System becomes extremely sluggish

[Taylor]

The interface experiences tremendous slowdown, where even basic tasks like opening nautilus sap resources.

The system never entirely freezes -- just moves extremely sluggishly. Using 'top', no single process saps it. Sometimes, top takes up 40%+, sometimes nautilus, whatever active process is happening. I haven't been able to figure out what causes it.

It did happen during natty, and I believe maverick before that.

ProblemType: Crash
DistroRelease: Ubuntu 11.10
Package: xserver-xorg-video-intel 2:2.15.901-1ubuntu2
ProcVersionSignature: Ubuntu 3.0.0-8.11-generic 3.0.1
Uname: Linux 3.0.0-8-generic i686
Architecture: i386
Chipset: i945gm
CompizPlugins: No value set for `/apps/compiz-1/general/screen0/options/active_plugins'
CompositorRunning: compiz
Date: Sun Aug 14 18:37:44 2011
DistUpgraded: Fresh install
DistroCodename: oneiric
DistroVariant: ubuntu
DuplicateSignature: [i945gm] GPU lockup EIR: 0x00000010 PGTBL_ER: 0x00000100 render.IPEHR: 0x01000000 Ubuntu 11.10
ExecutablePath: /usr/share/apport/apport-gpu-error-intel.py
ExtraDebuggingInterest: Yes, whatever it takes to get this fixed in Ubuntu
GpuHangFrequency: Several times a day
GpuHangReproducibility: Seems to happen randomly
GpuHangStarted: I don't know
GraphicsCard:
 Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller [8086:27a2] (rev 03) (prog-if 00 [VGA controller])
   Subsystem: Dell Device [1028:01d8]
   Subsystem: Dell Device [1028:01d8]
InstallationMedia: Ubuntu 11.10 "Oneiric Ocelot" - Alpha i386 (20110801)
InterpreterPath: /usr/bin/python2.7
MachineType: Dell Inc. MXC061
ProcCmdline: /usr/bin/python /usr/share/apport/apport-gpu-error-intel.py
ProcEnviron:

ProcKernelCmdLine: BOOT_IMAGE=/vmlinuz-3.0.0-8-generic root=UUID=a74c6c3a-ce67-44c9-b767-c4470aa4b79e ro quiet splash vt.handoff=7
RelatedPackageVersions:
 xserver-xorg 1:7.6+7ubuntu6
 libdrm2 2.4.26-1ubuntu1
 xserver-xorg-video-intel 2:2.15.901-1ubuntu2
SourcePackage: xserver-xorg-video-intel
Title: [i945gm] GPU lockup EIR: 0x00000010 PGTBL_ER: 0x00000100 render.IPEHR: 0x01000000
UpgradeStatus: No upgrade log present (probably fresh install)
UserGroups:

dmi.bios.date: 07/10/2007
dmi.bios.vendor: Dell Inc.
dmi.bios.version: A11
dmi.board.name: 0KG525
dmi.board.vendor: Dell Inc.
dmi.chassis.type: 8
dmi.chassis.vendor: Dell Inc.
dmi.modalias: dmi:bvnDellInc.:bvrA11:bd07/10/2007:svnDellInc.:pnMXC061:pvr:rvnDellInc.:rn0KG525:rvr:cvnDellInc.:ct8:cvr:
dmi.product.name: MXC061
dmi.sys.vendor: Dell Inc.
version.compiz: compiz 1:0.9.5.0-0ubuntu5
version.libdrm2: libdrm2 2.4.26-1ubuntu1
version.libgl1-mesa-dri: libgl1-mesa-dri 7.11-0ubuntu3
version.libgl1-mesa-dri-experimental: libgl1-mesa-dri-experimental N/A
version.libgl1-mesa-glx: libgl1-mesa-glx 7.11-0ubuntu3
version.xserver-xorg: xserver-xorg 1:7.6+7ubuntu6
version.xserver-xorg-input-evdev: xserver-xorg-input-evdev 1:2.6.0-1ubuntu13
version.xserver-xorg-video-ati: xserver-xorg-video-ati 1:6.14.99~git20110811.g93fc084-0ubuntu1
version.xserver-xorg-video-intel: xserver-xorg-video-intel 2:2.15.901-1ubuntu2
version.xserver-xorg-video-nouveau: xserver-xorg-video-nouveau 1:0.0.16+git20110411+8378443-1

[Bryce Harrington (bryce)]

Hi Taylor,

Your description is a bit terse, let me see if I understand correctly. Several times a day you experience a situation where it starts to tremendously slow down and then after a few minutes the whole system locks up and you have to power cycle the machine with the power button to get it back?

I see you don't know when it first started, but your system is a fresh installation of oneiric; did you have natty installed on it previously, and if so do you remember if it had freezes like this there too?

