Upgrade to leap 16.0 with opensuse-migration-tool hang after reboot

pier_andreit · January 27, 2026, 3:32pm

on my laptop 1=ASUS vivobook pro N552VW-FY204T (2016) running 15.6 KDE and dual boot with windows
CPU is x86-64-v3 (supported, searched)
GPU is GTX960M is currently supported by the Nvidia 580.126.09 driver.

used opensuse-migration-tool version 20260106.d2cfd39-lp156.20.1 by yast
executed sudo LANG=C opensuse-migration-tool --dry-run
selected migration to leap 16.0
selected all the repositories to remove deselected only chrome
selected keepselinux
selected 20_ia32
selected pulsepipewire

everything went smoothly

rebooted
executed sudo LANG=C opensuse-migration-tool
selected migration to leap 16.0
selected all the repositories to remove deselected only chrome
selected keepselinux
selected 20_ia32
selected pulsepipewire
after upgrade it asks me what to execute after upgrade
selectedd selinux
selected ia_32
rebooted executing sudo shutdown -r now

boot ok with grub with dual boot options as previously
hang at
booting ‘opensuse leap 16.0’
loading linux 6.12.0-160000.8-default
loading initial ramdisc …

tried also selecting grub options boot recovery mode but doesn’t works

what I have to do to have my leap 16.0 working??

Sauerland · January 27, 2026, 3:37pm

Do you have any entry with

selinux=1

in your boot line?
set it to

selinux=0

and start with F10

pier_andreit · January 27, 2026, 3:54pm

done pressing “e”, but hangs at the same point
there are also
security=selinux
enforcing=1
ia32_emulation=1

Sauerland · January 27, 2026, 4:12pm

delete them this is a tempary solution

Also set selinux=0 and add 3 at the end to go to Runlevel 3 (not in graphics Mode)
Do you get a login?

pier_andreit · January 27, 2026, 6:47pm

deleted security=selinux and enforcing=1
set selinux=0 and added 3 at the end of the line where selinux=0 and the others are
it hangs at the same point

deano_ferrari · January 27, 2026, 8:09pm

This points to an initramfs / NVIDIA breakage. You could try booting with the nomodeset kernel boot option. Does that help?

If it still hangs at the loading initial ramdisk stage, this Fedora Wiki might be useful to help gain further diagnostic information…
https://fedoraproject.org/wiki/How_to_debug_Dracut_problems

pier_andreit · January 27, 2026, 9:33pm

the line is:
linux /boot/vmlinuz-6.12.0-160000.8-default root=UUIDefa67809-7971-42b8-94f7-93179f4121f4 splash=silent resume=/dev/disk/by-uuid/2e52215c-5454-4a37-9150-224d4319eac5 preempt=full mitigations=auto quiet rd.driver.blacklist=nouveau security=selinux enforcing=1 selinux=1 ia32_emulation=1

I tested these

added nomodeset at the end
…quiet rd.driver.blacklist=nouveau security=selinux enforcing=1 selinux=1 ia32_emulation=1 nomodeset
deleted security=selinux enforcing=1 and set selinux=0 and nomodeset
…quiet rd.driver.blacklist=nouveau selinux=0 ia32_emulation=1 nomodeset
deleted security=selinux enforcing=1 and set selinux=0 and nomodeset and 3
…quiet rd.driver.blacklist=nouveau selinux=0 ia32_emulation=1 nomodeset 3
deleted rd.driver.blacklist=nouveau security=selinux enforcing=1 and set selinux=0 and nomodeset and 3
…quiet selinux=0 ia32_emulation=1 nomodeset 3

but it hangs at the same point

then I tried from your link:
Identifying your problem area

Remove rhgb and quiet from the kernel command line
there wasn’t any rhgb command and removed quiet but hangs at the same point
Add rd.shell to the kernel command line. This will present a shell should dracut be unable to locate your root device
removed quiet and added rd.shell but no shell was presented
Add rd.shell rd.debug log_buf_len=1M to the kernel command line so that dracut shell commands are printed as they are executed
not tested becouse no shell

4)Inspect the system logs:
less /run/initramfs/rdsosreport.txt
journalctl -a
dmesg
less /run/initramfs/init.log
not tested becouse no shell

deano_ferrari · January 28, 2026, 12:22am

Thanks for the update. From what you describe, it looks like the kernel hangs before it can execute the initramfs (no shell, no logs). It’s possible the initramfs wasn’t regenerated for the Leap 16 kernel during the migration, and may still reference modules from the old Leap 15.6 kernel.

