Mads

I can't replicate it with speaker-test, no.... I'm using pipewire 0.3.56 with default options, and audacious with pulseaudio output as mentioned.

Maybe you're onto something, that it doesn't have anything to do with powertop. I might just have had bad luck in triggering it when I tested it out and created...

I sat PIPEWIRE_DEBUG=5 in the service, and retried: ``` [...] juli 26 00:05:01 marhide pipewire[1303]: spa.loop: 0x5630a548a210: update 0000001d -> 00000019 juli 26 00:05:01 marhide pipewire[1303]: mod.protocol-native: connection 0x5630a534aea0: 22...

dmesg: ``` [ 476.736098] sof-audio-pci-intel-tgl 0000:00:1f.3: ipc tx timed out for 0x60050000 (msg/reply size: 12/12) [ 476.736111] sof-audio-pci-intel-tgl 0000:00:1f.3: ------------[ IPC dump start ]------------ [ 476.736143] sof-audio-pci-intel-tgl 0000:00:1f.3: hda irq...

1) booted 2) Started audacious, playing song. 3) Stream stops between 1 and 2 minutes played. Nothing in dmesg yet. 4) I try systemctl --user restart pipewire Then all those...

No, I did not edit it. It was the time it took before I ran systemctl --user restart pipewire. I checked many times before to be sure. It's the restart...

I had to trigger the stream stopping first with audacious, so it takes a little while. The stream stopping makes no output in dmesg.

Ah, sorry, I didn't catch that comment! I redid the audacious -> restart pulseaudio test now (went a bit faster this time): [dmesg_tgl_ipc_dbg.log](https://github.com/thesofproject/linux/files/9188177/dmesg_tgl_ipc_dbg.log)

Well, the stream still stops though...

I managed to get an error with dyndbg. 1. boot 2. systemctl --user restart pipewire-pulse pipewire wireplumber 3. play song in audacious 4. stream stops 5. systemctl --user restart pipewire-pulse...