I wanted to try out pipewire on Manjaro with Gnome 41.2, but the pipewire process is just stalling all my applications, not finding any audio device and leaks memory until pipewire consumes about all of my memory (128GB RAM) and makes the entire system freeze.
What I can do to prevent the system from crashing is freezing the pipewire process or regularly killing it before the system memory is exhausted by pipewire. Obviously, this isn’t a useable state so I want to remove pipewire and switch back to pulse.
I installed pipewire by simply installing manjaro-pipewire without the optional dependency wireplumber. Uninstalling manjaro-pipewire didn’t remove any other package, but installing manjaro-pulse fails due to a dependency error which I cannot make sense of:
sudo pacman -S manjaro-pulse
resolving dependencies...
looking for conflicting packages...
:: manjaro-pulse and pipewire-jack are in conflict. Remove pipewire-jack? [y/N] y
:: manjaro-pulse and pipewire-pulse are in conflict. Remove pipewire-pulse? [y/N] y
:: manjaro-pulse and pipewire-zeroconf are in conflict. Remove pipewire-zeroconf? [y/N] y
error: failed to prepare transaction (could not satisfy dependencies)
:: unable to satisfy dependency 'pulseaudio-bluetooth' required by manjaro-pulse
:: removing pipewire-pulse breaks dependency 'pulseaudio-bluetooth' required by gnome-bluetooth
Any idea how to fix the dependency issue and go back to PulseAudio?
I guess the process is misbehaving. It’s also constantly putting 100% load on a CPU core. It’s probably hanging in an unprotected while loop or something like that.
RME Hammerfall DSP MADI can support 64 channels of 24-Bit/48kHz audio (or 32 channels of 24 bit/96 kHz); card has a manufacturer-specific driver snd_hdspe for ALSA and hdspmixer from package alsa-tools to control the audio channels
I suspect that Pipewire may not be able to manage all the channels on this card effectively and/or the card does not work well with timer-based scheduling required by PipeWire
After the stuck-in-while-loop issue was fixed within the pipewire repository, there are audio issues when accessing the card via pipewire, indeed. I’ll try to contact the devs about this.
(Accessing the card from ALSA seems to be working fine, though)