No sound with pulseaudio after kernel 5.17.4+

[Saundersx]

Just like the title says, sound works fine with 5.17.3 but after upgrading to 5.17.4 through .6 pulseaudio is broken. Looking through the other issues on here and none of them apply to me, mainly pipewire was never installed and this isn't a pure bluetooth headset (has its own usb receiver that only works with these headphones).

Code: Select all

$ inxi -A
Audio:
  Device-1: NVIDIA GM204 High Definition Audio driver: snd_hda_intel
  Device-2: AMD Starship/Matisse HD Audio driver: snd_hda_intel
  Device-3: SteelSeries ApS Arctis 7 type: USB
    driver: hid-generic,snd-usb-audio,usbhid
  Sound Server-1: ALSA v: k5.17.6 running: yes
  Sound Server-2: PulseAudio v: 15.99.1 running: yes

I can boot back into 5.17.3 and audio magically works again, boot into 5.17.6 and it's broken. Usually when I run into an issue like this I can wait a few kernel revisions and it will sort itself out, not this time apparently. Checking dmesg between the two kernels gives no relevant differences.

One weird workaround I have found is loading pavucontrol and leaving it running makes everything work again. I tried some media players and watched the output.

vlc (only these two lines were different)

Code: Select all

- bad (no audio)
pulse audio output debug: changing sink 2: PulseEffects_apps (PulseEffects(apps))
- good
pulse audio output debug: changing sink 1: alsa_output.usb-SteelSeries_SteelSeries_Arctis_7-00.stereo-game (SteelSeries Arctis 7 Game)

mpv (bad one outputted this with no audio)

Code: Select all

AV: 00:00:00 / 00:03:39 (0%)
Audio device underrun detected.
AV: 00:00:00 / 00:03:39 (0%)
Audio device underrun detected.
AV: 00:00:00 / 00:03:39 (0%)
Audio device underrun detected.
AV: 00:00:00 / 00:03:39 (0%)
Audio device underrun detected.
AV: 00:00:01 / 00:03:39 (1%)
Audio device underrun detected.
AV: 00:00:01 / 00:03:39 (1%)

edit: almost forgot, lot of these in the syslog when trying to play something

Code: Select all

May  9 16:46:30 xx pulseaudio[4665]: [alsa-sink-USB Audio] alsa-sink.c: Resume failed, couldn't restore original fragment settings. (Old: 384000/192000, New 352800/176400)
May  9 16:46:31 xx pulseaudio[4665]: [alsa-sink-USB Audio] alsa-sink.c: Resume failed, couldn't restore original fragment settings. (Old: 384000/192000, New 352800/176400)
May  9 16:46:31 xx pulseaudio[4665]: [alsa-sink-USB Audio] alsa-sink.c: Resume failed, couldn't restore original fragment settings. (Old: 384000/192000, New 352800/176400)
May  9 16:46:31 xx pulseaudio[4665]: [alsa-sink-USB Audio] alsa-sink.c: Resume failed, couldn't restore original fragment settings. (Old: 384000/192000, New 352800/176400)

if you look at the kernel changelog for 5.17.4 https://cdn.kernel.org/pub/linux/kernel ... Log-5.17.4 there are a lot of alsa changes. I'm inclined to believe one of these created the problem.

[figueroa]

Did you run make oldconfig when upgrading from 5.17.3?

[Saundersx]

Did you run make oldconfig when upgrading from 5.17.3?

No real changes.

[garrison]

Did you run make oldconfig when upgrading from 5.17.3?

No real changes.

If 'pa-info' reports different set of devices or sinks and sources with different kernel versions, you can try collecting verbose log from daemon (i.e. like in stopping running daemon and then 'pulseaudio -vvvv > /tmp/pulse-verbose.log') to compare what is going on while devices are probed by pulseaudio daemon.

This hopefully should reveal what is changed and how; if you are stuck attach 'pa-info' and verbose log for each kernel version.

[omidxo]

Try to turn it on in /etc/pulse/client.conf set autospawn = yes. Remove any ; at the beginning of the line.

The default setting is off. This is for pipewire set. If you well enable pipeware to replace pulseaudio, see https://wiki.gentoo.org/wiki/PipeWire.

[Hu]

if you look at the kernel changelog for 5.17.4 https://cdn.kernel.org/pub/linux/kernel ... Log-5.17.4 there are a lot of alsa changes. I'm inclined to believe one of these created the problem.

Can you bisect between v5.17.3 and v5.17.4 to find which of these ALSA changes caused your problem? Per stable kernel rules, those patches should only be there if they are already in Linus' tree, so it is probable that v5.18, once released, will have the same problem for you. Finding the bad patch and arguing for its revert now, before v5.18 is released, will be easier than waiting and arguing later.

