asyncq.c: q overrun, queuing locally

[Micah Gersten]

Binary package hint: pulseaudio

Apr 2 19:36:25 defiant pulseaudio[5066]: asyncq.c: q overrun, queuing locally
Apr 2 19:36:40 defiant last message repeated 38398 times
--
Apr 2 19:36:41 defiant pulseaudio[5066]: asyncq.c: q overrun, queuing locally
Apr 2 19:36:45 defiant last message repeated 5111 times
Apr 2 19:36:45 defiant pulseaudio[5066]: cpulimit.c: Received request to terminate due to CPU overload.
Apr 2 19:36:45 defiant pulseaudio[5066]: asyncq.c: q overrun, queuing locally
Apr 2 19:36:45 defiant last message repeated 1735 times
Apr 2 19:36:45 defiant kernel: [23369.624826] pulseaudio[5066]: segfault at 7f4a1b005fa0 ip 00007f4a1b005fa0 sp 00007fff2b06f698 error 14 in module-null-sink.so[7f4a1b628000+3000]

ProblemType: Bug
Architecture: amd64
DistroRelease: Ubuntu 9.04
NonfreeKernelModules: fglrx
Package: pulseaudio 1:0.9.14-0ubuntu16
ProcEnviron:
 LANG=en_US.UTF-8
 SHELL=/bin/bash
SourcePackage: pulseaudio
Uname: Linux 2.6.28-11-generic x86_64

[Micah Gersten (micahg)]

I was attempting to start a Virtual machine in virtualbox.

[Daniel T Chen (crimsun)]

Please attach output from running the bash script (not sh but bash)
http://www.alsa-project.org/alsa-info.sh

On Apr 2, 2009 8:50 PM, "Micah Gersten" <email address hidden>
wrote:

I was attempting to start a Virtual machine in virtualbox.

-- asyncq.c: q overrun, queuing locally
https://bugs.launchpad.net/bugs/354220 You received this b...

[Micah Gersten (micahg)]

Daniel T Chen wrote:
> Please attach output from running the bash script (not sh but bash)
> http://www.alsa-project.org/alsa-info.sh
>
> On Apr 2, 2009 8:50 PM, "Micah Gersten" <email address hidden>
> wrote:
>
> I was attempting to start a Virtual machine in virtualbox.
>
> -- asyncq.c: q overrun, queuing locally
> https://bugs.launchpad.net/bugs/354220 You received this b...
>
>
Here's the requested information.

[Daniel T Chen (crimsun)]

Does this symptom appear only when starting vbox?

On Apr 2, 2009 9:15 PM, "Micah Gersten" <email address hidden>
wrote:

Daniel T Chen wrote: > Please attach output from running the bash script
(not sh but bash) > http://...
Here's the requested information.

** Attachment added: "alsa-info.txt"
  http://launchpadlibrarian.net/24732409/alsa-info.txt

-- asyncq.c: q overrun, queuing locally
https://bugs.launchpad.net/bugs/354220

You received this bug notification because you are a member of Ubuntu Audio
Team, which is subscribe...

[Micah Gersten (micahg)]

Daniel T Chen wrote:
> Does this symptom appear only when starting vbox?
>
> On Apr 2, 2009 9:15 PM, "Micah Gersten" <email address hidden>
> wrote:
>
> Daniel T Chen wrote: > Please attach output from running the bash script
> (not sh but bash) > http://...
> Here's the requested information.
>
>
> ** Attachment added: "alsa-info.txt"
> http://launchpadlibrarian.net/24732409/alsa-info.txt
>
> -- asyncq.c: q overrun, queuing locally
> https://bugs.launchpad.net/bugs/354220
>
> You received this bug notification because you are a member of Ubuntu Audio
> Team, which is subscribe...
>
>
No, it's happened a few times over the past couple of days per syslog.

