Pulse Audio dropouts after Manjaro Testing 2025.07.26 update
Yesterday, 11:31 PM
(This post was last modified: Yesterday, 11:39 PM by Dendrocalamus64.)
I see on the Manjaro forums they have a thread for every ARM Unstable Update.
https://forum.manjaro.org/c/arm/100
They say that only ARM Unstable is maintained anymore, but Testing seems to be getting updates too.
This Testing update arrived around 07.26 (Sat). Videos that used to play fine in mpv started getting debilitating amounts of audio dropouts with a buffer underrun message each time.
At first I thought it was due to video decompression getting less efficient and eating all the CPU, but playing the same files with --no-video and top running showed the system lightly loaded (~8% CPU), and yet the audio dropouts occured in the exact same places.
I tried updating to the Megi 6.15 kernel and rebooting. mpv no longer said "Audio device underrun detected" but the dropouts still occured, and pulseaudio was now dying all the time and needing to be manually restarted.
Start reading through this loooong page:
https://wiki.archlinux.org/title/PulseAu...leshooting
Try running with pulseaudio -v, notice that it seems to be suspending too much.
Section 4.8, unload-module module-suspend-on-idle
And that fixed the problem completely.
(I had tried various other things since 07.26 like turning timer-based scheduling on and off, and adjusting buffer sizes, and so far none of it had helped.)
So pulseaudio suspend-on-idle is currently completely broken, at least with the default config. It thinks it's "idle" while it's playing back a video and the laptop is actively being used.
Next problem was pulseaudio sometimes starts endlessly spamming to the console,
Solution to that was to edit /etc/pulse/daemon.conf and set both default-sample-rate and alternate-sample-rate to 44100.
https://forum.manjaro.org/c/arm/100
They say that only ARM Unstable is maintained anymore, but Testing seems to be getting updates too.
This Testing update arrived around 07.26 (Sat). Videos that used to play fine in mpv started getting debilitating amounts of audio dropouts with a buffer underrun message each time.
At first I thought it was due to video decompression getting less efficient and eating all the CPU, but playing the same files with --no-video and top running showed the system lightly loaded (~8% CPU), and yet the audio dropouts occured in the exact same places.
I tried updating to the Megi 6.15 kernel and rebooting. mpv no longer said "Audio device underrun detected" but the dropouts still occured, and pulseaudio was now dying all the time and needing to be manually restarted.
Start reading through this loooong page:
https://wiki.archlinux.org/title/PulseAu...leshooting
Try running with pulseaudio -v, notice that it seems to be suspending too much.
Section 4.8, unload-module module-suspend-on-idle
And that fixed the problem completely.
(I had tried various other things since 07.26 like turning timer-based scheduling on and off, and adjusting buffer sizes, and so far none of it had helped.)
So pulseaudio suspend-on-idle is currently completely broken, at least with the default config. It thinks it's "idle" while it's playing back a video and the laptop is actively being used.
Next problem was pulseaudio sometimes starts endlessly spamming to the console,
Code:
W: [alsa-sink-ff890000.i2s-ES8316 HiFi ES8316 HiFi-0] alsa-sink.c: Resume failed, couldn't restore original sample settings.
I: [alsa-sink-ff890000.i2s-ES8316 HiFi ES8316 HiFi-0] alsa-sink.c: Trying resume...
I: [alsa-sink-ff890000.i2s-ES8316 HiFi ES8316 HiFi-0] alsa-util.c: Cannot disable ALSA period wakeups
I: [alsa-sink-ff890000.i2s-ES8316 HiFi ES8316 HiFi-0] alsa-util.c: Device hw:rockchipes8316c doesn't support 48000 Hz, changed to 44100 Hz.Solution to that was to edit /etc/pulse/daemon.conf and set both default-sample-rate and alternate-sample-rate to 44100.
