[Unstable Update] March 2025

consider this blind searching,

my findings are that the said recorfail=1 setting was in so called “grub enviroment block”. a date check in the actual file that contains the said environment block /boot/grub/grubenv says it has been this way since i installed manjaro.

you are only supposed to “unset” the grub-enviroment-var with;

$ sudo grub-editenv - unset recordfail

on unsetting “recorfail” i see that the /boot/grub/grubenv file is modified to current date. so i think recorfail=1 was set all this time, it only triggered the unncessary issues only after the recent changes triggered on it.

current grub enviroment block cocntents can be seen by command;

$ grub-editenv list

or by simply

$ sudo cat /boot/grub/grubenv

ADDENDUM:
so i started(booted) the nextday with no issues (after unsetting ‘recodfail’) things were as they were before. however after booting when i checked the contents of the grub-environment-block; recordfail=1 was back. and no, the date modified was untouched on the file `/boot/grub/grubenv’.

so something indeed is setting recordfail=1 since the recent update.

ADDENDUM #2;
in my case, commenting line #134 in /etc/grub.d/00_header

if [ -n "\${have_grubenv}" ]; then if [ -z "\${boot_once}" ]; then save_env recordfail; fi; fi

prevented setting of recordfail=1 on each successive boot.

2 Likes

The new Ubuntu patches set recordfail. We will see what happens when we remove them.

1 Like

linux612-nvidia-open is listed as orphaned, although the nvidia open driver and kernel 6.12 are installed. Is that intentional?

Not sure why it would be an orphan. Set it to explicitly installed:

sudo pacman -D --asexplicit linux612-nvidia-open
1 Like

Because it’s not living on gitlab anymore:

It’s a split package with linux612-nvidia :wink:

Having issues with audio after the pipewire package updates today. No streaming playback possible through any browser, only able to play audio through audio players directly to ALSA devices. I’ve rolled back the updates for now.

Any advice on what to post for support with the issue?

A package being an oprhan means it was installed as a dependency of another package and is not needed by any other package now.
It has nothing to do with its PKGBUILD availability on any gitlab.

2 Likes

I also noticed an issue since the Pipewire update. When I switched user accounts then returned to my user, audio was off. I couldn’t find the devices, rather than change anything I logged out then back in and then the audio worked.
I have audio through Display Port to my monitor.

1 Like

On my system, the process /usr/bin/pipewire is holding 1 CPU core hostage with 100% usage. Downgrading pipewire* to previous version restores normal function…

Edit: on Debian SID the same thing happens, so I suppose it is a bug upstream ?

Cheers,
Eddy

1 Like

Installing pipewire-pulse, pipewire-alsa and pipewire-audio resolved the issue for me.

Had to remove manjaro-pulse, pulseaudio-bluetooth and pulseaudio.

Related arch forum thread: [FIXED] No audio since pipewire updated this morning / Multimedia and Games / Arch Linux Forums

1 Like

That is a manual approach for switching to pipewire with manjaro-pipewire meta package iirc.

From limited testing (gnome & firefox) I do not see this behavior (for reference)

Summary
gst-plugin-pipewire                 1:1.4.0-2                     extra     163,2 kB
lib32-libpipewire                   1:1.4.0-1                     multilib  898,0 kB
lib32-pipewire                      1:1.4.0-1                     multilib  4,9 MB
lib32-pipewire-jack                 1:1.4.0-1                     multilib  485,1 kB
libpipewire                         1:1.4.0-2                     extra     1,9 MB
manjaro-pipewire                    20231009-1                    extra     1,6 kB
pipewire                            1:1.4.0-2                     extra     3,0 MB
pipewire-alsa                       1:1.4.0-2                     extra     1,2 kB
pipewire-audio                      1:1.4.0-2                     extra     6,2 MB
pipewire-jack                       1:1.4.0-2                     extra     701,5 kB
pipewire-pulse                      1:1.4.0-2                     extra     525,9 kB
pipewire-zeroconf                   1:1.4.0-2                     extra     393,1 kB

1 Like

I can also confirm that the latest update broke the sound for Firefox and Spotify as well.
In my case, it was resolved by installing manjaro-pipewire as @Hanzel suggested, which uninstalled various PulseAudio-related packages.

I think the issue only exists if you are still on PulseAudio, which is probably the reason why you weren’t experiencing problems (I don’t see PulseAudio in your package list).

