Comment # 3 on bug 1191247 from Takashi Iwai

Thanks for quick testing.

Then could you load alsa-info.sh output from the patched kernel after
suspend/resume?  Run the script with --no-upload, and use Bugzilla attachment.

Also, for verifying whether it's a driver PM problem, try the following:

- Before and after suspend/resume, try to play without PA, PW or JACK but via
ALSA native API, e.g.
   % aplay -Dplughw:XXX foo.wav
  where XXX is the card number like 0, 1, 2, corresponding to the device index
appearing in /proc/asound/cards.

  I suppose this also results in the mono output?

- Get alsa-info.sh outputs before and after the suspend/resume.

- Check whether the problem persists with suspend-to-RAM and suspend-to-disk
(hibernate).  I assume you're dealing with the former, right?

- Last but not least, check whether re-loading the driver fixes the problem
without re-plugging.  You might need to go out of desktop session once, unload
snd-usb-audio module and re-load snd-usb-audio module on Linux console.