Karmic: frequent freezes

Is this issue reproducible with
linux-backports-modules-alsa-karmic-generic installed (and after a
fresh boot)?

Hi Daniel. If you're asking me, I don't have a clue what "linux-backports-modules-alsa-karmic-generic" is. I've never reproduced an issue, but just reported bugs through Apport. The system just freezes randomly. What should I do exactly?

Install it.

On Nov 9, 2009 2:15 PM, "lelamal" <email address hidden> wrote:

Hi Daniel. If you're asking me, I don't have a clue what "linux-
backports-modules-alsa-karmic-generic" is. I've never reproduced an
issue, but just reported bugs through Apport. The system just freezes
randomly. What should I do exactly?

-- Karmic: frequent freezes https://bugs.launchpad.net/bugs/479296 You
received this bug notificat...

OK. Installed. And rebooted. Should I just wait, now?

Unfortunately, yes.

Alright, thank you for your support so far!

I have had the same problem with my HP Pavilion 9600 Laptop since upgrading to Karmic. The system freezes at a random moment, the mouse can move but that's all, nothing can be clicked. When I've had System Monitor Open or top in a terminal they also freeze. I've not been able to get at /var/log/syslog yet. I have given up and reverted to 8.04 LTS and so far all seems to be working again.

I am having the same problem. The log file reports the same error as of Nov 7, 2009. Today (11th) the crashes seem more random now, and it is not reporting a problem with pulseaudio.

9.10, x64, onboard and discrete sound card. AMD Opteron 170 (939). This was a clean install.

I have done the advice of Daniel, installed the package "linux-backports-modules-alsa-karmic-generic". I guess I'll wait too.

Alternatively, you can get at the /var/log/syslog (and all other logs) at System > Administration > Log File Viewer.

I am still getting kernel panics. It has even reset itself without warning (no proper shutdown, as if I had pushed the reset button on a working system).

Well I spoke too soon and I did have the same problem with 8.04 LTS (i.e. a freeze, but mouse moving), but actually I now think that all my problems with the HP Pavilion laptop have been related to the NVIDIA graphics. I took apart the laptop as far as removing the system board, and rebuilt it. Then I upgraded the BIOS. I've done a fresh install of 9.10. Its when I install the NVIDIA Hardware drivers, either 173 or 185 that the system stops working. If I don't install those drivers it seems to work OK. Luckily I don't need high performance graphics, at the moment. I don't think any of these problems are related to pulseaudio.

I'm seeing the same messages in the syslog and also have been getting random cold restarts (no warning, just goes away) but no freezing so far.

As a side note, I've been playing around with alsa-base.conf, trying different models for an intel-hda ALC883 chipset (trying to fix the never-working microphone), I don't know if it can have anything to do with it.

@LeBurt
Are you using any non-Free/proprietary drivers?

I believe I am running into this bug as well. After doing a fresh install to upgrade from intrepid to karmic, when I watch a video the box will randomly freeze and display the "pulseaudio[1564]: ratelimit.c: 66 events suppressed" error. I have never seen any of the kernel errors, so I am not 100% sure if this is the same. I tried installing the linux-backports-modules-alsa-karmic-generic package with no change. It got frustrating to the point that I tried removing pulse audio. This stopped the pulseaudio errors, but the freezes persist, though now with no errors showing up in the log.

My setup is an AN-M2HD motherboard using the onboard audio. I am using the proprietary Nvidia graphics drivers.

@Belgarth Please note that you need to reproduce this bug while using PA and
the Free nv driver. The fact that your system still freezes without using PA
and while using the nvidia driver means that it has nothing to do with PA.

On Nov 13, 2009 12:50 AM, "Belgarth" <email address hidden> wrote:

I believe I am running into this bug as well. After doing a fresh
install to upgrade from intrepid to karmic, when I watch a video the box
will randomly freeze and display the "pulseaudio[1564]: ratelimit.c: 66
events suppressed" error. I have never seen any of the kernel errors, so
I am not 100% sure if this is the same. I tried installing the linux-
backports-modules-alsa-karmic-generic package with no change. It got
frustrating to the point that I tried removing pulse audio. This stopped
the pulseaudio errors, but the freezes persist, though now with no
errors showing up in the log.

My setup is an AN-M2HD motherboard using the onboard audio. I am using
the proprietary Nvidia graphics drivers.

-- Karmic: frequent freezes https://bugs.launchpad.net/bugs/479296 You
received this bug notificat...

Daniel: Yes, the Nvidia graphics driver version 185.

I had another restart last night after installing linux-backports-modules-alsa-karmic-generic.

Here's another observation about the problem:

After a fresh restart I went to the log viewer and had the syslog displayed. Everything appears ok. Then I opened up sound preferences and immediately the syslog started showing ratelimit.c messages. When I closed sound prefs, messages stopped.

I removed the proprietary drivers and the hangs/freezes stoped, but the ratelimit errors continued. I was originally using the 185 drivers. this morning I installed the 173 ones. So far no hangs, though the ratelimit errors still persist.

The ratelimit messages are not errors.

I spoke too soon. Still getting the freezes with 173 drivers.

But these are the *proprietary* Nvidia 173 drivers, no?

I think I might be getting this bug as well.
After upgrading to Karmic I get random freezes.

Most often everything but the mouse pointer freezes. Not even caps lock works.
Pressing the power button on the computer does not initialize a shutdown as it usually does.

When I get the problem I could see the screen flashing black once very rapidly.
I have not found any useful info in my syslogs...

I'm on the 185 NVidia driver.

I have a very similar problem. When watching video, everything is working fine for a while, but then the sound gets all choppy while the video freezes (mplayer). Fortunately, X doesn't freeze up and my machine doesn't reboot. I just have to close and re-start mplayer and wait for it to happen again. This happens every couple of minutes, but sometimes doesn't happen for about 10 minutes or so. I looked at /var/log/messages to see this:

----------------
... snip ...
Nov 16 22:21:53 koala pulseaudio[1312]: ratelimit.c: 7 events suppressed
Nov 16 22:21:58 koala pulseaudio[1312]: ratelimit.c: 5515 events suppressed
Nov 16 22:22:03 koala pulseaudio[1312]: ratelimit.c: 3625 events suppressed
Nov 16 22:22:09 koala pulseaudio[1312]: ratelimit.c: 449 events suppressed
Nov 16 22:23:10 koala pulseaudio[1312]: ratelimit.c: 1467 events suppressed
Nov 16 22:23:15 koala pulseaudio[1312]: ratelimit.c: 5020 events suppressed
Nov 16 22:23:21 koala pulseaudio[1312]: ratelimit.c: 2735 events suppressed
Nov 16 22:23:26 koala pulseaudio[1312]: ratelimit.c: 3 events suppressed
Nov 16 22:25:24 koala pulseaudio[1312]: ratelimit.c: 2262 events suppressed
Nov 16 22:25:30 koala pulseaudio[1312]: ratelimit.c: 5261 events suppressed
Nov 16 22:25:35 koala pulseaudio[1312]: ratelimit.c: 4069 events suppressed
----------------

There are *many* more, but I don't think posting more will be more helpful.

I'm using the Open Source ATI driver specified in xorg.conf as "ati" with a "ATI Technologies Inc Radeon RV100 QY [Radeon 7000/VE]" (as reported by lspci) graphics card, so linking it to the proprietary nvidia driver doesn't seem to be correct.

Any info I can provide to make this report more useful? I have not tried installing "linux-backports-modules-alsa-karmic-generic" since a user above said it didn't make a difference.

Thanks.

