SRU: [i915gm] GPU lockup (ESR: 0x00000001 IPEHR: 0x02000004)

David Coggins - I've forwarded this bug upstream to http://bugs.freedesktop.org/show_bug.cgi?id=34948 - please subscribe yourself to this bug, in case they need further information or wish you to test something. Thanks ahead of time!

Confirming this on the latest Natty daily build. I logged on and got a message about kernel problem, clicked "report problem", entered my password and got this crash.

Also, the system froze on the previous boot. The kernel was still responsive, I managed to get a proper shutdown, but X and even the keyboard NumPad toggling didn't work.

The system froze for me testing the latest natty 2.6.38-7.36 which should incorporate the fix for bug 717114

drm/i915: Fix tiling corruption from pipelined fencing

Mar 21 11:29:13 eee kernel: [ 0.000000] Linux version 2.6.38-7-generic (buildd@roseapple) (gcc version 4.5.2 (Ubuntu/Linaro 4.5.2-6ubuntu4) ) #36-Ubuntu SMP Fri Mar 18 22:05:25 UTC 2011 (Ubuntu 2.6.38-7.36-generic 2.6.38)

Mar 21 11:47:30 eee kernel: [ 1115.992048] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
Mar 21 11:47:30 eee kernel: [ 1115.998408] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -11 (awaiting 110179 at 110177, next 110180)

Apport is not generating a problem popup when I next reboot at the moment.
A small amount of testing with the terminal does not show any corruption which I was seeing 2 weeks ago bug 717114

Hi David, thanks for the feedback, sorry to hear it didn't solve it.

I've reopened the upstream bug report and included your comment. It would be very helpful if you could add yourself to the CC of the upstream bug report, so you can spot and respond to upstream's suggestions for things to test.

Still crashing. It happens on rapid mouse movement or scroll in Firefox.
2.6.38-7 worked fine for few days...

Reported on 2011-04-01 as Bug #747676, got the reply 'This bug has been marked a duplicate of bug 727594', so here we are

Again, this is not a critical bug for me.

I filed three reports in this bug which all happened when the crash reporter was reporting another bug. The crash reporter crashes and confronts me with this bug.

My computer doesn't stall or in any other way starts behaving badly, it's only when a diagnostic report on another application is being created that the bug reporter crashes.

System froze testing natty 2.6.38-8.41. I would be interested to know whether this kernel contains the Mar 25 patch

 drm/i915: Round-up GTT allocations for unfenced surfaces to the next tile row ?

Alternatively does the daily mainline kernel 2.6.39-999 contain the patch?

I gave up and went back to 10.04 LTS. I don't like Unity anyway.

Unsubscribe

On Fri, Apr 8, 2011 at 4:45 AM, David Coggins <email address hidden> wrote:
> System froze testing natty 2.6.38-8.41. I would be interested to know
> whether this kernel contains the Mar 25 patch
>
>  drm/i915: Round-up GTT allocations for unfenced surfaces to the next
> tile row ?
>
> Alternatively does the daily mainline kernel 2.6.39-999 contain the
> patch?
>
> --
> You received this bug notification because you are a direct subscriber
> of a duplicate bug (747676).
> https://bugs.launchpad.net/bugs/727594
>
> Title:
>  [i915gm] GPU lockup (ESR: 0x00000001 IPEHR: 0x02000004)
>
> Status in X.org xf86-video-intel:
>  Confirmed
> Status in “linux” package in Ubuntu:
>  New
> Status in “xserver-xorg-video-intel” package in Ubuntu:
>  Triaged
>
> Bug description:
>  Binary package hint: xserver-xorg-video-intel
>
>  As instructed by Bryce I installed the following kernel on my Asus eee pc 701 running natty:
>  " Please test with the following kernel:
>
>    http://kernel.ubuntu.com/~kernel-ppa/mainline/daily/current/ "
>
>  by doing sudo dpkg -i linux-image....deb
>
>  Then I ran update manager.
>
>  After rebooting I browsed with google chrome. The system froze after 5
>  minutes.
>
>  I rebooted the same kernel and immediately got this problem popup. In
>  addition there is font corruption - part of the letter t is missing in
>  the browser.
>
>  ProblemType: Crash
>  DistroRelease: Ubuntu 11.04
>  Package: xserver-xorg-video-intel 2:2.14.0-1ubuntu11
>  Uname: Linux 2.6.38-999-generic i686
>  Architecture: i386
>  Chipset: i915gm
>  CompizPlugins: No value set for `/apps/compiz-1/general/screen0/options/active_plugins'
>  CompositorRunning: None
>  DRM.card0.LVDS.1:
>   status: connected
>   enabled: enabled
>   dpms: On
>   modes: 800x480
>   edid-base64:
>  DRM.card0.VGA.1:
>   status: connected
>   enabled: enabled
>   dpms: On
>   modes: 1680x1050 1280x1024 1280x1024 1280x960 1152x864 1024x768 1024x768 1024x768 832x624 800x600 800x600 800x600 800x600 640x480 640x480 640x480 640x480 720x400
>   edid-base64: AP///////wBMLX4CMjJFTQ4RAQMOLx54KtUVpFVJmicUUFS/74CzAIGAgUBxTwEBAQEBAQEBfC6QoGAaHkAwIDYA2igRAAAaAAAA/QA4Sx5RDgAKICAgICAgAAAA/ABTeW5jTWFzdGVyCiAgAAAA/wBIOU5QNDAwMTk3CiAgAM8=
>  Date: Wed Mar  2 15:23:13 2011
>  DistUpgraded: Yes, recently upgraded Log time: 2011-02-17 19:00:58.983764
>  DistroCodename: natty
>  DistroVariant: ubuntu
>  DumpSignature: c38b9ae8 (ESR: 0x00000001 IPEHR: 0x02000004)
>  ExecutablePath: /usr/share/apport/apport-gpu-error-intel.py
>  GraphicsCard:
>   Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller [8086:2592] (rev 04) (prog-if 00 [VGA controller])
>     Subsystem: ASUSTeK Computer Inc. Device [1043:82d9]
>     Subsystem: ASUSTeK Computer Inc. Device [1043:82d9]
>  InstallationMedia: Ubuntu 11.04 "Natty Narwhal" - Alpha i386 (20110202)
>  InterpreterPath: /usr/bin/python2.7
>  MachineType: ASUSTeK Computer INC. 701
>  ProcCmdline: /usr/bin/python /usr/share/apport/apport-gpu-error-intel.py
>  ProcEnviron:
>
>  ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.38-999-generic root=UUID=3893e3fd-c2b4-40ec-9810-...

