[pulseaudio-tickets] [Bug 106557] power on / plugging in E-MU 0202 USB hangs up pulseaudio and USB bus

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=106557

--- Comment #4 from Dmitrii Sutiagin <f3flight at gmail.com> ---
(In reply to Tanu Kaskinen from comment #3)
> Apparently autospawning is enabled for gdm. That's weird. ... Is there some
> gdm specific configuration in
> /var/lib/gdm/.config/pulse/client.conf? 

[dsutyagin at pulsar ~]$ sudo find / -name 'client.conf' 2>/dev/null | grep pulse
| xargs -I@ bash -c 'echo @; grep autospawn @ | sed "s/^/--> /"'
/etc/pulse/client.conf
--> ; autospawn = yes
/home/dsutyagin/.config/pulse/client.conf
--> # Disable autospawning the PulseAudio daemon
--> autospawn = no
/home/dsutyagin/.local/share/flatpak/runtime/org.freedesktop.Sdk/x86_64/1.4/d6a8c46e30e0518c8f6c8278edf0bd59800193494bbdaeee7393646385247445/files/etc/pulse/client.conf
--> ; autospawn = yes

as you can see, gdm doesn't have "client.conf", to be more specific, here's the
contents:

[user at system ~]$ sudo ls -l /var/lib/gdm/.config/pulse
total 56
-rw-r--r-- 1 gdm gdm 20480 May  3 15:23
1475c2dd14c9474ab7bab828a842239e-card-database.tdb
-rw-r--r-- 1 gdm gdm    43 Jun 11 09:53
1475c2dd14c9474ab7bab828a842239e-default-sink
-rw-r--r-- 1 gdm gdm    42 Jun 11 09:53
1475c2dd14c9474ab7bab828a842239e-default-source
-rw-r--r-- 1 gdm gdm 12288 May  3 15:23
1475c2dd14c9474ab7bab828a842239e-device-volumes.tdb
-rw-r--r-- 1 gdm gdm 12288 May 29 19:48
1475c2dd14c9474ab7bab828a842239e-stream-volumes.tdb
-rw------- 1 gdm gdm   256 May  2  2017 cookie

Maybe the default is "enabled" so that's why. GDM's cwd is / so idk if it
should read from /var/lib/gdm/.config/pulse or /etc/pulse

[dsutyagin at pulsar ~]$ pgrep -f /usr/bin/gdm | sudo xargs -I@ ls -ld /proc/@/cwd
lrwxrwxrwx 1 root root 0 Jun 11 14:14 /proc/542/cwd -> /

But anyway.

> The log stops rather abruptly in the middle of the probing process after
> getting "Input/output error" from snd_pcm_hw_params(). It seems that
> snd_pcm_close() is hanging, because otherwise there should be more logs.

I've stopped collecting logs at that point, they do keep adding slowly, but
every call in pulseaudio log ends with long delay and then timeout, so further
logs won't add much (unless there's some useful debug info at the end of all
the attempts).

> To me this seems quite clearly a kernel bug, and I don't think I have much
> more to contribute here. You should report the problem to the alsa
> developers, here are some instructions:
> http://alsa-project.org/main/index.php/Bug_Tracking

I will try filing a bug to alsa, thank you.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/pulseaudio-bugs/attachments/20180611/b1b32b64/attachment.html>