Nov 17 10:50:03 mymachine pulseaudio[26963]: module.c: Failed to load module "module-native-protocol-unix" (argument: ""): initialization failed.
Nov 17 10:50:03 mymachine pulseaudio[26963]: main.c: Module load failed.
Nov 17 10:50:03 mymachine pulseaudio[26963]: main.c: Failed to initialize daemon.
Nov 17 10:50:04 mymachine pulseaudio[26961]: main.c: Daemon startup failed.
Nov 17 10:50:09 mymachine pulseaudio[26973]: pid.c: Daemon already running.
Nov 17 10:50:38 mymachine pulseaudio[26969]: ratelimit.c: 75 events suppressed
Nov 17 10:51:16 mymachine pulseaudio[26969]: ratelimit.c: 79 events suppressed
Nov 17 10:51:34 mymachine pulseaudio[26969]: ratelimit.c: 86 events suppressed
Nov 17 10:51:56 mymachine pulseaudio[26969]: ratelimit.c: 94 events suppressed
Nov 17 10:52:34 mymachine pulseaudio[26969]: ratelimit.c: 96 events suppressed
Nov 17 10:52:39 mymachine pulseaudio[26969]: ratelimit.c: 96 events suppressed
Nov 17 10:55:52 mymachine pulseaudio[26969]: ratelimit.c: 11 events suppressed
Nov 17 10:58:21 mymachine pulseaudio[26969]: alsa-sink.c: ALSA woke us up to write new data to the device, but there was actually nothing to write!
Nov 17 10:58:21 mymachine pulseaudio[26969]: alsa-sink.c: Most likely this is a bug in the ALSA driver 'snd_usb_audio'. Please report this issue to the ALSA developers.
Nov 17 10:58:21 mymachine pulseaudio[26969]: alsa-sink.c: We were woken up with POLLOUT set -- however a subsequent snd_pcm_avail() returned 0 or another value < min_avail.
Nov 17 12:30:13 mymachine pulseaudio[26969]: ratelimit.c: 213 events suppressed
Nov 17 16:25:55 mymachine pulseaudio[26969]: ratelimit.c: 457 events suppressed
Nov 17 17:02:17 mymachine pulseaudio[26969]: ratelimit.c: 461 events suppressed

$ uname -srm
Linux 2.6.31-14-generic x86_64

$ dpkg -l | grep pulse
ii gstreamer0.10-pulseaudio 0.10.16-1ubuntu3 GStreamer plugin for PulseAudio
ii libcanberra-pulse 0.15-0ubuntu7 a simple abstract interface for playing even
ii libpulse-browse0 1:0.9.19-0ubuntu4 PulseAudio client libraries (zeroconf suppor
ii libpulse-mainloop-glib0 1:0.9.19-0ubuntu4 PulseAudio client libraries (glib support)
ii libpulse0 1:0.9.19-0ubuntu4 PulseAudio client libraries
ii libpulsecore9 1:0.9.14-0ubuntu20.2 PulseAudio sound server core
ii pulseaudio 1:0.9.19-0ubuntu4 PulseAudio sound server
ii pulseaudio-esound-compat 1:0.9.19-0ubuntu4 PulseAudio ESD compatibility layer
ii pulseaudio-module-bluetooth 1:0.9.19-0ubuntu4 Bluetooth module for PulseAudio sound server
ii pulseaudio-module-gconf 1:0.9.19-0ubuntu4 GConf module for PulseAudio sound server
ii pulseaudio-module-hal 1:0.9.14-0ubuntu20.2 HAL device detection module for Pulse...

Similar freezes here since I upgraded to Karmic. I never had such a problem with Jaunty (or any previous version).
The typical freeze is a severe one (everything really dead, keyboard LEDs don't toggle, soft power off does nothing, forced power off needed).
The last items present in the log is always something like this:
Nov 18 07:49:18 Philinux pulseaudio[1854]: ratelimit.c: 5 events suppressed
and the next message is when I realized the machine was dead and rebooted it:
Nov 18 08:38:19 Philinux kernel: imklog 4.2.0, log source = /var/run/rsyslog/kmsg started.

Nov 18 10:29:49 Philinux pulseaudio[1825]: ratelimit.c: 229 events suppressed
Nov 18 10:56:55 Philinux kernel: imklog 4.2.0, log source = /var/run/rsyslog/kmsg started.

Nov 18 11:09:39 Philinux pulseaudio[1853]: ratelimit.c: 1 events suppressed
Nov 18 12:40:22 Philinux kernel: imklog 4.2.0, log source = /var/run/rsyslog/kmsg started.

When those freezes happened, no sound application such as Rythmbox or other was active. Sound is only used by the default system sounds (like on error, etc.).
So far, the freezes didn't happen when I was actively working on the machine (keyboard / mouse), so I suspected a relation with the power management. The machine is configured to never sleep when inactive, "Spin down hard disks" is not checked, but the display is config'd to turn off after 5 minutes.

Finally, looking at past logs, the machine doesn't always freeze after a "pulseaudio[1854]: ratelimit.c: 5 events suppressed" message, but most often, the next consecutive lines in the log indicate some severe problem: Either a reboot or some segfault (usually nautilus, sometimes a deskbar applet). The segfault or reboot might happen immediately (same second) or several minutes after the "event suppressed" message, though.

Quite often, the "events suppressed" message is followed by a "Clocksource tsc unstable":
Nov 18 08:39:17 Philinux pulseaudio[1825]: ratelimit.c: 5 events suppressed
Nov 18 08:39:25 Philinux kernel: [ 92.232052] Clocksource tsc unstable (delta = -109978147 ns)

I'm not sure this makes sense or is any useful, just trying to identify patterns...

Hi Daniel and all, I'm back just to report that, since I installed linux-backports-modules-alsa-karmic-generic, the issue is no more reproducible at my end. During last week, I only experienced 2 freezes that were somewhat different from those frequent ones already described: one didn't involve any audio, plus I could move the mouse pointer (like described by Simon Nuttall at #7, for example), but nothing else worked; the other happened after logging out, before I could enter my password again (no mouse moving here). No revealing trace was left in the log these times, before freezing.

All in all, I can say things are much better now, although I'm not entirely sure that audio was the only cause of the issue, although apparently a major one (at least in my case). Also, I keep reading the same lines in the System Log (as detailed in my report), although I don't worry any longer, after you wrote that the ratelimit messages are not errors. But what about:

Nov 3 02:47:30 ECO pulseaudio[1564]: ratelimit.c: 153 events suppressed
Nov 3 02:49:19 ECO kernel: [59669.083978] [drm] LVDS-8: set mode 1024x768 d
Nov 3 02:49:21 ECO kernel: [59671.086380] [drm] DAC-6: set mode 640x480 0
Nov 3 02:49:22 ECO kernel: [59671.576399] [drm] DAC-6: set mode 640x480 0
Nov 3 02:49:22 ECO kernel: [59672.002693] i2c-adapter i2c-1: unable to read EDID block.

They always come together, although I clearly don't understand their origin, nor could I say that they are related to one another. Since 640x480 reminds me of the display resolution (mine is in fact 1024x768), does it have anything to do with the Graphics Device? And may it be responsible for the freezes? Many thanks again.

On Wed, Nov 18, 2009 at 8:51 AM, lelamal <email address hidden> wrote:
> All in all, I can say things are much better now, although I'm not
> entirely sure that audio was the only cause of the issue, although

It's unlikely and a red herring.

> Nov 3 02:47:30 ECO pulseaudio[1564]: ratelimit.c: 153 events suppressed
> Nov 3 02:49:19 ECO kernel: [59669.083978] [drm] LVDS-8: set mode 1024x768 d
> Nov 3 02:49:21 ECO kernel: [59671.086380] [drm] DAC-6: set mode 640x480 0
> Nov 3 02:49:22 ECO kernel: [59671.576399] [drm] DAC-6: set mode 640x480 0
> Nov 3 02:49:22 ECO kernel: [59672.002693] i2c-adapter i2c-1: unable to read EDID block.

Different subsystem -- graphics: i915/drm/intel X driver. No idea if
it is the culprit; graphics aren't my expertise.

Follow up to my own #24:

I installed
linux-backports-modules-alsa-karmic-generic
but it didn't prevent my station to freeze again (same messages in the log). :(

Another one found in this morning logs:
Nov 19 09:06:49 Philinux pulseaudio[1851]: ratelimit.c: 3 events suppressed
Nov 19 09:07:15 Philinux kernel: [ 147.347561] __ratelimit: 3 callbacks suppressed
Nov 19 09:07:15 Philinux kernel: [ 147.347575] deskbar-applet[2141]: segfault at 4fc00c ip 08094a8e sp bfab3cd0 error 6 in python2.6[8048000+1d3000]
Nov 19 09:07:57 Philinux kernel: [ 189.158323] apport-checkrep[2957]: segfault at 6d016c ip 08106ad8 sp bf9ea720 error 4 in python2.6[8048000+1d3000]
Nov 19 09:07:57 Philinux kernel: [ 189.508341] apport-checkrep[2958]: segfault at 7b5d1c ip 0810602a sp bf9b19d0 error 4 in python2.6[8048000+1d3000]

On Wed, 2009-11-18 at 14:36 +0000, Daniel T Chen wrote:
On Wed, Nov 18, 2009 at 8:51 AM, lelamal <email address hidden> wrote:
> > All in all, I can say things are much better now, although I'm not
> > entirely sure that audio was the only cause of the issue, although
>
> It's unlikely and a red herring.
>
And a red herring it was, indeed. I had one of those freezes just a couple of hours later. I was transferring MP3 files from an external EXT4 HD to a FAT32 USB pen-drive through Karmic main HD (in EXT4 as well, of course). Music got stuck between two notes, and system became unresponsive. I confess that whenever it happens, the feeling is similar to the one I had with Windows' Blue Screen Of Death, and it isn't pleasant at all.