Interesting insights from the Arch-Linux forums:

gst-plugin-pipewire got a hard dep on the audio part, which is probably intentional, if you get the audio parts then most things will try to connect to pulseaudio via pipewire which gets blocked by pulse or vice versa. If you actually want to stay on pulseaudio for some reason you should ideally remove gst-plugin-pipewire and dependants, or potentially better yet, switch to wireplumber and use the @video-only service variant

I was getting constant 100% usage on one core by pipewire 1.4.0 also (but sound worked). Pipewire-pulse, pipewire-alsa, and pipewire-audio were already installed. Pulseaudio is not installed. Downgrading pipewire packages to 1.2.7 restored normal function.

Yes, for me there were also no Pulseaudio libraries installed, and pipewire-pulse, pipewire-alsa and pipewire-audio were also already there. Sound is playing fine indeed. The strange thing is that my other linux install - Debian (unstable/sid) - is doing the same thing with 1 CPU core @ 100%. Therefore I supposed a bug upstream, but not many other users are reporting the same bug so far ? :face_with_monocle:

Greetings,
Eddy

1 Like

I am getting a dependency breaks with libcap and lib31-libcap:

libcap (2.75-1) breaks dependency 'libcap=2.71' required by lib32-libcap  

Both packages are available at version 2.75-1.
Your mirror seems to be partially synced.

Indeed, french mirror seems to have issues. Thanks !!

I wish pacman and pamac would verify that a mirror is up to date before proceeding and switch if necessary (or at least offer a warning to the user). It would fix issues like this. I’ve gotten into the habit of always manually checking, but it would be nice to have it automated.

2 Likes

I thought I’d share a script I started working on for my parents as a “workaround” for the recent pamac issues; as they pretty much rely on the ease of the pamac-gtk3 GUI.

when I’ve run sudo pacman-mirrors -f in the terminal I’ve often seen it omitting one of the 3 servers for my country, which I presume means that unsync’d server(s) were omitted.

assuming that is correct… what I’m not sure about is what if all the mirrors were not sync’d? Would the sudo pacman-mirrors -f command return an error?

If these assumptions are true, a bash script might get you the automation you are looking for…

System Update Script
#!/bin/bash
trap_error() {
        ERRORCODE=$?
        echo "Errorcode $ERRORCODE in $0 at line $BASH_LINENO"
        echo "Failed/Aborted command: $BASH_COMMAND"
        echo "Exiting script in 15 seconds." && sleep 15
        exit $ERRORCODE
}

trap trap_error ERR INT TERM

echo "<= Start of System Update Script =>"
echo && echo "Checking/Updating mirror list..."
sudo pacman-mirrors -f

echo && echo "Updating Manjaro package databases..."
sudo pacman -Syy

echo && echo "Updating Manjaro packages..."
sudo pacman -Syu

echo && echo "PACNEW check/resolution..."
DIFFPROG=meld pacdiff -s

echo && echo "Orphan List..."
pacman -Qdt

echo && echo "Foreign Package List..."
pacman -Qm

echo && read -p "Considering the Orphan/Foreign lists above... continue to update AUR packages? [Y/n] " yn_aur
  if [[ -z "$yn_aur" ]] || [[ $yn_aur == "y" ]] || [[ $yn_aur == "Y" ]]; then
    pamac update -a
  else
    echo "AUR package updates aborted..."
    echo "Exiting script in 15 seconds."
    echo "Note: After manually dealing with Orphan/Foreign packages, remember to update AUR and reboot." && sleep 15
    exit
  fi

echo "<= End of System Update Script =>"
read -p "Reboot now? [Y/n] " yn_reboot
  if [[ -z "$yn_reboot" ]] || [[ $yn_reboot == "y" ]] || [[ $yn_reboot == "Y" ]]; then
    sudo systemctl reboot now
  else
    echo "Remember to reboot soon!"
    echo "Exiting script in 5 seconds." && sleep 5
    exit
  fi

Edit it to your taste if the script is doing more (or less) than you would like it to do… or maybe you just need to add sudo pacman-mirrors -f before your pamac/pacman commands?

Food for thought.

P.S. I thought the “error trap” logic I found @ only run command after if first command is successful - #16 by gerardbeekmans - Linux - Spiceworks Community was pretty slick and helpful as I was learning to write this script… all props to that contributor.