Comment 9 for bug 2061091

Revision history for this message
Richard van der Hoff (richvdh) wrote :

> And this call trace looks is related to current issue, I guess snd_power_ref_and_wait was waiting for snd_card_disconnect which wakes up power_sleep at the end of the code, but snd_card_disconnect -> snd_device_disconnect_all -> snd_pcm_dev_disconnect was blocked for some reason.

Yes, interesting. Good spot.

> Looks suspend/resume can be completed sometimes.

Yes. I think what causes the problem is when I disconnect from my USB dock. After that, I can't suspend until I reboot. Once I reboot, it can suspend again until I once more disconnect the dock.

> Could you rebuild kernel with CONFIG_PROVE_LOCKING option to discover locking related deadlocks? Then upload the log after reproduce the issue by shut down laptop with the new kernel. Also please attach the output of lsusb given it could be usb relevant.
>
> And it also would be helpful to try with latest noble kernel or recent upstream kernel, thanks.

Sure. I'll try an upstream kernel with CONFIG_PROVE_LOCKING.