Solus Updates for Week 10, 2025

pomon

AlphaElwedritsch The /boot partition does not mount at system startup (that's the case in my situation). Unless you mounted it manually later. Therefore, there shouldn't be any issues performing operations like resizing this partition. However, if any problems arise, then yes, you should definitely use Solus Live.

snowee As I mentioned earlier, any operation on system files, including the /boot partition, can lead to system corruption, so I recommend extra caution.
I removed unnecessary entries in /boot/EFI/com.solus-project/, I ran sudo clr-boot-manager update, and restarted. Everything is fine.

minh

Just to update the issue I have since the update of Linux Kernel in previous post.
https://discuss.getsol.us/d/11355-solus-updates-for-week-6-2025/18.
During the weekend, I have chance to try the Plasma on beside current Xfce that I have been used for some months. Just to check if reinstall fix this issue.

What I found is not problem of Nvidia to combine with the latest kernel. It's something between latest kernel vs the /boot partition on my M2 (nvme) instead.

I have tried to boot Plasma or Xfce from the /boot partition setup on SSD. I could boot them both (use latest kernel).
And then I tried to boot them from the old /boot partition setup on M2, they both having same error.

Error preparing initrd: Device error
start_image () returned Device Error

So my conclusion is that something to do with latest Linux Kernel vs /boot partition on M2 drive. It works fine with SSD. And nothing to do with Nvidia drive. Hope it helps someone to troubleshoot this issue.

Phiqu

I also reinstalled Solus xfce now with Week10 besides CachyOS on my nvme.
I still get the:
Error preparing initrd: Device error start_image () returned Device Error
But Solus finally booting again with current kernel and current nvidia driver for my 1070.

Edit.
With "Solus Linux Bootloader" I get the Device Error from above. When I choose the "Linux Boot Manager" as priority boot option for Solus it is booting without any error.

brent

@BuzzPCSOS @AlphaElwedritsch

actually pretty common to use PSA at this joint.
But when i did go online, like you, I got nowhere.
I will spell it out next time I have a fix for someone, noted. Common at Solus, uncommon on Google. Makes sense.

Animpacz

pomon

Switching (set-kernel) to the lates kernel prompts a similar error message demanding more space.

As I always want to use the latest kernel available I removed the package linux-lts.

Also I removed all lts and older concurrent kernels with clr-boot-manager together with the entries under
'/boot/EFI/com.solus-project/'.
After cleaning up and a successful execute sudo clr-boot-manager update and sudo usysconf run -f the latest kernel was selected and I could reboot into the new kernel.

The output of df -h | grep boot is now:

/dev/nvme0n1p1 489M 183M 306M 38% /boot

..and from sudo du -d1 -h /boot is now:

183M /boot/EFI
24K /boot/loader
183M /boot

Your instruction were clear and worked well, it looks good now and I got a better understanding of it. Thank you a lot!

Harvey

Cherry-picked vivaldi-stable 7.1.3570.60

WetGeek

Now downloading the Vivaldi upgrade at 340 KB/s on a fast i7 laptop, and has been at it for several minutes. Something seems very familiar about that.

Harvey

WetGeek

Probably just the connection between you and the server, as it downloaded at 15.03MB/s for me. It is also no where near the file size limit issue that caused issues with our CDN recently.

AlphaElwedritsch

WetGeek
vivaldi-stable-7.1.3570.60-284-1-x86_64.eopkg (147.1 MB)100% 49.72 MB/s [00:00:00]
Just a few seconds here...

786psi

Hello,

The update did not go well for me.
Lightdm did not want to start. I ran the eopkg check/reinstall command from the troubleshooting page, it gave no output.
I am on the current branch, checking uname vs kernel command shows that I was still running on 6.12.12 hence why.
I use efi/esp, clr-boot-manager update tells me there's no space left on the partition. I checked the list of kernels and 6.12.9 was also still there. I manually removed it, 6.12.17 installed correctly after that.
That's all 🙂

brent

786psi make sure your boot partition is 1GB. 512MB no longer enough.