I remember reading somewhere that EXT4 file systems had some issues with FAT and NTFS file systems. I reformatted the pen drive in EXT4, just in case. I don't know what else to think.

> Nov 3 02:47:30 ECO pulseaudio[1564]: ratelimit.c: 153 events suppressed
> > Nov 3 02:49:19 ECO kernel: [59669.083978] [drm] LVDS-8: set mode 1024x768 d
> > Nov 3 02:49:21 ECO kernel: [59671.086380] [drm] DAC-6: set mode 640x480 0
> > Nov 3 02:49:22 ECO kernel: [59671.576399] [drm] DAC-6: set mode 640x480 0
> > Nov 3 02:49:22 ECO kernel: [59672.002693] i2c-adapter i2c-1: unable to read EDID block.
>
> Different subsystem -- graphics: i915/drm/intel X driver. No idea if
> it is the culprit; graphics aren't my expertise.
>
Then might it not be wrong to file the bug under pulseaudio, if we don't even know what causes it exactly? If the solution you proposed didn't bring the expected result, and everybody seems to keep experiencing freezes under Karmic, maybe the culprits have yet to be identified.

I'm currently working through a myriad of suspicious events in my syslog to try and troubleshoot major system freezes etc. I am also suffering from these outputs and was wondering if they could cause server banning from VOIP services such as Ventrilo? Is something spamming the audio server and transmitting it over the VOIP connection? I'm also getting the sound stuttering at regular intervals.

This is a side issue that arises just before my system malfunctions with graphics problems, which a restart doesn't solve due to failure to detect power connectors to my graphics cards, but i can do nothing and once in a safe low graphics ubuntu environment (2.6.31.15, nvidia-glx-185-latest) i can restart again and be back to my normal high graphics power connector enabled 7800 GT SLI loveliness.

So, you know... wtf? I am willing to provide more information if this is tantalizing.

Everything gets worse over time, which doesn't make sense. This is a fresh install from two days ago and the first day was fine from a user point of view (haven't checked old logs) but now things are getting worse. It took 5 hours to crash catastrophically yesterday and today it takes as little as maybe 15 minutes to get to major system load, sound stuttering etc. Strangely i'm still on the same system without a reboot and everything is working fine.

What is the link between the graphics and the sound issue? Why does it screw with X? It may also be screwing up the resume image and manual resume from disk fails.

Also, is anyone having problems with IRQ/wireless RT-chipset/audio configuration in their syslogs? Just wondering if it is relevant.

OK, sorry, this is going to be a long one. So here is a follow-up, and a possible fix to my problem. Without having identified the culprit, I've tried a number of blind guesses, and this is only the last attempt. I want to stress the issue I reported is inevitably related to MY own hardware, hence what I'm going to report may not work for others. Also, I want to point out that mine are only speculations, and all my conclusions may as well be completely wrong: I'm an ordinary user, I'm no expert.

Two days ago I had another freeze. The System Log showed nothing, but a couple of those ratelimit.c: messages, a few minutes earlier. Actually, what I've found out is that these freezes leave no traces in the log, really. However, I originally connected the freezes to those ratelimit.c: messages only because the latter were the last things reported in the Log. In turn, this must have misled Daniel who filed the bug under Pulseaudio. But other times I noticed those messages were reported tens of minutes earlier, and in other occasions preceding messages related to something different entirely.

But last freeze actually told me something different... This was a good instance of those classic freezes I was experiencing involving both audio and graphics. I was raising Rhythmbox window to play some music (but no music was playing at that moment), when the system froze between the two notes of the system notification sound. But what was more revealing to me was that it froze while the window was half way, meaning that the default Compiz effect that comes with a normal installation of Karmic was tilting the window upwards (the effect is called "Glide 2"), but couldn't finish the operation. The freezes have got to be connected to graphics, especially Compiz. Being on a Thinkpad R50e, maybe my system is too old for such things, I don't know - though I can't recall any of this on Jaunty, where I even had the cube. In addition, going back to the previous day's freeze, it happened when I was transferring data to a USB pen drive, so I incorrectly connected it to its NTFS file system: instead, it occurred when the small window of the file transfer popped up. In this light, the two freezes together now made much more sense.

For all this, I disabled Compiz. I'm back on Metacity, and I've got a feeling things are changing. I enabled compositing_manager in the Configuration Editor for simple effects like shadows, transparency, smoother movements, etc. and do not miss Compiz eye candy effects at all. The system appears much more efficient, things are way quicker. With Compiz Fade effect out, for example, that overall sleepy feeling Ubuntu had acquired has finally gone. Now everything responds immediately, from menus to windows, rather than waiting to show the fading effect first. But what's more important, audio stopped stuttering, and system stopped freezing. Of course I'm aware it's still too early to say, and I myself had decided to report back only after a week, at least. But I thought that maybe other people could be suffering from the same problem, and try this too in the meantime. If this is the fix to the freezes I was experiencing on Karmic, I swear I'll ditc...

Hi Lelamal,

It's not clear from your message what got you to suspect Compiz, but just as a test, I disabled it: On the next reboot, I got the following in the log:

Nov 21 14:37:20 Philinux pulseaudio[1857]: ratelimit.c: 2 events suppressed
Nov 21 14:37:27 Philinux pulseaudio[1857]: ratelimit.c: 4 events suppressed
Nov 21 14:37:35 Philinux kernel: [ 93.228051] Clocksource tsc unstable (delta = -249963054 ns)
Nov 21 14:37:51 Philinux kernel: [ 109.760217] __ratelimit: 3 callbacks suppressed
Nov 21 14:37:51 Philinux kernel: [ 109.760230] deskbar-applet[2181]: segfault at 34 ip 00000034 sp bfa7432c error 4 in libc-2.10.1.so[110000+13e000]

I haven't got a freeze yet, but still having various apps segfaulting here and there makes me think that it will sure happen in the coming hours.
Here, the segfaults are often enough happening in the same second as a "__ratelimit: X callbacks suppressed", BTW.

Is there a way to prevent pulseaudio for loading at all in the machine (without disabling anything else than the sound subsystem)?
A manual killing of pulsaudio happens too late, since pulseaudio is already loaded then, and I see the suspicious "ratelimit.c" messages at several places during the boot.

Trying this could at least give an indication of whether pulseaudio is really the main suspect here:
If the problem still exists after pulseaudio has been fully removed, drivers and all, it will clearly indicate that the problem comes from something else...

Guys, the ratelimit messages are not errors.

On Sat, 2009-11-21 at 13:51 +0000, Filofel wrote:
Hi Lelamal,
>
> It's not clear from your message what got you to suspect Compiz, but
> just as a test, I disabled it

Hi Filofel, sorry for my convoluted explanation. What made me suspect was that the freezes described occurred when a Compiz effect was at work, either when raising a window ("Glide 2" effect), or popping up a window during file transfer (same effect, if I remember correctly), or a drop-down menu ("Fade" effect) - see my original report. Also, having excluded the ratelimit.c: messages as potential errors following Daniel's message #18, I'm still left clueless as to why the freezes I experience involve (most of the times) audio being stuck in an endless loop. But, let me stress it once again, I'm not sure this switch to Metacity will help, for I have no clue what's triggering the freezes in the first place.

Regarding the endless loop of audio: I think that could be fairly normal behaviour from sound hardware with samples in its buffer.

What you hear is what is in the sound hardwares buffers playing over and over again. I don't think there needs to be any pulse audio driver code involved.

I remember getting the same kind of effect years ago when Windows XP froze.

Triggered a reboot:
Nov 22 11:34:38 Philinux kernel: [ 106.085237] __ratelimit: 3 callbacks suppressed
Nov 22 11:34:38 Philinux kernel: [ 106.085250] python[1917]: segfault at 900c0c ip 0808e110 sp bf8c0530 error 4 in python2.6[8048000+1d3000]
Nov 22 11:38:10 Philinux kernel: imklog 4.2.0, log source = /var/run/rsyslog/kmsg started.