[Micah Gersten (micahg)]

Micah Gersten wrote:
> Daniel T Chen wrote:
>
>> Does this symptom appear only when starting vbox?
>>
>> On Apr 2, 2009 9:15 PM, "Micah Gersten" <email address hidden>
>> wrote:
>>
>> Daniel T Chen wrote: > Please attach output from running the bash script
>> (not sh but bash) > http://...
>> Here's the requested information.
>>
>>
>> ** Attachment added: "alsa-info.txt"
>> http://launchpadlibrarian.net/24732409/alsa-info.txt
>>
>> -- asyncq.c: q overrun, queuing locally
>> https://bugs.launchpad.net/bugs/354220
>>
>> You received this bug notification because you are a member of Ubuntu Audio
>> Team, which is subscribe...
>>
>>
>>
> No, it's happened a few times over the past couple of days per syslog.
>
>
I take that back, it does seem to just be happening when I run virtualbox.
Here's all the related errors:
Apr 6 21:26:32 defiant pulseaudio[4913]: asyncq.c: q overrun, queuing
locally
Apr 6 21:26:46 defiant last message repeated 12880 times
Apr 6 21:26:46 defiant pulseaudio[4913]: cpulimit.c: Received request
to terminate due to CPU overload.
Apr 6 21:26:46 defiant pulseaudio[4913]: asyncq.c: q overrun, queuing
locally
Apr 6 21:26:46 defiant last message repeated 2150 times
Apr 6 21:26:47 defiant pulseaudio[4913]: memblock.c: Memory pool
destroyed but not all memory blocks freed! 3135 remain.

[Emmanuel JOORIS (ejooris)]

I have this problem too, when a computer share his local sound peripheral with discover and no auth with an another computer who have the automatic connect to discovered peripheral.

Ubuntu 9.04 amd64
RTP disabled

[quincy (dotcop)]

This too has been happening to me while listening to Pandora Radio.

[...]
Jun 3 08:33:49 catie pulseaudio[3366]: protocol-native.c: Failed to push data into queue
Jun 3 08:33:54 catie last message repeated 25650 times
Jun 3 08:33:54 catie pulseaudio[3366]protocol-native.c: Failed to push data into queue
Jun 3 08:33:54 catie pulseaudio[3366]: protocol-native.c: Failed to push data into queue
Jun 3 08:33:56 catie last message repeated 18786 times

~$ uname -a
Linux catie 2.6.28-11-generic #42-Ubuntu SMP Fri Apr 17 01:57:59 UTC 2009 i686 GNU/Linux

[Jeremy Visser (jeremy-visser)]

I get this when attempting to use module-bluetooth-device. When I try and load the module, PulseAudio hangs, with its last message being the "queuing locally" message, and it eventually eats up all the RAM.

[Matt Moen (matt-moen)]

Also running virtual box and seeing the same issue ... In my case it amounts to a DOS on my machine. The hard disk is trashing, which may be an indication that my ram is full and it's moving to swap. The system is unresponsive however so I can't look at what is consuming the resources. Sometimes it seems to recover, however often I'm forced to hard restart.

I did recently enable Audio in a virtual box machine that's just about always running, which corresponds to starting to see these issues. I've disabled that audio and i'll see if the problem recurs.

[Daniel T Chen (crimsun)]

On Wed, Aug 5, 2009 at 12:07 PM, Matt Moen<email address hidden> wrote:
> Also running virtual box and seeing the same issue ... In my case it
> amounts to a DOS on my machine. The hard disk is trashing, which may be
> an indication that my ram is full and it's moving to swap. The system is
> unresponsive however so I can't look at what is consuming the resources.
> Sometimes it seems to recover, however often I'm forced to hard restart.

There's a fix for VMs that I need to pull for linux.

[BrowneR (chris-scotland)]

Seeing these errors when playing audio from Spotify running under Wine. I don't see them except when using wine. winecfg is setup to play through esd. Sound is fine except for this error and 100% cpu.

[Daniel T Chen (crimsun)]

This symptom should be fixed in current Karmic + the version of PulseAudio in the ~ubuntu-audio-dev PPA. Please confirm.

[Philippe Delodder (lodder)]

Could it cause a system lock up?

[Mark (darck1)]

I'm getting the same error. Mine mostly happens when I play X3: Reunion (not Wine - native). Here is my alsa info:
http://www.alsa-project.org/db/?f=323c83769abcad15c4b81bbd7d048d296a30e003

[Daniel T Chen (crimsun)]

What audio backend does X3: Reunion use? Is it SDL-based?

