Again sometimes audio stopped (pipewire 0.3.28, PulseEffects 5.0.3). I also use 3.5mm jack output directly from my motherboard for my speakers.
Did the upgrade with an nvidia GTX 1070, 2x 27 inch 1440p 165hz (MSI Optix MAG274QRF-QD) and Iām using the displayports! All went fine for me, I can make fully us of my system.
I have one problem thoughā¦ Iām using a wireless headset (corsair virtuoso se) with kernel 5.4 and I can make use of the profile āDigital audio output + Multichannel inputā (pulse audio), I tried out kernel 5.10 (with this new rolling release) and I canāt make use of that profile, at least I donāt have sound then. I only have sound with the profile āAnalog Stereo output + Multichannel inputā in kernel 5.10 (pulse audio) and the sound is a little bit lower compared to the digital profile from 5.4. Anyone else has that problem?
I gave up with pipewire, pipewire has to many problems and bugs to replace the old Alsa/Pulse sound system, so i removed most of pipewire and went back to alsa/pulse and all Audio works again with Zero Issues.
Issue from stable-update 2121-04-28 occurs again:
āIn some rare cases you may face the following error on your KDE Plasma Desktop: usr/lib/libKF5KIOCore.so.5: undefined symbol: _ZN4KJob14percentChangedEPS_mNS_14QPrivateSignalE
.
Please click the āŗ at the beginning of this line to view this solutionā¦ā
Solution (downgrade to kio back to 5,79.0) this time doesnt work. - Any ideas?
I also did this,
pacman -Q | grep nvidia
added all those to ignore packages in etc/pacman.conf and the system updated and booted right up
Check if you have any -git packages of KDE software or plasma-desktop-primex installed. If so, replace them with the regular packages.
Works. (replacement -git packages) Thank you!
Dang, I also experienced this when I came back home today. I had to restart PulseEffects and then it all worked. I used these 2 commands in a terminal to fix it:
$ pulseeffects -q
$ pulseeffects --gapplication-service & disown
Maybe this is a PulseEffects bug though. Iām using pulseeffects-git
from the AUR with @philmās build of PipeWire 0.3.28. Iāll check the PulseEffects bug tracker on GitHub.
Iāve got to say this situation with the Nvidia drivers is pretty terrible. As a somewhat literate user, Iāve read this thread multiple times to determine what the best course of action is to not end up with a system that wonāt boot, and Iāve found conflicting information about which packages to ignore, then due to kernel updates, maybe having to get a dkms Nvidia driver from sourceforge for the non upgraded driver to work. I just donāt think this is even somewhat of a workable solution for someone relatively new to linux. What is needed is some sort of consistent message as to the best course of action for the users that do end up at this forum to take.
For right now Iām choosing to simply not upgrade, but Iām sure there are more than a few users out there who are just sitting at a black screen and have no way to figure out how to get back to a working system.
I love manjaro but these kind of situations make it difficult to recommend to users wanting to try linux.
@CarloChaos: thank you, you are right!
Above, you can see, I was one of the owners of blackscreen (DP/4K) . Now I am back to a working system thanks Timeshift and before a working update (like you)
The bug with NVIDIA is critical, but the Dev-communication to this is a disaster. I use manjaro for more than three years (and it makes mainly a fantastic job). Since the first NV-disaster at the end of 2020 I look for alternatives, cause Stable (for me) seems more and more experimental.
A second problem, for a longer time Manjaro donĀ“t initialize sound with two different Gigabyte-Boards (with LiveStick!) ALC1220 Soundchip. But all functions are o.k. with MX-Linux and Fedora, Win10.
After request here (German section) the reaction: DAU and the question āYou work for Microsoft?ā - N.C. - Enough is enough!
I have an issue with Kdenlive update, video transitions are not working anymore.
Subject can be found here.
I could not boot because of the displayport bug (nvidia 1050 ti, 1920x1080). Fortunately I still had a hdmi cable. With that connected it boots correctly.
ya, a real mess
absolutely true MichaelP, I dont know why , but āstableā has lost its usual robustness. I suggest that more time be allocated to intensive testing. otherwise Manjaro reputation will suffer. āstableā must be rock solid.
Donāt know why, but the āofficialā mesa provided by Manjaro - Mesa 21.0.3-3 - does not work very well with Vega 64. Several of games refuse to start at all (HZD; Control). Kind of workaround so far was to replace them with mesa-git and lib32-mesa-git from AUR and it has been working quite well, however, being devel branch and git and whatever, it is also kind of unstable and subject to crashes. Lately too often sadly.
Is it okay for now to hold off on this update and wait until a next stable release with nvidia fixes? I donāt want to take the risk tbh.
Or is it better to update everything except for the nvidia drivers?
This nvidia fiasco has been handled exceptionally poorly.
Deliberately including a widely reported broken driver in a stable release known to cause a failure to boot is unforgivably blithe and only serves the damage the reputation of the distribution.
The solution of offering vague and often incompletely explained instructions concerning pacman exclusions is one thing, recommending users transition to nouveau is arrogance beyond reprieve. How anyone could recommend a driver set known to be a poorly performing, functionally incomplete alternative with limited hardware support for any GTX900 maxwell class cards or newer is unfathomable, especially in light of the ease of bundling the working nvidia drivers. Note this is a statement of fact not a criticism of the efforts of those attempting to reverse engineer, deliberately hindered by nvidia.
I can only surmise the arrogance with which this has been imposed is due to some sort of distaste for nvidia by those making package inclusion decisions.
Sorry, but I just donāt understand you.
Linux is not Windows. Linux is free, and is maintained by willing people who spend their free time running an operating system as complex as Manjaro.
The Nvidia driver is proprietary, and in Manjaro āthey tryā to make it work, but it is not the fault of the Manjaro maintainers if Nvidia does not cooperate.
Rather blame Nvidia.
Or, much more easily, dual-boot Windows and use it for those games (or applications) that donāt work in Manjaro.
Itās not very difficult, is it?
On shutdown, I had the message
Failed unmounting /tmp:
systemd-shutdown[1]: Waiting for process: (about 15 different processes here)
and it hung for about 5 minutes before finally shutting down.
Itās quite simple, why release a known broken driver for Nvidia in stable when a perfectly suitable working alternative exists? Nouveau is virtually unusable for modern nvidia cards.
This is not a complex decision.
This has nothing to do with your irrelevant statements about linux being free, proprietary drivers, complexity of the OS or some vague recriminations about using windows. Thereās no ātrying to make it workā in this case, just bundling of a working driver. Nvidia co-operating, what do you mean? Somethings political about proprietary versus free?
To reiterate, evil Nvidia release a broken driver, itās widely acknowledged as such and well known, someone includes it in stable?
You seem to be of the opinion that because linux is free, expect it to be deliberately broken, questionable actions are immune from criticism and just install an entire alternative OS should it break?