Triggered a reboot:
Nov 22 11:39:35 Philinux kernel: [ 113.942371] __ratelimit: 3 callbacks suppressed
Nov 22 11:39:35 Philinux kernel: [ 113.942385] deskbar-applet[2308]: segfault at c35d5f9e ip 00d1ade7 sp bfafe724 error 5 in libc-2.10.1.so[c14000+13e000]
Nov 22 11:40:03 Philinux kernel: [ 142.089533] BUG: Bad page map in process apport-checkrep pte:00c92025 pmd:32b91067
Nov 22 11:40:03 Philinux kernel: [ 142.089549] page:c104b240 flags:00000004 count:0 mapcount:-1 mapping:(null) index:0
Nov 22 11:40:03 Philinux kernel: [ 142.089559] addr:002e0000 vm_flags:08000075 anon_vma:(null) mapping:f6db833c index:df
Nov 22 11:40:03 Philinux kernel: [ 142.089574] vma->vm_ops->fault: filemap_fault+0x0/0x340
Nov 22 11:40:03 Philinux kernel: [ 142.089582] vma->vm_file->f_op->mmap: generic_file_mmap+0x0/0x50
Nov 22 11:40:03 Philinux kernel: [ 142.089591] Pid: 2988, comm: apport-checkrep Tainted: P 2.6.31-15-generic #50-Ubuntu
Nov 22 11:40:03 Philinux kernel: [ 142.089598] Call Trace:
Nov 22 11:40:03 Philinux kernel: [ 142.089613] [<c01c8a89>] print_bad_pte+0x139/0x1f0
Nov 22 11:40:03 Philinux kernel: [ 142.089622] [<c01ca5e7>] zap_pte_range+0x3d7/0x3e0
Nov 22 11:40:03 Philinux kernel: [ 142.089631] [<c01ca866>] unmap_vmas+0x186/0x350
Nov 22 11:40:03 Philinux kernel: [ 142.089640] [<c01cf780>] exit_mmap+0xa0/0x160
Nov 22 11:40:03 Philinux kernel: [ 142.089651] [<c014361e>] mmput+0x2e/0xa0
Nov 22 11:40:03 Philinux kernel: [ 142.089659] [<c0147bb2>] exit_mm+0xe2/0x120
Nov 22 11:40:03 Philinux kernel: [ 142.089667] [<c0147ddc>] do_exit+0xfc/0x2e0
Nov 22 11:40:03 Philinux kernel: [ 142.089676] [<c01527dd>] ? dequeue_signal+0x2d/0x170
Nov 22 11:40:03 Philinux kernel: [ 142.089685] [<c0161dab>] ? sched_clock_cpu+0xeb/0x1a0
Nov 22 11:40:03 Philinux kernel: [ 142.089693] [<c0147ffa>] do_group_exit+0x3a/0xb0
Nov 22 11:40:03 Philinux kernel: [ 142.089702] [<c015403f>] get_signal_to_deliver+0x18f/0x300
Nov 22 11:40:03 Philinux kernel: [ 142.089712] [<c03b27b8>] ? scsi_finish_command+0x98/0x100
Nov 22 11:40:03 Philinux kernel: [ 142.089721] [<c010304b>] do_signal+0x6b/0x160
Nov 22 11:40:03 Philinux kernel: [ 142.089730] [<c0152f88>] ? send_signal+0x38/0x70
Nov 22 11:40:03 Philinux kernel: [ 142.089738] [<c0154227>] ? force_sig_info+0x77/0xc0
Nov 22 11:40:03 Philinux kernel: [ 142.089747] [<c015427f>] ? force_sig+0xf/0x20
Nov 22 11:40:03 Philinux kernel: [ 142.089756] [<c0571479>] ? do_trap+0xb9/0xc0
Nov 22 11:40:03 Philinux kernel: [ 142.089764] [<c010318d>] do_notify_resume+0x4d/0x60
Nov 22 11:40:03 Philinux kernel: [ 142.089772] [<c0103448>] work_notifysig+0x13/0x1b
Nov 22 11:41:52 Philinux kernel: imklog 4.2.0, log source = /var/run/rsyslog/kmsg started.

Triggered a reboot:
Nov 22 11:42:28 Philinux pulseaudio[1842]: ratelimit.c: 1 events suppressed
Nov 22 11:42:34 Philinux pu...

PulseAudio doesn't have any sound drivers. See
https://lists.ubuntu.com/archives/ubuntu-devel-discuss/2009-October/009531.html
for instructions to disable PA.

Peter,

Thanks for the link. I'm going to try this and see is that makes a change.
I'm not just looking for a quick way to get rid of the problem on the one machine I have that exhibits the problem. Even if this makes the problem go away, I'm ready to come back to the problem configuration if I can help in pinpointing it.
Since I have a machine that does exhibit the problem, is there anything special I can do to help?

About "> PulseAudio doesn't have any sound drivers."
What if I replace "drivers" in my previous message by "kernel module"?

PulseAudio is entirely userspace. Are you referring to the ALSA kernel
modules (sound drivers)?

On Nov 22, 2009 10:36 AM, "Filofel" <email address hidden> wrote:

Peter,

Thanks for the link. I'm going to try this and see is that makes a change.
I'm not just looking for a quick way to get rid of the problem on the one
machine I have that exhibits the problem. Even if this makes the problem go
away, I'm ready to come back to the problem configuration if I can help in
pinpointing it.
Since I have a machine that does exhibit the problem, is there anything
special I can do to help?

About "> PulseAudio doesn't have any sound drivers."
What if I replace "drivers" in my previous message by "kernel module"?

-- Karmic: frequent freezes https://bugs.launchpad.net/bugs/479296 You
received this bug notificat...

Daniel,

I applied the method indicated at the link you sent.
Copied / pasted / executed:

touch ~/.pulse_a11y_nostart
echo autospawn = no|tee -a ~/.pulse/client.conf
killall pulseaudio

and rebooted the machine.
That didn't prevent pulseaudio from starting, though, as my log still contains:

Nov 22 16:44:53 Philinux kernel: [ 28.182386] __ratelimit: 3 callbacks suppressed
(...)
Nov 22 16:45:05 Philinux pulseaudio[1697]: ratelimit.c: 1 events suppressed
Nov 22 16:45:58 Philinux kernel: [ 93.244043] Clocksource tsc unstable (delta = -216722338 ns)

What gives?...

> PulseAudio is entirely userspace. Are you referring to
> the ALSA kernel modules (sound drivers)?

Well, the reason I'm asking is that I repeatedly see various processes segfaulting within a dozen microseconds after a
__ratelimit: X callbacks suppressed
messages, I inferred that some pulseaudio kernel module was involved: How could the pulseaudio usermode process interfere with a whole set of other user processes otherwise? Or send the whole OS to a freeze?
Example follows:

Nov 21 14:09:42 Philinux pulseaudio[1842]: ratelimit.c: 2 events suppressed
Nov 21 14:10:05 Philinux kernel: [ 94.732041] Clocksource tsc unstable (delta = -223109891 ns)
Nov 21 14:10:38 Philinux kernel: [ 127.890435] __ratelimit: 3 callbacks suppressed
Nov 21 14:10:38 Philinux kernel: [ 127.890447] apt-check[2885] general protection ip:805b17c sp:bf88202c error:0 in python2.6[8048000+1d3000]
Nov 21 14:11:35 Philinux kernel: [ 184.778508] update-manager[2935] general protection ip:809494a sp:bfae3134 error:0 in python2.6[8048000+1d3000]

Finally, my apologies for answering to you as "Peter" in my previous message. We have a Peter Chen in our team, and it looks like I blew a couple of synapses while typing the previous message!

Which process is launching pulseaudio? ps should identify it easily.

PulseAudio is 100% userspace. Provided you haven't done any fiddling
with PAM's limits.conf to enable RT privileges, the only culprits
would then be alsa-lib and alsa-kernel (linux).

> Nov 21 14:10:38 Philinux kernel: [ 127.890447] apt-check[2885] general protection ip:805b17c sp:bf88202c error:0 in python2.6[8048000+1d3000]
> Nov 21 14:11:35 Philinux kernel: [ 184.778508] update-manager[2935] general protection ip:809494a sp:bfae3134 error:0 in python2.6[8048000+1d3000]

See python2.6.