[chocolateboy (chocolateboy)]

@Daniel T Chen: I get the same error under Karmic. The only thing mentioned here that I'm using is Spotify under wine (using the winepulse PPA), although the error occurs without Spotify running.

[Daniel T Chen (crimsun)]

In PA 0.9.20, glitch-free is disabled if PA detects that it's running in a
vm. You may want to test this new version (in the ubuntu-audio-dev PPA
available for Karmic).

On Nov 13, 2009 3:11 PM, "chocolateboy" <email address hidden> wrote:

@Daniel T Chen: I get the same error under Karmic. The only thing
mentioned here that I'm using is Spotify under wine (using the winepulse
PPA), although the error occurs without Spotify running.

-- asyncq.c: q overrun, queuing locally
https://bugs.launchpad.net/bugs/354220 You received this b...

Status in “pulseaudio” package in Ubuntu: New Bug description:

Binary package hint: pulseaudio Apr 2 19:36:25 defiant pulseaudio[5066]:
asyncq.c: q overrun, queu...

[Daniel Asarnow (dasarnow)]

Just got this as I attempted to use VLC.
I installed PulseAudio /today/. Karmic AMD64 with custom kernel, everything otherwise up to date.

My error output includes the line:
W: ratelimit.c: 129 events suppressed

129 was just a representative line, the number of events increases as the bug continuous. I managed to kill pulseaudio after a couple of minutes.

No VMs running, no wine apps running...

[FiX (fix-kowalski)]

Also facing this exact same issue, with vbox.

- Karmic/x86_64
- Dual-core Xeon 2.8 GHZ (in case it matters, for SMP locks...)
- virtualbox-3.1 (3.1.0-55467_Ubuntu_karmic)
- linux-image-2.6.31-16-generic (2.6.31-16.52)
- pulseaudio (1:0.9.19-0ubuntu4)

I get tons of the below:
[,,,]
Dec 9 11:35:41 tonarra pulseaudio[2143]: asyncq.c: q overrun, queuing locally
Dec 9 11:35:46 tonarra pulseaudio[2143]: ratelimit.c: 78084 events suppressed
Dec 9 11:35:51 tonarra pulseaudio[2143]: ratelimit.c: 156553 events suppressed
Dec 9 11:35:51 tonarra pulseaudio[2143]: asyncq.c: q overrun, queuing locally
Dec 9 11:35:51 tonarra pulseaudio[2143]: asyncq.c: q overrun, queuing locally
Dec 9 11:35:51 tonarra pulseaudio[2143]: asyncq.c: q overrun, queuing locally
Dec 9 11:35:51 tonarra pulseaudio[2143]: asyncq.c: q overrun, queuing locally
[...]

The pulseaudio process was taking 115% CPU (more than one core). Kill -15 did not stop it. Kill -9 produced the following OOPS (not sure it is related).

