There is only one provider in the official repository. I suggest that one. It’s a 600Kb file more or less. I don’t see the issue.
Surprised that 6.18 LTS is a no-go for your setup
Just as i thought and researched but i needed confirmation from user. Ok it’s dependency spaghetti. I will add this to the known problems later.
2 Likes
From the Debian wiki:
For people that need to stay on 6.12, a backported ntsync module has been created as a DKMS package that will be automatically installed together with the below described patched Wine packages (unless 6.14.5 or newer kernel is already installed).
This was brought up in the Arch Forums, marked solved, without a good solution.
https://bbs.archlinux.org/viewtopic.php?id=309389
1 Like
All good on the update but again I got bumped up to the latest kernel 7.1.6-1
This is after I removed the meta kernels as previously noted.
What are the main benefits of kernel 7.1.6-1?
You can always remove 7.1 and go back to 6.18 lts.
As for the changes
P.s. the problem is grub always tries to boot the newest installed. The choice cannot be customized on btrfs so if you are annoyed by manually having to select 6.18 each boot you have to remove 7+
3 Likes
I’m still having problems with wine updates.
this update brought wine 11.14-2 which doesn’t work well either with gta4: the game doesn’t receive mouse movements, but receives mouse clicks and wheel scrolling, and keyboard presses.
I had to once again roll back to wine 11.12-1. but this time it was harder to do because of the ntsync-autoload dependency. I ended up removing wine completely together with ntsync-autoload and other dependencies, and installing back the older version.
kernel 6.18 lts
This article has a nice summary of all the benefits 
Blockquote AMD GPU and display drivers pulled the highest single-driver patch count at 80 commits. Alex Deucher leads with 15 patches across the amdgpu and DCN32/35 display stack. Zhao Li follows with 9 patches.
Networking wasn’t far behind at roughly 169 patches total. Guangshuo Li contributed 12. Ruoyu Wang and Pengpeng Hou each brought 9. Eric Dumazet and Matthieu Baerts round out the top networking contributors with 7 each.
WiFi drivers ate up about 68 of those networking patches. The ath12k, ath6kl, carl9170, and ath11k families were the primary targets. Use-after-free and OOB vulnerabilities were the recurring theme across those drivers.
Everything works fine.
Thanks for the update guys!
I need a little help please.
I have an old nvidia card and can’t use anything newer than 470, which is fine for me.
The current nvidia driver I am running is: 470.256.02
However, running pacman -Syu now lists this package as an update:
mhwd-nvidia-580xx-580.178.04-1
I am not on nvidia-580xx specifically because of that known incompatibility with my card.
I find that nvidia-580xx is listed as a dependency for mhwd-db.
Question: If I proceed with the update and that 580xx package gets installed, will my nvidia driver still remain on the current 470xx version (470.256.02) or will it automatically change to 580xx and break my display?
The 470xx driver should not be changed
mhwd-nvidia packages provide module IDs of compaitble GPUs for Manjaro Hardware Detection
$ pamac info mhwd
Name : mhwd
Version : 0.6.6-6
Description : Manjaro Linux Hardware Detection library and application
Required By : mhwd-nvidia mhwd-nvidia-390xx mhwd-nvidia-470xx mhwd-nvidia-580xx
1 Like
I confirm. After the update, graphical artifacts started appearing.
System
user@desktop
--------------
OS: Manjaro Linux x86_64
Kernel: Linux 7.1.6-1-MANJARO
Uptime: 12 hours, 41 mins
Packages: 1203 (pacman)[stable]
Shell: vim
Display (EI322QUR): 2560x1440 in 31", 60 Hz [External]
DE: KDE Plasma 6.7.4
WM: KWin (Wayland)
WM Theme: Breeze
Theme: Breeze (BreathLight) [Qt]
Icons: breeze [Qt]
Font: Noto Sans (10pt) [Qt]
Cursor: breeze (24px)
Terminal: tmux 3.7b
CPU: AMD FX(tm)-8350 (8) @ 4.00 GHz
GPU: AMD Radeon RX 9060 XT [Discrete]
Memory: 4.77 GiB / 31.30 GiB (15%)
Swap: 0 B / 8.00 GiB (0%)
Disk (/): 133.21 GiB / 1.11 TiB (12%) - xfs
Local IP (enp5s7): 192.168.1.13/24
Locale: en_US.UTF-8
The memory and processor are not overclocked. The graphics card is at stable undervolting level. Before the update, with the same settings, there were no artifacts.
I suspect it’s a KDE issue, as there are no artifacts in the kernel console, including when using mpv.
Update: It looks like it’s not KDE, as in the Fedora thread mentioned above, people are having the same problems with other DEs. Could it be Vulkan?
After the latest update I am seeing graphical artifacts in the bottom of my screen (blinking black lines)
Kernel: 6.18.42-1-MANJARO
DE : Plasma 6.7.4
GPU: AMD ATI Radeon RX 7900 XT/7900 XTX/7900 GRE/7900M
WM: kwin
After the update i have a strange output from pacdiff:
$ pacdiff -s
==> pacnew file found for /etc/locale.gen
:: (V)iew, (M)erge, (S)kip, (R)emove pacnew, (O)verwrite with pacnew, (Q)uit: [v/m/s/r/o/q] v
find: ‘/var/cache/pacman/pkg/download-ifuBNq’: Permission denied
find: ‘/var/cache/pacman/pkg/download-lXSlWs’: Permission denied
[sudo] password for koboldx:
sudoedit: /etc/locale.gen.pacnew unchanged
:: (V)iew, (M)erge, (S)kip, (R)emove pacnew, (O)verwrite with pacnew, (Q)uit: [v/m/s/r/o/q] r
removed '/etc/locale.gen.pacnew'
find: ‘/var/cache/pacman/pkg/download-ifuBNq’: Permission denied
find: ‘/var/cache/pacman/pkg/download-lXSlWs’: Permission denied
It is safe to remove those empty directories:
sudo rm -fr /var/cache/pacman/pkg/download-*/
3 Likes
I can observe them too with my 9060XT but with kernel 7.1, sporadically also when switching between fullscreen / fullscreen windowed apps to and from desktop.
Might be related to 2512106 – [amdgpu regression] Kernel 7.1.6-201 causes KWin atomic commit failures and dual-monitor corruption; 7.1.5-201 works.
It is most certainly related, but mine is 6.18 and I’m not sure which kernel should I use.
Everything went fine with the update. My system so far running flawlessly.
2 Likes
Well, I have no idea which part of the kernel the issue originates from, but I’ve rolled back to earlier version of the same kernel, and the issue is gone.
It looks like the issue lies somewhere at the intersection of the amdgpu driver and Mesa: the problem doesn’t appear in the kernel console, only when a desktop environment is running - and all of them rely on rendering acceleration via graphics APIs.
A working fix that doesn’t require a kernel rollback was suggested in a related thread: change the “Color accuracy” setting from efficiency to color accuracy. In KDE, this option can be configured under “System Settings → Display & Monitor”.
2 Likes
DELL 5530 all ok (updated together with previous update),
LOQ 15IAX9I all ok,
Desktop (MSI Tomahawk wifi Z790, i5-12600k, RTX 2060 SUPER OC 8GB on U3423, ram 64 GB DDR5) all ok. Now no massage on running windecfg
nice job!
1 Like