That's a good question.
I see the pulseaudio messages in the log (the one below is from the latest / current boot), but I see no pulseaudio process in ps when I'm using the machine.
So it seems pulseaudio was active in the machine at boot, 14 seconds after the very first boot-time log message was written, but isn't when I look at it.

The only mention of pulse I find in ps are in ssh-agent, gpg-agent and dbus-launch: All three have been launched with a cmd containing a pulse related option:
--exit-with-session /usr/bin/pulse-session gnome-session

On Sun, Nov 22, 2009 at 1:26 PM, Filofel <email address hidden> wrote:
> I see the pulseaudio messages in the log (the one below is from the latest / current boot), but I see no pulseaudio process in ps when I'm using the machine.
> So it seems pulseaudio was active in the machine at boot, 14 seconds after the very first boot-time log message was written, but isn't when I look at it.

Oh, that's gdm, then, which uses PA (as the gdm user) for login
"ready" alert. Once you've logged in as your user, you can probably
verify with "pgrep pulseaudio" that it is not running as your user.

> Once you've logged in as your user, you can probably
> verify with "pgrep pulseaudio" that it is not running as your user.

Sure, but what I wanted was to prevent pulseaudio from starting at all in the machine. Not even once.
I first created file /var/lib/pulseaudio/pulse_a11y_nostart: Better than the $HOME/.pulse_a11y_nostart because it prevents logging whatever $HOME resolves to. But that wasn't enough. It seems that pulseaudio was still called, presumably from the login screen (when it rolled the Ubuntu drum).
I ended up using the brute force approach and renaming the the pulseaudio binary to pulseaudio.off. That worked.

I'll be back here within at most 72 hours, to report either yet another freeze or random segfault (and pulseaudio presumably innocent) or no problem (and pulseaudio still under suspicion).

pulseaudio not guilty, at least not in my case.
I'm still experiencing random freezes...

At least one of the problems I had was a failing RAM stick.
I could isolate this 100% by swapping it with a perfectly identical machine running Windows (it immediately started to BSoD all around the place).

I have replaced the RAM and reenabled pulseaudio.

Is your audio provided by snd-hda-intel? If not then disregard this.

I think these issues could be caused by alsa kernel modules, as suggested earlier.

My current workaround is

i) Ensure correct audio chipset model is loaded:
sudo nano /etc/modprobe.d/alsa-base.conf
(confirm audio chipset with "cat /proc/asound/card0/codec#* | grep Codec" and start googling)
ii) If you changed "option snd-hda-intel model=XXX" to something more suitable then reboot

1) sudo killall pulseaudio (and anything else using audio)
2) sudo rmmod snd_hda_intel
3) sudo modprobe snd_hda_intel

Obviously, 3-5 imply that Daniel's analysis in #41 is spot-on, at least in my case.

Rock steady, no ratelimit.c messages using xbmc, no audio drops and my cpu usage graph doesn't look like the output of a 555.

HTH.

Hmmm, here I have an "NVidia nForce3 with AD1981B sound card" that takes snd-intel8x0m.
Aka ac97.
According to the alsa docs, this is the right driver for that one.
I still the pulseaudio messages in the log, but nothing segfault anymore, and the freezes are gone.
Not sure what I can improve there.

Anyhow, the machine works much better with 512 MiB of healthy RAM than with 1280 MiB including a bad stick!

In my view, the fact that this bug has been reported manually makes it more difficult to identify the culprits. More specifically, the issues described in my original report are quite general, and leave ample room both to interpretations, and to a series of other issues that may well be triggered by different culprits. In turn, this has become an issue report (not different from a forum exchange), rather than a bug report, for the bug itself is rather difficult to pinpoint, and some digging is necessary.

This was one thing I wanted to clarify before the following: I have been experiencing a series of other freezes that led me to report a number of other bugs. At first, it didn't occur to me to see similarities between them, but the comments of other users left me wondering whether these freezes are similar indeed, and whether the root cause of them all is a kernel problem.

Apport reported those bugs with my laptop model [IBM 1834RTG], but as I understand it this is not hardware-specific. If anyone is interested and is experiencing the same problems, please visit the following pages and comment accordingly: Bug #485108, Bug #488673.

 Many thanks in advance!

I also have same problems.
I have an Intel quad core 2 with Nvidia 9600 graphic and Realtec sound.
I use karmic with the proprietary Nvidia 185 driver.
It seems to me that this is a very strange mixup of Nvidia 185 driver, Compiz Window/Desktop Effects and Pulseaudio.
On my computer I could definitely see this phenomena:
When deactivating the Compiz effects (but still using the Nvidia 185 driver) two things happen:
- I get a lot of more desktop sound effects (things I never heard in jaunty and intrepid!)
- this pulseaudio: ratelimit.c: XXX events suppressed messages are lesser and the counts within are smaller.
When activating the Compiz Window/Desktop effects:
- Most desktop sounds are gone
- The pulseaudio: ratelimit.c: XXX events suppressed comes more often and with much higher counts within.
It seems there is any kind of problem between Compiz and Pulseaudio.

Try of compiz

*sorry
Try OFF Compiz

Hi,
I experience the same problems on my Thinkpad R50e. I don't know if I can add any new information. Not sure if this is important: I tried to run different window managers (KDE4 and XFCE) and it happened in both cases. I never have any music programs running when the crash happens. I'm not a pro, so if I can add any logs or something please tell me what to do...
Cheers,
Olf

I can reproduce the freezes on either of 2 PCs, both P4 era boxes with built in NIC/video/sound.
reproduce as follows :
Install 9.10 fresh + 159 updates + configure firefox + install thunderbird and configure that too.
Start GUI as normal, use 2 consoles (Ctrl-Alt F4 & F5) to watch the tail of xsession-errors and syslog.
Yes, syslog has many ratelimit messages and I do think they're irrelevant.

The last gasp coming out in xsession-errors at freeze time says :
** Message: NM disappeared

** (polkit-gnome-authentication-agent-1:1423): WARNING **: no sender

the NM message seems to be right at crash time.

What I see at freeze/crash time is that the mouse still moves, but mouse clicks and keyboard activity is ignored.
After the freeze - when I plug in a USB flash drive the hard drive rattles, something is doing an fsck (or similar) so the whole OS is not dead, my guess is that its only the GUI that is stuck.

I removed ndisrapper on my Thinkpad R50e and freezing stopped.
And messages pulseaudio fewer.

I can confirm what tomato-paste said about just the GUI being stuck. I have Karmic installed on my desktop, and at random times, everything but the mouse gets locked up. I can move the mouse around, and sometimes even carets in their textboxes are still blinking, but I cannot click or type anything. However, if sshd is running, I can log into the machine, and everything appears to be fine. When logged in like this, I can run "sudo shutdown -r now" to bring the system down gracefully.

Every once in a while, when brought down like this, when it reboots, it will hang on a black screen that says:

"fsck ...
n files, m directories..."

I have no clue where to start debugging this, and am on the verge of reverting back to Jaunty.

I have (had?) this problem as well. But for the last month or so I have not seen it.
I haven't done anything special really. The only thing I can thing of that could have affected this is that I once pushed a bit on my graphics card.
I seemed to be completely in its socket, and I could not feel it move. But after that I have had no more of these freezes...
(Cross my fingers)

I'm still getting the freeze at random times. Just when I think I've narrowed the problem down, it proves otherwise.

Also, it's almost as if the freeze comes in 'spurts' and then goes away for awhile.

Attached is a handful of files from my /var/log directory.

Hello,
The symptoms are I think associated with those in bug #495420: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/495420.
Sometimes massive segmentation faults happen (several processes), sometimes it's kernel panic. I had messages like "Nov 2 14:16:36 ECO pulseaudio[1564]: ratelimit.c: 158 events suppressed" recently just before "kernel bye bye" (freeze, two blinking LEDs).

Often messages like:
Dec 18 22:39:02 mini kernel: [90090.360114] avahi-daemon[729] general protection ip:cae330 sp:bf9f320c error:0 in libdaemon.so.0.4.0[cad000+5000]
(general protection ip) are occurs with messages about segfaults.

I have Karmic, proprietary nVidia graphic driver.
Last Ubuntu version before Karmic I had was Hardy Heron - it was all OK there.