The next step might be to boot from a live distro (eg TW)…

Once you have a shell, identify the partitions for your Leap 16 install (/ and, if separate, /boot). Then mount them:

mount /dev/<root-partition> /mnt
# If you have a separate /boot
mount /dev/<boot-partition> /mnt/boot

Bind the virtual fiesystems…

sudo mount --rbind /dev /mnt/dev
sudo mount --rbind /proc /mnt/proc
sudo mount --rbind /sys /mnt/sys
chroot /mnt /bin/bash

From there try to build the initramfs with dracut -f or specify the kernel explicitly with dracut -f --kver 6.12.0-160000.8-default.

Once that completes, exit the chroot, unmount, and try booting the Leap 16 install again.

knurpht · January 28, 2026, 2:23am

I always do this by
mount -a
right after the chroot.

susejunky · January 28, 2026, 5:04am

deano_ferrari:

Once you have a shell, identify the partitions for your Leap 16 install (/ and, if separate, /boot). Then mount them:
mount /dev/<root-partition> /mnt
# If you have a separate /boot
mount /dev/<boot-partition> /mnt/boot
Bind the virtual fiesystems…
sudo mount --rbind /dev /mnt/dev
sudo mount --rbind /proc /mnt/proc
sudo mount --rbind /sys /mnt/sys
chroot /mnt /bin/bash

Will this work even if OP is UEFI-booting and using btrfs for “/”?

So far I did chroot into btrfs like this:

> su -
# mount -o subvolid=<ID of @> /dev/<root-partition> /mnt
# mount -o subvolid=<ID of @/var> /dev/<root-partition> /mnt/var
# mount -o subvolid=<ID von @/boot/grub2/x86_64-efi> /dev/<root-partition> /mnt/boot/grub2/x86_64-efi
# mount -o subvolid=<ID von @/boot/grub2/i386-pc> /dev/<root-partition> /mnt/boot/grub2/i386-pc
# mount /dev/<ESP> /mnt/boot/efi
# for i in dev proc run sys; do mount -o rbind /$i /mnt/$i ; done
# chroot /mnt /bin/bash
# mount -a

To identify the subvolume IDs I did

# mount -t btrfs /dev/<root-partition> /mnt
# btrfs subvolume list /mnt
# umount /mnt

and took a note of the subvolume IDs.

So is all this not necessary?

deano_ferrari · January 28, 2026, 5:20am

Well, yes…short answer is it will work with UEFI + btrfs…they may need to adjust instructions to suit.

For example, the root filesystem…
mount -o subvol=@ /dev/<btrfs-partition> /mnt
Only need the root subvolume and /boot partition if separate.

Also

# If using UEFI, mount ESP
mount /dev/<ESP> /mnt/boot/efi

pier_andreit · January 28, 2026, 5:29am

by the way, I’m with ext4 filesystem on / and /home

pier_andreit · January 28, 2026, 5:31am

used TW openSUSE-Tumbleweed-KDE-Live-x86_64-Snapshot20260126-Media.iso
written on the USB drive with suse studio imagewriter
booted TW slowly but black screen with mouse corsor in graphic mode
pressed ctrl alt F3 to get commandline login
at localhost login logged as root seems regularly

this is not the true output but copyed and elaborated

fdisk -lx
Device         Start        End   Sectors Type-UUID                            UUID                                 Name
/dev/sda1       2048    1026047   1024000 EBD0A0A2-B9E5-4433-87C0-68B6B72699C7 AEE9C626-69B9-433E-B849-994A886011BB nEFI
/dev/sda2    1026048    1230847    204800 C12A7328-F81F-11D2-BA4B-00A0C93EC93B 214D1A7A-AB7B-4C0F-801E-5D8935C9AB46 EFI system partition
/dev/sda3    1230848    1263615     32768 E3C9E316-0B5C-4DB8-817D-F92DF00215AE 69210CD3-6B32-458A-845C-065969CDD4E5 Microsoft reserved partition
/dev/sda4    1263616  122345471 121081856 EBD0A0A2-B9E5-4433-87C0-68B6B72699C7 28904F0D-9574-40A6-8DEB-03A9E0692ECD Basic data partition
/dev/sda5  122345472  123903999   1558528 DE94BBA4-06D1-4D40-A16A-BFD50179D6AC EC5D5565-75C0-4149-B77C-0CB1841F5451
/dev/sda6  123906048  246786047 122880000 0FC63DAF-8483-4772-8E79-3D69D8477DE4 715E40B1-4154-4C49-9593-CD05149CBFA2 nSUSE1
/dev/sda7  246786048  308226047  61440000 0FC63DAF-8483-4772-8E79-3D69D8477DE4 57B9B075-B304-4206-87D4-39C054BD0789 nHOME1
/dev/sda8  308226048  373762047  65536000 0657FD6D-A4AB-43C4-84E5-0933C84B4F4F C06A39E6-1587-40B7-9FC7-28BEBA0BFFD2 nSWAP
/dev/sda9  373762048 1000214527 626452480 EBD0A0A2-B9E5-4433-87C0-68B6B72699C7 0A8C3B0D-BEA0-4427-8659-728D8428AF57 nDATI