Dec 9 11:36:35 tonarra kernel: [88441.160129] wpa_supplican D 00000000ffffffff 0 1175 1 0x00000000
Dec 9 11:36:35 tonarra kernel: [88441.160140] ffff88019ae15a38 0000000000000082 ffff88019ae159b8 0000000000015880
Dec 9 11:36:35 tonarra kernel: [88441.160151] ffff8801b0cf1a60 0000000000015880 0000000000015880 0000000000015880
Dec 9 11:36:35 tonarra kernel: [88441.160160] 0000000000015880 ffff8801b0cf1a60 0000000000015880 0000000000015880
Dec 9 11:36:35 tonarra kernel: [88441.160170] Call Trace:
Dec 9 11:36:35 tonarra kernel: [88441.160189] [<ffffffff815286b7>] __mutex_lock_slowpath+0xd7/0x160
Dec 9 11:36:35 tonarra kernel: [88441.160200] [<ffffffff8142a241>] ? sock_alloc_send_pskb+0x1c1/0x310
Dec 9 11:36:35 tonarra kernel: [88441.160208] [<ffffffff815285b6>] mutex_lock+0x26/0x50
Dec 9 11:36:35 tonarra kernel: [88441.160250] [<ffffffffa0b4ecc5>] iwl_mac_config+0x35/0x2a0 [iwlcore]
Dec 9 11:36:35 tonarra kernel: [88441.160261] [<ffffffff81045d09>] ? __wake_up_common+0x59/0x90
Dec 9 11:36:35 tonarra kernel: [88441.160293] [<ffffffffa00f6af5>] ieee80211_hw_config+0x85/0xd0 [mac80211]
Dec 9 11:36:35 tonarra kernel: [88441.160323] [<ffffffffa0103199>] ieee80211_recalc_idle+0x49/0x60 [mac80211]
Dec 9 11:36:35 tonarra kernel: [88441.160351] [<ffffffffa00fb80d>] __ieee80211_start_scan+0xad/0x230 [mac80211]
Dec 9 11:36:35 tonarra kernel: [88441.160379] [<ffffffffa00fba85>] ieee80211_request_scan+0x35/0x60 [mac80211]
Dec 9 11:36:35 tonarra kernel: [88441.160408] [<ffffffffa01072b0>] ieee80211_scan+0x60/0x70 [mac80211]
Dec 9 11:36:35 tonarra kernel: [88441.160436] [<ffffffffa00c94db>] cfg80211_wext_siwscan+0x18b/0x210 [cfg80211]
Dec 9 11:36:35 tonarra kernel: [88441.160446] [<ffffffff8150a89c>] ioctl_standard_iw_point+0x1bc/0x370
Dec 9 11:36:35 tonarra kernel: [88441.160468] [<ffffffffa00c9350>] ? cfg80211_wext_siwscan+0x0/0x210 [cfg80211]
Dec 9 11:36:35 tonarra kernel: [88441.160477] [<ffffffff8150aa50>] ? ioctl_standard_call+0x0/0xd0
Dec 9 11:36:35 tonarra kernel: [88441.160485] [<ffffffff81509ac0>] ? ioctl_private_call+0x0/0xa0
Dec 9 11:36:35 tonarra kernel: [88441.160492] [<ffffffff8150ab01>] ioctl_standard_call+0xb1/0xd0
...

[Daniel T Chen (crimsun)]

Are you able to reproduce this symptom using the most current daily-live desktop Lucid image?

[FiX (fix-kowalski)]

I did not test with the latest alpha, because that would require upgrading my personal productivity system to a non release version. That being said, I no longer see these error messages. They likely were fixed by some update. Here are my new versions:

virtualbox-3.1 3.1.4-57640_Ubuntu_karmic (upgraded since pb)
pulseaudio 1:0.9.19-0ubuntu4.1 (not upgraded since pb)
linux-image-2.6.31-19-generic 2.6.31-19.56 (upgraded since pb)

Pb automagically disappeared. Either fixed in kernel/vbox or worked-around by vbox team. You can close the defect as far as I am concerned.

[BrowneR (chris-scotland)]

I have not noticed these errors on my system since upgrading to Karmic. They were present when using spotify under wine but have since disappeared. This seems fixed in my opinion.

[Scott MacKay (scottmackay)]

I have the same problem, has it in 9 and now with a just fresh download of 10, installed on a pretty generic box.

It happens with my logitech USB headset. Under 9 it seems to happen when I used a media player...it also keeps my machine from powering down without a hardware reset.

Under 10 I had been able to listen to alot using the headset, but as soon as I tried to use it for Skype, selecting it as the output device for mic, these errors started up and seem to have hung the sound system

-Scott

[Shock (mmiron)]

Just happened to me on up to date Lucid. indicator-sound-service is now using 100% cpu and sound is no longer working.
If I kill indicator-sound-service it restarts and continues using 100% cpu. Turns out pulseaudio is dead. If I run "pulseaudio -D" indicator-sound-service cpu usage drops to normal but the sound indicator menu is not appearing when I click on the indicator. Let me know how I can help get this fixed.

[Shock (mmiron)]

Sometime after running "pulseaudio -D" I noticed this in syslog:
pulseaudio[29286]: alsa-sink.c: ALSA woke us up to write new data to the device, but there was actually nothing to write!
pulseaudio[29286]: alsa-sink.c: Most likely this is a bug in the ALSA driver 'snd_emu10k1'. Please report this issue to the ALSA developers.
pulseaudio[29286]: alsa-sink.c: We were woken up with POLLOUT set -- however a subsequent snd_pcm_avail() returned 0 or another value < min_avail.