I have the same bug, although I experience no difference whether music is being played or not. I suspect this has something to do with the NVIDIA driver and Karmic.

I use an HP DV5 Pavilion with NVIDIA 9600M GT and Ubuntu Karmic. Prior to Karmic I used 9.04 and everything was working fine. On Karmic, the screen would randomly flicker and sometimes freeze. This happens more often whenever I perform any activity which requires more work from the window manager - e.g. moving / opening windows. When the screen freezes, either the computer is completely non-responsive, or sometimes I can use CTRL-ALT-1 to escape it, then return to the normal display and it somehow causes a small temporary improvement. Sometimes, instead of freezing, the X server completely restarts and I get to the login screen.

That is the exact same behaviour as I have as well.

And I can add that there is no problem at all when running Windows, so this seems to be a software related issue.

Tal: Do you also get XID 13 in the log?

I have the same problem when I'm watching a movie with totem, ubuntu starts to go very slow.

Ubuntu 9.10 64bits

>> lspci
00:00.0 Host bridge: Advanced Micro Devices [AMD] RS780 Host Bridge
00:01.0 PCI bridge: Advanced Micro Devices [AMD] RS780 PCI to PCI bridge (int gfx)
00:0a.0 PCI bridge: Advanced Micro Devices [AMD] RS780 PCI to PCI bridge (PCIE port 5)
00:11.0 SATA controller: ATI Technologies Inc SB700/SB800 SATA Controller [AHCI mode]
00:12.0 USB Controller: ATI Technologies Inc SB700/SB800 USB OHCI0 Controller
00:12.1 USB Controller: ATI Technologies Inc SB700 USB OHCI1 Controller
00:12.2 USB Controller: ATI Technologies Inc SB700/SB800 USB EHCI Controller
00:13.0 USB Controller: ATI Technologies Inc SB700/SB800 USB OHCI0 Controller
00:13.1 USB Controller: ATI Technologies Inc SB700 USB OHCI1 Controller
00:13.2 USB Controller: ATI Technologies Inc SB700/SB800 USB EHCI Controller
00:14.0 SMBus: ATI Technologies Inc SBx00 SMBus Controller (rev 3a)
00:14.1 IDE interface: ATI Technologies Inc SB700/SB800 IDE Controller
00:14.2 Audio device: ATI Technologies Inc SBx00 Azalia (Intel HDA)
00:14.3 ISA bridge: ATI Technologies Inc SB700/SB800 LPC host controller
00:14.4 PCI bridge: ATI Technologies Inc SBx00 PCI to PCI Bridge
00:14.5 USB Controller: ATI Technologies Inc SB700/SB800 USB OHCI2 Controller
00:18.0 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] HyperTransport Technology Configuration
00:18.1 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] Address Map
00:18.2 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] DRAM Controller
00:18.3 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] Miscellaneous Control
01:05.0 VGA compatible controller: ATI Technologies Inc Radeon HD 3200 Graphics
01:05.1 Audio device: ATI Technologies Inc RS780 Azalia controller
02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B PCI Express Gigabit Ethernet controller (rev 02)
03:07.0 Multimedia controller: Philips Semiconductors SAA7131/SAA7133/SAA7135 Video Broadcast Decoder (rev d1)
03:0e.0 FireWire (IEEE 1394): Texas Instruments TSB43AB23 IEEE-1394a-2000 Controller (PHY/Link)

>> tail -100 /var/log/syslog
Mar 14 19:46:50 pcpedro pulseaudio[4100]: ratelimit.c: 9 events suppressed
Mar 14 19:46:55 pcpedro pulseaudio[4100]: ratelimit.c: 47 events suppressed
Mar 14 19:47:00 pcpedro pulseaudio[4100]: ratelimit.c: 200 events suppressed
Mar 14 19:47:05 pcpedro pulseaudio[4100]: ratelimit.c: 196 events suppressed
Mar 14 19:47:10 pcpedro pulseaudio[4100]: ratelimit.c: 112 events suppressed
Mar 14 19:47:15 pcpedro pulseaudio[4100]: ratelimit.c: 130 events suppressed
Mar 14 19:47:20 pcpedro pulseaudio[4100]: ratelimit.c: 1 events suppressed
Mar 14 19:47:25 pcpedro pulseaudio[4100]: ratelimit.c: 97 events suppressed
Mar 14 19:47:30 pcpedro pulseaudio[4100]: ratelimit.c: 84 events suppressed
Mar 14 19:47:35 pcpedro pulseaudio[4100]: ratelimit.c: 50 events suppressed
Mar 14 19:47:40 pcpedro pulseaudio[4100]: ratelimit.c: 87 events suppressed
Mar 14 19:47:45 pcpedro pulseaudio[4100]: ratelimit.c: 332 events suppressed
Mar 14 19:47:52 pcpedro pulseaudio[4100]: ratelimit.c: 62 events suppressed...

I forgot to say that if restarted pulseaudio go back to normal, but after a time the same thing happen again

I have the same problem - freezing Ubuntu 9.10.
I have the same thing in the log:
Mar 16 11:04:33 ubuntu pulseaudio[2293]: ratelimit.c: 157 events suppressed
Mar 16 11:06:46 ubuntu pulseaudio[2293]: ratelimit.c: 159 events suppressed
Mar 16 11:06:54 ubuntu pulseaudio[2293]: ratelimit.c: 158 events suppressed

I have disabled the ACPI in BIOS and no reboot for an hour. I will tell you later if it helped me or not.

No in the logs I have:
Mar 16 11:46:20 ubuntu pulseaudio[1722]: alsa-sink.c: Disabling timer-based scheduling because high-resolution timers are not available from the kernel.
Mar 16 11:46:20 ubuntu pulseaudio[1722]: alsa-source.c: Disabling timer-based scheduling because high-resolution timers are not available from the kernel.
Mar 16 11:46:26 ubuntu pulseaudio[1958]: alsa-sink.c: Disabling timer-based scheduling because high-resolution timers are not available from the kernel.
Mar 16 11:46:26 ubuntu pulseaudio[1958]: alsa-source.c: Disabling timer-based scheduling because high-resolution timers are not available from the kernel.

No reboot after disabling the ACPI.

Now I have:
Mar 18 17:08:18 ubuntu pulseaudio[2213]: alsa-sink.c: Disabling timer-based scheduling because high-resolution timers are not available from the kernel.
Mar 18 17:08:18 ubuntu pulseaudio[2213]: alsa-source.c: Disabling timer-based scheduling because high-resolution timers are not available from the kernel.
Mar 18 17:08:25 ubuntu pulseaudio[2213]: alsa-sink.c: ALSA woke us up to write new data to the device, but there was actually nothing to write!
Mar 18 17:08:25 ubuntu pulseaudio[2213]: alsa-sink.c: Most likely this is a bug in the ALSA driver 'snd_intel8x0'. Please report this issue to the ALSA developers.
Mar 18 17:08:25 ubuntu pulseaudio[2213]: alsa-sink.c: We were woken up with POLLOUT set -- however a subsequent snd_pcm_avail() returned 0 or another value < min_avail.

I have a similar problem with occasional crashes on Karmic (9.10). Sorry to say "me too" but I'm stumped on this one. Don't remember having this problem with Jaunty.

System hangs sometime occur a few seconds after selecting a new track in Rhythmbox (but can happen randomly at other times). Music carries on playing until end of track, mouse moves ok but everything else is stuck, including caps lock. Only way out is to use magic sys-request key to put keyboard into raw mode then 3-finger salute or sync-unmount-boot. However, graphics chipset status is evidently fubar because system must be physically power-cycled otherwise it will re-enter the same (hung) state once ubuntu starts X (gdm) again.

I also get pulseadio ratelimit messages but I think they might be a coincidence. Annoyingly, there are no other messages in the system log at the time of the crash.

Running Karmic on a toshiba laptop (Satellite Pro A10) with Intel sound and video (i810/i915) chipset.

lspci reports:
 00:02.0 VGA compatible controller: Intel Corporation 82852/855GM Integrated Graphics Device (rev 01)
 00:1f.5 Multimedia audio controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) AC'97 Audio Controller (rev 03)

Uname says:
 Linux blackrock 2.6.31-20-generic #58-Ubuntu SMP Fri Mar 12 05:23:09 UTC 2010 i686 GNU/Linux

I have a strong feeling these bugs:
https://bugs.launchpad.net/ubuntu/+source/libsdl1.2/+bug/203158
https://bugs.launchpad.net/ubuntu/+source/dosbox/+bug/429373
https://bugs.launchpad.net/ubuntu/+source/pulseaudio/+bug/368941
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/444950

As well as several other bugs I am not aware of, and several problems for which bug reports may or may not be filed yet (it's difficult to judge bug reports with similar symptoms)---are all related to this one.

