I am using pacman -Syyu on console to update my systems.
Last stable update, I compiled old nvidia driver (450) for linux54. This update ships with a newer linux54 version, and as told in the support topic:
You may need to recompile nvidia-4YYxx whenever we release a new linux54 kernel.
I indeed needed to recompile nvidia-450xx after this update. Here’s how to do it (you can change all occurrences of “450” with your desired driver):
- Add the line
IgnorePkg = lib32-nvidia-utils lib32-opencl-nvidia mhwd-nvidia opencl-nvidia
in file/etc/pacman.conf
- Upgrade your packages (graphical interface or
sudo pacman -Syyu
) - Reboot (
sudo reboot
) on kernel 5.4 (see Note 1 below on how to do this) - Boot will be stuck because no driver has been found, just open a console with CTRL + ALT + F2.
- First read Note 2 below, then get the sources for your driver (in my case, 450xx):
git clone https://gitlab.manjaro.org/packages/extra/nvidia-450xx-utils.git
git clone https://gitlab.manjaro.org/packages/multilib/lib32-nvidia-450xx-utils.git
git clone https://gitlab.manjaro.org/packages/extra/linux54-extramodules/nvidia-450xx.git
- Compile the driver (if you get an error, see Note 3):
cd nvidia-450xx-utils && makepkg -si
cd ../lib32-nvidia-450xx-utils && makepkg -si
cd ../nvidia-450xx && makepkg -sif
- Reboot (
sudo reboot
) on kernel 5.4 (see Note 1 below on how to do this)
Now, everything is working again! To check if you’re indeed running kernel linux 5.4, you can run uname -r
in a console, and to check if you’re running the driver you just compiled, you can run inxi -G --display
.
Note 1: to choose which kernel to boot on, you need to edit those two lines in /etc/default/grub
to match those (5 as in 5 seconds to make your choice on the menu, you can put whatever):
GRUB_TIMEOUT=5
GRUB_TIMEOUT_STYLE=menu
Then, run sudo update-grub
. When booting, you’ll now see a menu with all installed kernels so you can choose which one to use for this session.
Note 2: During step 5 and 6: If you had already compiled the nvidia driver last update, you can skip cloning nvidia-450xx-utils
and lib32-nvidia-450xx-utils
(you already have the sources, you just need to compile them again), but I suggest you remove your older nvidia-450xx
folder (rm -rf nvidia-450xx
) and clone it again before compiling this one, otherwise you will get an error like “folder already exists” when trying to compile.
Note 3: When trying to compile the driver, you might get an error with two nvidia packages creating conflicts (I think lib32-nvidia-utils
was one of them, I don’t remember the other one). To fix that, just uninstall those packages with sudo pacman -R <package>
.
$ pactree -rl qt4
qt4
libdbusmenu-qt4
sni-qt
phonon-qt4
I guess it’s safe to remove qt4
altogether now?
There are some minor cosmetic problems that I realized.
- xfce4-power-manager system tray icons are different from my choosen icon package. It wasn’t changed when I’d chosen a different icon package, it uses the same icon whatever I choose things.
Downgrade package is resolve the problem for temporary. - Standard favorite apps seen on Whisker Menu has disappeared.
I did not do anything.
As far as I can see, the update went fine on three computers:
1 * KDE and 2 * XFCE.
Thanks!
Yes, I rebuilt teamviewer but the crash still occurs. I’ll follow your advice and report to the bugtracker too.
Edit: Found an open bug report on this at Status Notifier crashes opening TeamViewer 13 o 14 (#6) · Issues · Panel Plugins / xfce4-statusnotifier-plugin · GitLab
After this update, on reboot I have :
Failed to start load kernel modules
and the system stop.
I don’t know to fix that.
After doing
pacman -Syyu
I get
error: gptfdisk: signature from "Evangelos Foutras <evangelos@foutrelis.com>" is invalid :: File /var/cache/pacman/pkg/gptfdisk-1.0.6-1-x86_64.pkg.tar.zst is corrupted (invalid or corrupted package (PGP signature)). Do you want to delete it? [Y/n]
Help appreciated.
On the upstream there is a fix for the xfce panel: statustray: Prevent crash when parsing properties (Fixes #379) (e4616236) · Commits · Xfce / xfce4-panel · GitLab
which has been appplyed two weeks ago; can you consider to push it on Manjaro’s branches?
Sorry: maybe is not related:
Everything went fine, apart from a broken power icon on XFCE.
Internal microphone down after the update.
Tried solving with ArchWiki/PulseAudio/TroubleShooting/Microphone_not_detected, now the microphone can be detected, but no sound can be recorded.
Working on it…
:: Synchronizing package databases...
error: failed retrieving file 'core.db' from mirror.deace.id : error:1408F10B:SSL routines:ssl3_get_record:wrong version number
error: failed to update core (download library error)
error: failed retrieving file 'extra.db' from mirror.deace.id : error:1408F10B:SSL routines:ssl3_get_record:wrong version number
error: failed to update extra (download library error)
error: failed retrieving file 'community.db' from mirror.deace.id : error:1408F10B:SSL routines:ssl3_get_record:wrong version number
error: failed to update community (download library error)
error: failed retrieving file 'multilib.db' from mirror.deace.id : error:1408F10B:SSL routines:ssl3_get_record:wrong version number
error: failed to update multilib (download library error)
error: failed to synchronize all databases
Tried to update using discover, and end up in this. Any solution? Sorry if I have made any mistake.
sudo pacman-mirrors --country “your country or a near one”
sudo pacman -Syyuu
Fixed the problem for me
Updated just now, no issues.
Is the black screen bug on boot with AMD and kernel 5.10 still persist?
Edit: nvm I saw it still persists, so I’m stuck in 5.9 for now unfortunatelly
Thank you that solved my problem too
There were 2 packages, … deleted and seems fine now.
OpenVPN, update-resolv-conf failing after this update.
2021-01-19 15:05:25 us=516871 /etc/openvpn/up.sh tun0 1500 1585 172.21.0.44 255.255.0.0 init
+ run /etc/openvpn/update-resolv-conf
+ /etc/openvpn/update-resolv-conf
dhcp-option DNS 172.21.0.1
Cannot write to /run/resolvconf/lock
+ '[' 1 -ne 0 ']'
+ echo '/etc/openvpn/update-resolv-conf Failed with exit code 0'
/etc/openvpn/update-resolv-conf Failed with exit code 0
+ return 1
2021-01-19 15:05:25 us=531116 WARNING: Failed running command (--up/--down): external program exited with error status: 1
2021-01-19 15:05:25 us=531161 Exiting due to fatal error
Probably related to:
https://bugs.archlinux.org/task/68480
Update went smooth af on Manjaro KDE from the 2020-12-30 update. Thanks!
Such a huge update for me!
And no problem so far on xfce!
Thank you Manjaro team!