots the default menu entry. This is the default when only one bootable OS is identified by GRUB 2.

I was unable to capture looking at the start anything, so I recorded a video and then found, that for a fraction of a second a message is flashed:

GRUB something
Please press 't' to show the boot menu on this console

However, I’m unable to press the T, it seems the loop may actually 0 second long. I’ll might need to find out how to change the configuration. The question is, if this is related to the problem.

You need to try booting the prior kernel, from which the kernel panic can probably be fixed. You just need to experiment with the keystroke timing to get to the Grub menu. Once you’ve managed to boot, consider to reconfigure /etc/default/grub so that the menu always appears.

Yep, I already found the setting and changed and managed to display it for couple of seconds. However, it does NOT respond to ‘t’ press or shift or Esc. I tried to further modify the defaults, but it doesn’t help.

Looks like you’ll need to try to fix this via chroot (as explained on https://doc.opensuse.org/documentation/leap/startup/html/book-startup/cha-trouble.html section 17.5.2.3 Accessing the installed system).

I would suggest downloading the full installer and running that. First attempt, try choosing the option to Upgrade. Don’t allow online repos during this attempt, go with what comes with the installer. This should replace any corrupted or incorrect files. Another thing: Watch whether you are using UEFI or Legacy. If your BIOS is set to Legacy, and you boot the UEFI part of the installer, you could possibly run into problems, and vice versa. Just make sure you match what your BIOS is set to. (On my machine, I had set the BIOS to Legacy, but in the advanced boot options, it still gave me the choice of booting the USB key as Legacy or UEFI. Of course, I chose the Legacy device.)

I’ll check that out, thank you for suggestion.

I did the full installer - problem from the start was how to fire it up, since Ihave tons of 4GB flash disks, but no functioning larger. Finally I got it up fired from external HD via boot from a matching network installation on the USB flash disk. I have UEFI BIOS.

I think it all may come to boot loader part in the header of the HD not being updated. So I’ll start from the most simple and standard thing - I will reinitiate the HD install and will try to modify something in the boot configuration to force the installer to modify and save it.
If that doesn’t work, I’ll go through the list of your suggestions. Thank you for the moment.

OK, forcing to rewrite Grub settings via the installer FIXED booting, but it only goes into tty1. There are no errors, apparently. Plasma is installed (I checked via YaST). So I’ll try to run installer once again and make sure KDE is configured. Hopefully it will bring it to life.

I can start the system with KDE by changing the default Display Manager to something different but default SDDM, like XDM or LightDM. But my 4K screen is at horrible 1024x768 resolution, so the problem has just evolved again, not being solved :-(. Now looking for solution for that. I think it’s all caused by Wayland, that thing is a trouble maker…
If I wasn’t sick, I wouldn’t have time to play with this and would have to revert to 15.2.

Well, it seems the problem is somewhere with the nVidia drivers:

**linux-e53g:/home/**me** #** **lsmod** | grep nvidia
i2c_**nvidia**_gpu         20480  0

**linux-e53g:/home/**me** #** **nvidia-xconfig**

WARNING: Unable to locate/open X configuration file.

Package xorg-server was not found in the pkg-config search path.
Perhaps you should add the directory containing `xorg-server.pc'
to the PKG_CONFIG_PATH environment variable
No package 'xorg-server' found
New X configuration file written to '/etc/X11/xorg.conf'

**linux-e53g:/home/me #** **nvidia-settings**

ERROR: NVIDIA driver is not loaded


ERROR: Unable to load info from any available system

**linux-e53g:/home/**me** #** 


Not sure, whether I should keep it in one thread or start new one, but it’s still the same fight with the 15.3 upgrade, though about 20th issue.

As for this latest problem, I already tried:

  • reinstall the nVidia drivers from KDE/YaST. No change.
  • remove all the nVidia drivers and reinstall them from downloaded nVidia recommended package from tty1. Failed to build the driver.
  • KDE settings see only 1 screen size (1024x768px), so it cannot be changed, I think it runs in some fail-safe mode, because even default driver should offer some basic resolutions, like FullHD. Even the installer can run in 4K on my scren!

some more information…

I uninstalled all nVidia packages and tried to install only the driver and dependants. Now it look like this.

linux-e53g:/home/me **#** **zypper se -si nvidia**

S  | Name                      | Type    | Version                      | Arch         | Reposi-> 
---+---------------------------+---------+------------------------------+--------------+--------- 
i  | nvidia-computeG05         | package | 470.74-lp153.44.1            | x86_64       | nVidia 
i+ | nvidia-gfxG05-kmp-default | package | 470.74_k5.3.18_57-lp153.44.1 | x86_64       | nVidia 
i  | nvidia-glG05              | package | 470.74-lp153.44.1            | x86_64       | nVidia 
i  | x11-video-nvidiaG05       | package | 470.74-lp153.44.1            | x86_64       | nVidia 

**linux-e53g:/home/me #** **/sbin/lspci -nnk | grep -EiA3 'vga|3d|display'** 
02:00.0 USB controller [0c03]: Advanced Micro Devices, Inc. [AMD] 400 Series Chipset USB 3.1 XHCI Controller [1022:4[b]3d5] (
rev 01) 
       Subsystem: ASMedia Technology Inc. Device [1b21:1142] 
       Kernel driver in use: xhci_hcd 
       Kernel modules: xhci_pci 
-- 
07:00.0 **VGA** compatible controller [0300]: NVIDIA Corporation TU104 [GeForce RTX 2070 SUPER] [10de:1e84] (rev a1) 
       Subsystem: Gigabyte Technology Co., Ltd Device [1458:4001] 
       Kernel modules: nouveau 
07:00.1 Audio device [0403]: NVIDIA Corporation TU104 HD Audio Controller [10de:10f8] (rev a1)

You never mentioned nVidia before. That might have helped.

nVidia makes it a whole new ballgame, and I am in a different league.

Good luck, the others can help you here.