Something is not right with SDL+pulseaudio+karmic's kernel.

According to bug 203158, the sdl issue is not present in Lucid.

What was changed that resolved this issue?

I know we only have a few weeks until the official release of Lucid, but I tend to keep my distro about six months behind to avoid the first wave of bugs. Therefore I'd like to fix this in Karmic before upgrading to Lucid.

This bug is present in Lucid as well, unfortunately. My symptoms are largely the same:

* Complete system freeze, including X and the mouse
* Kernel panic (cannot SSH into the box)
* No obvious pattern; will reboot multiple times in an hour, and then be fine for five or six hours or more
* Occasionally system hangs with music looping, but is more often random and not clearly associated with an activity
* Last reported issue is the ratelimit.d problem with Pulseaudio

I have filed what may be a duplicate bug at: <a href="https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/568549">https://bugs.edge.launchpad.net/ubuntu/+source/linux/+bug/568549</a>

It does not appear to be related to the proprietary NVidia drivers, as I've downgraded to the non-proprietary versions and am still experiencing the issue.

Audio device is as follows:

23:00.0 Audio device: Creative Labs X-Fi Titanium series [EMU20k2] (rev 03)
 Subsystem: Creative Labs Device 0044
 Flags: bus master, fast devsel, latency 0, IRQ 56
 Memory at dfcf0000 (64-bit, non-prefetchable) [size=64K]
 Memory at dfa00000 (64-bit, non-prefetchable) [size=2M]
 Memory at de000000 (64-bit, non-prefetchable) [size=16M]
 Capabilities: <access denied>
 Kernel driver in use: SB-XFi
 Kernel modules: snd-ctxfi

I have a similar issue. System freezes. Last lines in messages file are:

May 3 17:50:46 bar-prec-m4300 pulseaudio[2354]: ratelimit.c: 392 events suppressed
May 3 17:51:15 bar-prec-m4300 pulseaudio[2354]: ratelimit.c: 405 events suppressed
May 3 17:57:31 bar-prec-m4300 pulseaudio[2354]: ratelimit.c: 425 events suppressed
May 3 18:01:53 bar-prec-m4300 pulseaudio[2354]: ratelimit.c: 406 events suppressed
May 3 18:03:03 bar-prec-m4300 pulseaudio[2354]: ratelimit.c: 430 events suppressed
May 3 18:09:15 bar-prec-m4300 pulseaudio[2354]: ratelimit.c: 412 events suppressed
May 3 18:11:37 bar-prec-m4300 pulseaudio[2354]: ratelimit.c: 397 events suppressed
May 3 18:14:59 bar-prec-m4300 kernel: imklog 4.2.0, log source = /proc/kmsg started.
May 3 18:14:59 bar-prec-m4300 rsyslogd: [origin software="rsyslogd" swVersion="4.2.0" x-pid="807" x-info="http://www.rsyslog.com"] (re)start
May 3 18:14:59 bar-prec-m4300 rsyslogd: rsyslogd's groupid changed to 103
May 3 18:14:59 bar-prec-m4300 rsyslogd: rsyslogd's userid changed to 101
May 3 18:14:59 bar-prec-m4300 kernel: [ 0.000000] Initializing cgroup subsys cpuset
May 3 18:14:59 bar-prec-m4300 kernel: [ 0.000000] Initializing cgroup subsys cpu

I have a SigmaTel STAC9205 Sound Card. Worked fine with Ubuntu 9.x now I'm on 10.04, but with a clean install.
Dell Precision M4300.

Random, but a complete freeze. Unable to SSH and no mouse movement. I'm assuming this could be audio.

+ 1 on those random freezes -

I'll see if i can get a panic

Details:

Lucid 2.6.33-020633-generic
M4A89GTD PRO/USB3 motherboard
Phenom II X4 965
sda (ahci0): 750GB 7.2K SATA/300 Samsung SpinPoint F1 DT 32MB
sdb (ahci0): 750GB 7.2K SATA/300 Seagate 7200.10 16MB
sdc (ahci0): 80GB SSD SATA/300 2.5" Intel X25-M G2 MLC
Disk controller: ahci0: ATI SB700/SB800 SATA Controller [AHCI mode]
Network eth0 (r8168): Realtek RTL8111/8168B PCI-E Gigabit

running proprietary fglrx (based on catalyst 10.4 pre-release).

00:00.0 Host bridge: Advanced Micro Devices [AMD] RS780 Host Bridge Alternate
        Subsystem: Advanced Micro Devices [AMD] RS780 Host Bridge Alternate
        Flags: bus master, 66MHz, medium devsel, latency 0
        Capabilities: [c4] HyperTransport: Slave or Primary Interface
        Capabilities: [54] HyperTransport: UnitID Clumping
        Capabilities: [40] HyperTransport: Retry Mode
        Capabilities: [9c] HyperTransport: #1a
        Capabilities: [f8] HyperTransport: #1c

00:03.0 PCI bridge: Advanced Micro Devices [AMD] RS780 PCI to PCI bridge (ext gfx port 1)
        Flags: bus master, fast devsel, latency 0
        Bus: primary=00, secondary=01, subordinate=01, sec-latency=0
        I/O behind bridge: 0000c000-0000cfff
        Memory behind bridge: fe600000-fe6fffff
        Prefetchable memory behind bridge: 00000000d0000000-00000000dfffffff
        Capabilities: [50] Power Management version 3
        Capabilities: [58] Express Root Port (Slot+), MSI 00
        Capabilities: [a0] Message Signalled Interrupts: Mask- 64bit- Queue=0/0 Enable+
        Capabilities: [b0] Subsystem: Advanced Micro Devices [AMD] Device 9601
        Capabilities: [b8] HyperTransport: MSI Mapping Enable+ Fixed+
        Capabilities: [100] Vendor Specific Information <?>
        Capabilities: [110] Virtual Channel <?>
        Kernel driver in use: pcieport
        Kernel modules: shpchp

00:09.0 PCI bridge: Advanced Micro Devices [AMD] RS780 PCI to PCI bridge (PCIE port 4)
        Flags: bus master, fast devsel, latency 0
        Bus: primary=00, secondary=02, subordinate=02, sec-latency=0
        Memory behind bridge: fe700000-fe7fffff
        Capabilities: [50] Power Management version 3
        Capabilities: [58] Express Root Port (Slot+), MSI 00
        Capabilities: [a0] Message Signalled Interrupts: Mask- 64bit- Queue=0/0 Enable+
        Capabilities: [b0] Subsystem: Advanced Micro Devices [AMD] Device 9601
        Capabilities: [b8] HyperTransport: MSI Mapping Enable+ Fixed+
        Capabilities: [100] Vendor Specific Information <?>
        Capabilities: [110] Virtual Channel <?>
        Kernel driver in use: pcieport
        Kernel modules: shpchp

00:11.0 SATA controller: ATI Technologies Inc SB700/SB800 SATA Controller [AHCI mode] (rev 40) (prog-if 01)
        Subsystem: ATI Technologies Inc SB700/SB800 SATA Controller [AHCI mode]
        Flags: bus master, 66MHz, medium devsel, latency 64, IRQ 27
        I/O ports at b000 [size=8]
        I/O ports at a000 [size=4]
        I/O ports at 9000 [size=8]
        I/O ports at 8000 [size=4]
        I/O ports at 7000 [size=16]
        Memory at fe5ffc00 (32-bit, non-prefetchable) [size=1K]
      ...

I'm with Sog et al. Random freezes of the GUI, unable to switch to other consoles (via ctrl+alt+Fnx). I'm running Lucid but I believe these freezes only started happening the last 1-2 weeks. Perhaps a recent update? I did not have these freezes at all when the machine was running Karmic.