I identify / as /dev/sda6 and boot as /dev/sda2 as /boot

this is not the true output but writte by me
dracut -f
failed
[numbers]dracut f3043] No '/dev/logger' or 'logger' included for syslog logging
realpath /lib/modules/6.18.7-1-default no such file or directory
dracut[F]: cannot find module directory
[numbers]dracut f3043] cannot find module directory
dracut[F]: and --no-kernel was not specified
[numbers]dracut f3043]  and --no-kernel was not specified

dracut -f --kver 6.12.0-160000.8-default
dracut [E] No '/dev/logger' or 'logger' included for syslog logging
cp: error writing '/boot/initrd-6.18.7-1-default': no space left on device
dracut[F]: creation of '/boot/initrd-6.18.7-1-default' failed
[numbers]dracut f3110] creation of '/boot/initrd-6.18.7-1-default' failed

so I tried to boot with another kernel
at boot I selected other options for opensuse 16.0
selected the last linux 6.4.0-150600.23.81-default
it booted but hanged at another point
pressed ESC to see what happens and there was something abot selinux (by the way, why thw splas=silent didn’t worked?)
edited the line with selinux=0

and it booted!!!

plasma is without tray bar

and I discovered that the boot partition wasn’t separate

so I made your suggestion dracut with 16.0
open konsole as supeuser and

Password: 
localhost:~ # dracut -f
libkmod: ERROR: kmod_module_parse_depline: ctx=0x55c48384fe00 path=/usr/lib/modules/6.4.0-150600.23.81-default/weak-updates/updates/nvidia-modeset.ko error=No such file or directory
realpath: weak-updates/updates/nvidia-modeset.ko: No such file or directory
realpath: weak-updates/updates/nvidia-drm.ko: No such file or directory
realpath: weak-updates/updates/nvidia-uvm.ko: No such file or directory
realpath: weak-updates/updates/nvidia.ko: No such file or directory
dracut[F]: installkernel failed in module kernel-modules-extra
localhost:~ # dracut -f --kver 6.12.0-160000.8-default
localhost:~ #

rebooted
edited the line with selinux=0
but it hangs at the same time

it is a good step, have you other suggestion to boot with the normal kernel?

deano_ferrari · January 28, 2026, 5:53am

That reads like progress, albeit with an old kernel left in place.

I think best to boot with that kernel, and remove nvidia packages completely…
zypper rm '*nvidia*'

List the installed kernels…
ls /usr/lib/modules
and rebuild initramfs for the newest one present eg
dracut -f --kver 6.12.0-160000.8-default

Regenerate GRUB…
grub2-mkconfig -o /boot/grub2/grub.cfg
and reboot.

If the OS boots ok with the new kernel, the system is fixed, and NVIDIA packages can be reinstalled.

pier_andreit · January 28, 2026, 11:28am

manythanks,
to boot I have also to set selinux=0 in the boot line, is there a way to solve also this or I have to remove nvidia and do all the rest before?

ElectrodelessBird · January 28, 2026, 11:42am

i am commend you reship the system, try leap 15.4,
because my laptop is not loaded to install the 16, another two old computer is same trouble.
my old computer installed the 15.4-version , well done !

Sauerland · January 28, 2026, 11:53am

Dead Distribution, use for the rest of lifetime Leap 15.6 or install Slowroll/Tumbleweed.

ElectrodelessBird · January 28, 2026, 12:41pm

i don’t like the Tumbleweed version, because my all computer is not successfully to run it !
my laptop is new cpu , it 's not work at soud and wife .

knurpht · January 28, 2026, 3:01pm

Then you cannot run openSUSE. Leap 15.4 is dead, and will never ever be revived. Nor will 15.5 and in a couple of months 15.6

deano_ferrari · January 28, 2026, 6:11pm

Stabilise the system first. NVIDIA is currently breaking dracut and initramfs generation.