[Taylor (homeofpoe)]

Sorry for the brief description -- I was on my way out the door and didn't have time to do a thorough response. Wanted to get the bug up while I could.

You're correct, though the system never entirely freezes -- just moves extremely sluggishly. Using 'top', no single process saps it. Sometimes, top takes up 40%+, sometimes nautilus, whatever active process is happening. I haven't been able to figure out what causes it.

It did happen during natty, and I believe maverick before that. I'm happy to do whatever testing I can, I'm just not entirely sure where to start or how to single out the cause. I'll try to expand on this if I can when I get home, as well.

[Bryce Harrington (bryce)]

Ah, ok that makes more sense.

I think what might be going on here is two separate bugs. GPU lockup bugs generally involve the whole system locking up, however we have a known problem where the lockup handler sometimes has false-positives for some reason we haven't figured out. I think the crash popups you're seeing are these false positive type bugs.

The slowdown is probably a second unrelated problem. Might not even be X causing it. As we have plenty of bug reports about the false gpu lockups, I think this bug report should focus on the slowdown issue instead; sounds like that's what you care about.

[Taylor (homeofpoe)]

Works for me.

The problem I've had is how to try to isolate the issue. There doesn't seem to be a root cause - sometimes it slows down from the start, sometimes a few minutes after I've been logged in, sometimes hours later. It *seems* to happen most when I sit down to play a video, either via an application or flash... though that might just be me thinking that, as I haven't counted to confirm anything. I'll start making notes of when this occurs and try to jot down ongoing activities, system uptime; are there any other good-to-know factors?

[Bryce Harrington (bryce)]

Oh yeah there are lots of tips and tricks for optimizing performance. Some standard stuff to check:

* Any processes gobbling cpu, memory, or disk IO? (top, ps, nmon, atop, free -m, nice/ionice, etc.)
* Got any unusual cron jobs set up?
* Is it swapping to disk? (top or vmstat should show this, or the %wa number in top)
* Can you repro it in a guest session?
* Can you repro it on a Live CD?
* Any firmware updates for your motherboard?

dmesg is a good thing to check; especially keep an eye on that GPU lock message. If you're able to correlate that to when the sluggishness happens that would be quite interesting. If the message is there when the system is working fine, that would reinforce my guess that the bugs are separate.

Hardware's also a possibility. Switching to a SSD helped a ton with performance just in general, and particularly with swapping. Increasing ram can help in some situations. On laptops a failing power supply can cause weird bugs (esp. if it's fine when on battery).

Good luck, let me know what you find.

[Taylor (homeofpoe)]

* Any processes gobbling cpu, memory, or disk IO? (top, ps, nmon, atop, free -m, nice/ionice, etc.)
Whatever active applications are running take up whatever resources there are. If I'm running Pidgin, Pidgin will be taking 40%+. If it's nautilus, nautilus takes up 40%+. Even if I run 'top' to check with, it becomes the dominating process. Other than whatever I'm actively using (don't understand why that happens), there's no unusual or background activities taking up all the resources.

* Got any unusual cron jobs set up?
I haven't touched cron jobs at all since my install.

* Is it swapping to disk? (top or vmstat should show this, or the %wa number in top)
I checked last time this happened, it was at 0.0%.

* Can you repro it in a guest session?
* Can you repro it on a Live CD?
Haven't tried either of these yet - I'll try to give them a shot, but as the problem shows up seemingly at random, not sure when I'll be able to turn up results.

* Any firmware updates for your motherboard?
All my firmware should be fully up to date. I check for BIOS updates every few months just to be on top of things, but nothing since 2007, I think!

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[EricDHH (ericdhh)]

Simple applications cause Xorg to hose all memory and swap, the systems runs to death. I feel that's not okay, we should have proper ulimits to block a full system kill by single applications. Any ideas for ulimits are welcome.

[Bryce Harrington (bryce)]

Hey Taylor, just checking back for status; has this problem persisted through the release?

Given that you're seeing 40%+ loads in various applications, it's really starting to sound like a bug in something lower down than X, although I'm out of ideas on how to narrow it down further. But let me know if it's still occurring first, and we'll go from there.

@Eric, hard to be sure but your symptoms sound quite different than Taylors so you probably have some other bug.

[Taylor (homeofpoe)]

I actually haven't had this issue in the past month or so (maybe more), so I haven't had a chance to test circumstances. I wasn't able to narrow a cause or correlate any of the stuff you suggested to find something to look into.

If it shows up again I'll definitely bring it here, but I haven't recently. Which (hopefully!) is a good thing!

[Launchpad Janitor (janitor)]

[Expired for xserver-xorg-video-intel (Ubuntu) because there has been no activity for 60 days.]