Comment # 3 on bug 1040048 from Moritz Rakow

Lost audio again, this time after suspending in standard GNOME. When waking up
the machine, audio was lost in GNOME.

Differences to before:

*   This time, no microphone was detected at all.
*   The audio device was named "Dummy-Ausgabe". Note that neither my system
language nor user language are set to German, but the user language had been
set to German previously.
*   After logging out, sound was working fine on the login screen.
*   Loggin back in, sound is fine.

I've checked for running pulseaudio, as you suggested, on several occasions.


## Summary of findings

*   After suspend/wake-up, there was an old pulseaudio that did not work. It is
plausible that when that was started the user language *had* been German. This
would explain the device name "Dummy-Ausgabe". This process had still been
alive after logout. Not sure when it died.
*   gdm started its own pulseaudio, which worked. Note that the old pulseaudio
was still running at that time.
*   On logging in at 09:20, a new pulseaudio was started that also works. At
this point, the old pulseaudio is not running anymore.


## Output of `ps aux | grep pulseaudio` at different times

My user name is "ram".

After suspend/wake-up when audio was dead: 

```
ram       2256  1.3  1.1 792188 40816 ?        S<l  May23  31:15
/usr/bin/pulseaudio --start --log-target=syslog
ram      10116  0.0  0.0  10548  1620 pts/0    S+   09:15   0:00 grep
--color=auto pulseaudio
```

Then in VT1, after log-out, when sound was working fine in gdm:

```
ram       2256  1.3  1.1 398948 40516 ?        S<l  May23  31:21
/usr/bin/pulseaudio --start --log-target=syslog
gdm      10354  0.0  0.3 567220 11156 ?        S<l  09:15   0:00
/usr/bin/pulseaudio --start --log-target=syslog
ram      10502  0.0  0.0  10548  1608 tty1     S+   09:16   0:00 grep
--color=auto pulseaudio
```

Finally, when audio was back after logging in again:

```
ram       2210  0.0  0.3 518420 11964 ?        S<l  09:20   0:00
/usr/bin/pulseaudio --start --log-target=syslog
ram       3057  0.0  0.0  10548  1596 pts/0    S+   09:35   0:00 grep
--color=auto pulseaudio
```

Should we change the name of the bug report to include this after-suspend
scenario?