I hope it's relevant.

[David Henningsson (diwic)]

Hi folks,

This problem has been going on for far too long, and last week I spent a lot of time trying to nail it down, and trying to find the causes for these problems. I'm talking of the case where PulseAudio get stuck in rewind after rewind, finally getting killed by the kernel for using too much RT Prio time. (It might be that for some people, PA crashes for other reasons.)
Unfortunately, it's a quite complex problem, probably without one easy fix that works for all. I have also had some problems reproducing it reliably here.

Anyway, I have made a test ppa with five fixes in; two in gstreamer and three in pulseaudio. My guess is that using them will improve the situation for some people, but perhaps not for everyone. The ppa is here: https://launchpad.net/~diwic/+archive/fighting-rewinds and is meant for users running Maverick. I would be glad for some feedback of whether they improve the situation or not.
To test: install either the gstreamer pulseaudio plugin, pulseaudio, or both, from my ppa. (For pulseaudio, make sure you upgrade both libpulse and pulseaudio.) Then logout/login (or reboot).

[Launchpad Janitor (janitor)]

[Expired for pulseaudio (Ubuntu) because there has been no activity for 60 days.]

[Shock (mmiron)]

Sorry for the late response, somehow I missed the comment about the fixes in the ppa. I've just installed the packages from the ppa and will watch out for other occurrences.

FYI, the problem is still present on my system:

$ grep asyncq.c messages*
messages:Apr 19 16:54:03 shock pulseaudio[2348]: asyncq.c: q overrun, queuing locally
messages:Apr 19 16:56:15 shock pulseaudio[2348]: asyncq.c: q overrun, queuing locally
messages:Apr 19 16:56:20 shock pulseaudio[2348]: asyncq.c: q overrun, queuing locally
messages:Apr 19 16:56:48 shock pulseaudio[14315]: asyncq.c: q overrun, queuing locally
messages:Apr 19 16:57:49 shock pulseaudio[14386]: asyncq.c: q overrun, queuing locally
messages.1:Apr 14 01:41:55 shock pulseaudio[2321]: asyncq.c: q overrun, queuing locally
messages.1:Apr 14 01:42:00 shock pulseaudio[2321]: asyncq.c: q overrun, queuing locally
messages.1:Apr 14 01:42:06 shock pulseaudio[2321]: asyncq.c: q overrun, queuing locally
messages.1:Apr 14 01:42:12 shock pulseaudio[2321]: asyncq.c: q overrun, queuing locally
messages.1:Apr 14 01:42:17 shock pulseaudio[2321]: asyncq.c: q overrun, queuing locally

[Vidar Braut Haarr (vhaarr+launchpad)]

This is still a problem.
My PA keeps dying on me every day with the "async q overrun" stuff.

[Vidar Braut Haarr (vhaarr+launchpad)]

Some developer needs to jump in here and tell us what information he needs; I'm changing this bug from "Incomplete" to "New", because I have no idea what information to provide.

[Shock (mmiron)]

In my case, I'm not able to find the "asyncq.c: q overrun, queuing locally" message in logs after Apr 19th. I installed the packages from the ppa on Apr 21st, so it's been almost a month without the message in the logs.

I'm not sure if we can conclude that the bug is fixed solely on my account, but I think it's probable.

[Shock (mmiron)]

Vidar, what version of the pulseaudio package do you have installed?

[Vidar Braut Haarr (vhaarr+launchpad)]

I have the Ubuntu oneiric packages; 1:0.9.22+stable-queue-24-g67d18-0ubuntu4.

[Shock (mmiron)]

Vidar, you need to install the packages from the ppa mentioned in comment #28 and test using those. I have 1:0.9.22+stable-queue-18-geb966-0ubuntu3~diwic~maverick installed.

[Vidar Braut Haarr (vhaarr+launchpad)]

Okay, thanks I will!
That won't solve this bug though, it will just work around the problem. The patches from that PPA - if they work - need to be incorporated into pulseaudio upstream or the ubuntu repos.