[Saundersx]

Turns out this was something more simple. Had added a few lines for doom eternal back when it came out to fix the sound issues proton had with that game.

from /etc/pulse/daemon.conf , removed

Code: Select all

default-sample-rate = 48000
alternate-sample-rate = 44100

I still think this is a bug but just glad I got it working again (hopefully with my short testing).

[garrison]

Turns out this was something more simple. Had added a few lines for doom eternal back when it came out to fix the sound issues proton had with that game.

from /etc/pulse/daemon.conf , removed

Code: Select all

default-sample-rate = 48000
alternate-sample-rate = 44100

I still think this is a bug but just glad I got it working again (hopefully with my short testing).

These overrides are usually not a problem unless your audio driver stopped working with 48000 sample rate.

[Roman_Gruber]

Thanks for the fix. It seems the massive spam in /var/log/messages is reduced.



For a long time I had similar behaviour. For half a year or longer ...

I found this post by looking at a SUSE post.

Code: Select all

Sienna_Cichlid /home/roman # uname -a
Linux Sienna_Cichlid 6.15.4-gentoo_06_28_2025 #1 SMP PREEMPT Sat Jun 28 04:57:19 CEST 2025 x86_64 AMD Ryzen 5 7600X 6-Core Processor AuthenticAMD GNU/Linux

before i made those changes

massive spam in /var/log/messages

e.g. from today before the change

Code: Select all

Jul  1 11:38:52 Sienna_Cichlid pulseaudio[6459]: [alsa-source-USB Audio] alsa-source.c: Resume failed, couldn't restore original sample settings.
Jul  1 11:38:52 Sienna_Cichlid pulseaudio[6459]: [alsa-source-USB Audio] alsa-source.c: Resume failed, couldn't restore original sample settings.
Jul  1 11:38:52 Sienna_Cichlid pulseaudio[6459]: [alsa-source-USB Audio] alsa-source.c: Resume failed, couldn't restore original sample settings.
Jul  1 11:38:52 Sienna_Cichlid pulseaudio[6459]: [alsa-source-USB Audio] alsa-source.c: Resume failed, couldn't restore original sample settings.
Jul  1 11:38:52 Sienna_Cichlid pulseaudio[6459]: [alsa-source-USB Audio] alsa-source.c: Resume failed, couldn't restore original sample settings.
Jul  1 11:38:52 Sienna_Cichlid pulseaudio[6459]: [alsa-source-USB Audio] alsa-source.c: Resume failed, couldn't restore original sample settings.
Jul  1 11:38:52 Sienna_Cichlid pulseaudio[6459]: [alsa-source-USB Audio] alsa-source.c: Resume failed, couldn't restore original sample settings.
Jul  1 11:38:52 Sienna_Cichlid pulseaudio[6459]: [alsa-source-USB Audio] alsa-source.c: Resume failed, couldn't restore original sample settings.
Jul  1 11:38:52 Sienna_Cichlid pulseaudio[6459]: [alsa-source-USB Audio] alsa-source.c: Resume failed, couldn't restore original sample settings.

kill the process was a temporary fix. But i had to restart e.g. the webbrowser which played the audio.

seems now its a different issue in /var/log/messages

Code: Select all

Jul  1 12:55:58 Sienna_Cichlid pulseaudio[5971]: [alsa-sink-USB Audio] alsa-sink.c: ALSA woke us up to write new data to the device, but there was actually nothing to write.
Jul  1 12:55:58 Sienna_Cichlid pulseaudio[5971]: [alsa-sink-USB Audio] alsa-sink.c: Most likely this is a bug in the ALSA driver 'snd_usb_audio'. Please report this issue to the ALSA developers.
Jul  1 12:55:58 Sienna_Cichlid pulseaudio[5971]: [alsa-sink-USB Audio] alsa-sink.c: We were woken up with POLLOUT set -- however a subsequent snd_pcm_avail() returned 0 or another value < min_avail.

this bug happened with previous ASUS prime x670-p mainboard. For a few days i have now a MSI x670E gaming plus wifi mainboard.

soundcard - according to lsusb

Bus 003 Device 006: ID 0763:410c M-Audio AIR 192 6

This bug also happens when the dp audio from the radeon 7800xt is used also.

wiring

mainboard - usb -> asus monitor usb hub pa278QV -> m-audio air 192|6 audio interface = that is another word for usb sound card

On the previous asus mainboard the asus mainboard audio card was disabled in uefi. i did not bother in the past two weeks to disable the msi mainboard audio. I own that msi mainboard for around 2-3 weeks.

--

Thanks for the fix. It seems the massive spam in /var/log/messages is reduced.

I do delete messages from time to time. grep and using head shows entries from 10th of september. this bug was not there around 10th of september.

A few warning or bug messages do not bother me. But it annoys me to see spam in messages or dmesg. SPAM = same message line from the same process more than two times.

to make it 100% clear

comment out those two lines:
from /etc/pulse/daemon.conf , removed
Code:
default-sample-rate = 48000
alternate-sample-rate = 44100