Last log entries before freeze:

May 29 22:37:35 user1-desktop pulseaudio[1447]: ratelimit.c: 483 events suppressed
May 29 22:41:23 user1-desktop pulseaudio[1447]: ratelimit.c: 126 events suppressed
May 29 22:43:02 user1-desktop pulseaudio[1447]: ratelimit.c: 22 events suppressed

System is Intel D945GCLF2 w/ Atom 330

lspci says:
00:1b.0 Audio device: Intel Corporation N10/ICH 7 Family High Definition Audio Controller (rev 01)

uname:
Linux user1-desktop 2.6.32-22-generic #33-Ubuntu SMP Wed Apr 28 13:27:30 UTC 2010 i686 GNU/Linux

$ pulseaudio -v
I: main.c: setrlimit(RLIMIT_NICE, (31, 31)) failed: Operation not permitted
I: main.c: setrlimit(RLIMIT_RTPRIO, (9, 9)) failed: Operation not permitted
I: core-util.c: Successfully gained nice level -11.
I: main.c: This is PulseAudio 0.9.21-63-gd3efa-dirty
I: main.c: Page size is 4096 bytes

Hi I have it on my Mythbuntu Lucid.

It already started some weeks ago when it still was on Karmic. The trigger for me was XBMC. I'm not sure it's because the mythbackend/frontend is on the same machine and because it's accessing the mysql from it. When XBMC is not started, and I only use mythfrontend it's less frequent or even not occuring.

I use spdif for the my Surround.
It's happening even when the xbmc is on it's main screen for some hours. Only hard reset works. No network (ssh/vnc) / keyboard or mouse works.

I see a IRQ message for my soundcard before the events happen.

May 30 11:57:40 MythMan kernel: [ 2699.932606] sd 8:0:0:0: [sdb] Write Protect is off
May 30 11:57:40 MythMan kernel: [ 2699.937433] sdb: sdb1
May 30 11:57:40 MythMan kernel: [ 2699.961431] sd 8:0:0:0: [sdb] Attached SCSI disk
May 30 11:57:45 MythMan kernel: [ 2704.685371] hda-intel: IRQ timing workaround is activated for card #0. Suggest a bigger bdl_pos_adj.
May 30 14:04:43 MythMan pulseaudio[2234]: ratelimit.c: 394 events suppressed
May 30 16:12:16 MythMan pulseaudio[2234]: ratelimit.c: 181 events suppressed
May 30 18:44:41 MythMan pulseaudio[2234]: ratelimit.c: 228 events suppressed
May 30 19:45:42 MythMan pulseaudio[2234]: ratelimit.c: 221 events suppressed
May 30 19:48:47 MythMan pulseaudio[2234]: ratelimit.c: 300 events suppressed
May 30 19:57:19 MythMan pulseaudio[2234]: ratelimit.c: 230 events suppressed
May 30 20:01:40 MythMan pulseaudio[2234]: ratelimit.c: 226 events suppressed
May 30 20:23:26 MythMan pulseaudio[2234]: ratelimit.c: 221 events suppressed
May 30 20:40:27 MythMan pulseaudio[2234]: ratelimit.c: 222 events suppressed
May 30 20:56:53 MythMan kernel: imklog 4.2.0, log source = /proc/kmsg started.

This is my card 0

00:10.1 Audio device [0403]: nVidia Corporation MCP51 High Definition Audio [10de:026c] (rev a2)
        Subsystem: Micro-Star International Co., Ltd. Device [1462:3290]
        Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0 (500ns min, 1250ns max)
        Interrupt: pin B routed to IRQ 22
        Region 0: Memory at fead8000 (32-bit, non-prefetchable) [size=16K]
        Capabilities: [44] Power Management version 2
                Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot+,D3cold+)
                Status: D0 PME-Enable- DSel=0 DScale=0 PME-
        Capabilities: [50] Message Signalled Interrupts: Mask+ 64bit+ Queue=0/0 Enable-
                Address: 0000000000000000 Data: 0000
                Masking: 00000000 Pending: 00000000
        Capabilities: [6c] HyperTransport: MSI Mapping Enable- Fixed+
        Kernel driver in use: HDA Intel
        Kernel modules: snd-hda-intel

mythman
    description: Desktop Computer
    product: MS-7329
    vendor: MSI
    version: 2.0
    serial: To Be Filled By O.E.M.
    width: 64 bits
    capabilities: smbios-2.4 dmi-2.4 vsyscall64 vsyscall32
    configuration: chassis=desktop uuid=00020003-0004-0005-0006-000700080009
  *-core
       description: Motherboard
       product: MS-7329
...

I've been getting the same thing. After being in sleep mode over the entire weekend I froze just a few minutes after waking the system up. The freeze is always directly preceded with the pulseaudio "events suppressed" lines.

Jun 1 07:56:32 casutton-desktop rsyslogd: [origin software="rsyslogd" swVersion="4.2.0" x-pid="893" x-info="http://www.rsyslog.com"] rsyslogd was HUPed, type 'lightweight'.
Jun 1 08:22:15 casutton-desktop pulseaudio[1696]: ratelimit.c: 1897 events suppressed
Jun 1 08:22:15 casutton-desktop kernel: [413057.692073] device eth0 entered promiscuous mode
Jun 1 08:22:24 casutton-desktop pulseaudio[1696]: ratelimit.c: 1092 events suppressed
Jun 1 08:22:45 casutton-desktop pulseaudio[1696]: ratelimit.c: 1409 events suppressed

i am using a Asus UL30a and had similar issues and system freezes .. with kernel version 2.6.35 i had no fruther freezes but high system\processor activity when watching videos or listening to audiofiles. i did come out of the blue after 10-20min that the audio and\or video playback stumbles. restarting the audio file doesnt make a differencen, even closing the application doesnt.

the pulseaudio events start to increase and wont stop. what i've done is, to kill the pulseaudio process restart the application and it works.

i've the impression, that this only happens when the battery level is below a certain range but i've to doublecheck this. It never happen to me when connected to a power source.

Back to report a couple additional observations...

- when the desktop & mouse freeze occurs, I *am* still able to ssh in to the system. So it appears that it's not a complete system hang

- while connected remotely, if I run 'shutdown -r now' on the afflicted machine, I notice that the Desktop goes through a graceful shutdown -- the apps start closing one-by-one; and most interestingly, the mouse becomes responsive again!

- I've a gut feeling that video playback activity may be what causes the freeze. It appears to occur most often (after I've been playing a video from within a browser (e.g. youtube); I don't recall it having happened after I've played a video from a locally-stored file (e.g. VLC playing a .wmv or .avi) and I've played a lot of music via Rhythmbox since the last time it freezed. And have gone through dozens of suspend / resume cycles.

The afflicted Lucid system has been through a number of updates since my last post. Here's where it's now at:

Linux user1-desktop 2.6.32-23-generic #37-Ubuntu SMP Fri Jun 11 07:54:58 UTC 2010 i686 GNU/Linux

I had this problem, but it has disappeared totally for me. I had the same symptoms: I could ssh into the frozen machine but video was frozen. Only mouse would work.

It disappeared when I changed my computer case.

It may be hard to believe but the reason (I think) was that in my previous case the graphics card (nVidia 260) was so cramped that it actually had physical contact with the HD-bay. I guess that caused some glitchy contact in some rare occations.

Just my 2 öre.

I've been able to prevent this issue from happening, by disabling Pulseaudio (as per the excellent guide at http://www.jeffsplace.net/node/12 )

I should probably clarify that this initially started happening to me with Karmic, but continued to occur after having upgraded to Lucid. My system (Intel D945GCLF2, Atom 330, running Lucid w/ 2.6.32-24-generic) has had no problems at all except when I'd recently been running some audio or video-with-audio; then I'd randomly fall into a situation desribed above where the screen and mouse would freeze / lock up (but I could still ssh in to it and restart it etc.) . But, again, now that Pulseaudio is out of the picture, I've not had a single lock up.

Also getting this issue in Ubuntu 10.10 Maveric on 2.6.35-23 kernel amd 64

pulseaudio[2728]: ratelimit.c: 16 events suppressed

Along with the desktop freeze about 2 secs after I start a music file, e.g. using aplay on a wav file.

Didn't have the problem prior to upgrade.
Have reinstalled alsa 1.0.23 and alsamixer works fine.

Update: the freeze has nothing to do with the message. It's due to C1E being enabled in the BIOS on an AMD Athlon II X2 based system. Disabling this stopped the freeze issue. Gory details here: https://bugzilla.kernel.org/show_bug.cgi?id=15289
Still getting the message but audio playback is fine now.

This bug was filed against a series that is no longer supported and so is being marked as Won't Fix. If this issue still exists in a supported series, please file a new bug.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.