There's a command somewhere to re-marry lightdm. will try to dust off my notes.

786psi

brent My install is from 2020. I checked I made it at 512MB at the time, so there's that. But it seems I was not only one to have older kernel not properly removed.

Sebastian

@Animpacz
On the issue of kernel and EFI partition running out of space I'd like to mention that there seems to be some kind of bug currently as was reported here.

qk4-li3

All went well with graphical user interface, Software Center or whatchamacallit.

brent

786psi absolutely both are true.

banger

brent

I've just increased my boot partition from 600MB to 1.7GB to try and avoid the space problem in future but so far been unaffected.

zmaint

Need some help:

Upgraded kwallet-pam
[✓] Syncing filesystems success
[✓] Updating dynamic library cache success
[✗] Updating clr-boot-manager failed

A copy of the command output follows:

[FATAL] cbm (../src/bootman/kernel.c:L740): Failed to install extra initrds /usr/lib64/kernel/initrd-com.solus-project.current.6.12.17-314: No space left on device
[FATAL] cbm (../src/bootman/update.c:L250): Failed to install default-current kernel: ///usr/lib/kernel/com.solus-project.current.6.12.17-314

[✗] Updating clr-boot-manager failed

A copy of the command output follows:

[✓] Running depmod on kernel 6.12.17-314.current success
[✓] Updating graphical driver configuration success
[✓] Updating system users success
[✓] Updating systemd tmpfiles success
[✓] Reloading systemd configuration success
[✓] Re-starting vendor-enabled .socket units success
[✓] Compiling and Reloading AppArmor profiles success
[✓] Compiling glib-schemas success
[✓] Rebuilding font cache success
Failed to record path /usr/share/fonts/conf.d/66-noto-color-emoji.conf.newconfig
[✓] Updating mimetype database success
[✓] Updating icon theme cache: hicolor success
[✓] Updating icon theme cache: breeze_cursors success
[✓] Updating icon theme cache: Breeze_Light success
[✓] Updating icon theme cache: default success
[✓] Updating desktop database success
[✓] Updating manpages database success
[✓] Reloading udev rules success
[✓] Applying udev rules success

/dev/sdb1 489M 440M 49M 91% /boot
zsolus@zsolus ~ $ ls /boot/EFI/com.solus-project/
bootloaderx64.efi initrd-com.solus-project.current.6.12.12-313.nvidia kernel-com.solus-project.current.6.12.17-314
BOOTX64.CSV initrd-com.solus-project.current.6.12.17-314 kernel-com.solus-project.current.6.12.9-312
freestanding-com.solus-project.nvidia-firmware initrd-com.solus-project.current.6.12.9-312 mmx64.efi
grubx64.efi initrd-com.solus-project.current.6.12.9-312.nvidia
initrd-com.solus-project.current.6.12.12-313 kernel-com.solus-project.current.6.12.12-313

Animpacz

zmaint
Seems like you've the same problem as I had.

maybe have a look on the discussion between pomon and me Animpacz in this threat a bit earlier.

zmaint

Animpacz Thanks. I was able to piece together what to do. I deleted the .312 kernel and files manually from the folder, told it to use the .314 kernel, then re-updated clr and usysconf, crossed my fingers, puckered up and booted... and it it worked fine. I reckon I'll have to fire up a live session and see about increasing that partition size. It's set to whatever Solus defaulted to at the time of install (like 4 years ago).

zmaint

Quick followup. Disks shows the EFI partition to be 513mb. I have a 4gb swap file next to it. Is it ok to cut into the swap to make the EFI partition larger? This is a desktop, but I do have a UPS that might have to shut things down if I lose power (live out in the middle of nowhere). I have 32gb of ram, would a 3gb swap be sufficient?

Drive has 447.6GB free total space.

brent

zmaint I've done this to the ext4 to make boot bigger when the ext4 was next to my 512MB boot long ago...I don't see why you could shave a little off swap. Gparted should let you methinks.

« Previous Page