Is there anyone who can help to release a bug fixed package for Natty? My laptop crashed almost every day and every boot up. I have to remote login (ssh) to my laptop, kill X and restart my GUI every time.
Why should I use a Natty beta version? Because it seems only Natty's graphics chip-set driver support my Sandy Bridge GPU, ( VESA is too slow for me, even I don't need compiz and any 3D effect ),
Thank you.

I backported the patch from upstream that they believe will resolve this issue. It might be too late to include this in natty, however if people test it quickly and if it is found to eliminate the freezes on i915 maybe there's a chance.

My backport patch is available in my PPA here: https://launchpad.net/~bryce/+archive/fig

If you test it and find it resolves the issue sufficiently let me know.

Installed the patch from PPA
Still crashing -- scrolling on a page in Firefox with flash-video content freezes all.
The problem cannot be reported (not genuine Ubuntu package)

The PPA patch seems to have fixed the problem for me. Using Google Chrome would cause a lockup within five minutes before, so far no lockups after one hour of use. I'll update if anything changes.

Thanks!

Hi, unfortunately, the PPA patch did not fix the problem for me.
I am running Firefox, LibreOffice, gnome-terminal, gedit and a third part IM software, when I am typing in the gnome-terminal, the X locked, I have to force power off my laptop (by push the power button) because I don't have another computer to remote login this time.
Now I set "fbdev" as my GPU driver. It runs OK except the speed.
For my case, please refer to:
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/754777

@Robin, hmm, looks like your bug got mistakenly duped to this one. You have a sandybridge system, whereas near as we can tell this bug is specific to the i915/i945 architecture. I'll undupe your bug report.

That leaves us with one yes, one no... can anyone else provide feedback? I'd like to see a few yes's (and more yes's than no's) before proceeding with this fix.

I'm not seeing this on i945gm with the final freeze kernel.

Bryce, this time I booted on 2.6.38-8 with your patch from PPA. Seems to be OK, even after some 'stressing' in Firefox (multiple tabs, video, scrolling...).
Thanks!

My eeepc is running ok after installing the backport from the fig ppa. I used the system for several hours last night in google chrome and again this morning. Previously the bug could take all day to show up. Scrolling is a little laggy. However this evening I am seeing slight font corruption in the letter p in the console and gedit.

i always keep getting the fonts displayed weirdly..sometimes the upper part
of 't' is invisible, other times the lower part of letters like 'p' is
invisible. Sometimes strange weird lines appear.
wonder what wrong.

On Thu, Apr 21, 2011 at 12:10 PM, David Coggins
<email address hidden>wrote:

> My eeepc is running ok after installing the backport from the fig ppa. I
> used the system for several hours last night in google chrome and again
> this morning. Previously the bug could take all day to show up.
> Scrolling is a little laggy. However this evening I am seeing slight
> font corruption in the letter p in the console and gedit.
>
> --
> You received this bug notification because you are a direct subscriber
> of the bug.
> https://bugs.launchpad.net/bugs/727594
>
> Title:
> [i915gm] GPU lockup (ESR: 0x00000001 IPEHR: 0x02000004)
>
> Status in X.org xf86-video-intel:
> Fix Released
> Status in “linux” package in Ubuntu:
> New
> Status in “xserver-xorg-video-intel” package in Ubuntu:
> Incomplete
>
> Bug description:
> Binary package hint: xserver-xorg-video-intel
>
> As instructed by Bryce I installed the following kernel on my Asus eee pc
> 701 running natty:
> " Please test with the following kernel:
>
> http://kernel.ubuntu.com/~kernel-ppa/mainline/daily/current/ "
>
> by doing sudo dpkg -i linux-image....deb
>
> Then I ran update manager.
>
> After rebooting I browsed with google chrome. The system froze after 5
> minutes.
>
> I rebooted the same kernel and immediately got this problem popup. In
> addition there is font corruption - part of the letter t is missing in
> the browser.
>
> ProblemType: Crash
> DistroRelease: Ubuntu 11.04
> Package: xserver-xorg-video-intel 2:2.14.0-1ubuntu11
> Uname: Linux 2.6.38-999-generic i686
> Architecture: i386
> Chipset: i915gm
> CompizPlugins: No value set for
> `/apps/compiz-1/general/screen0/options/active_plugins'
> CompositorRunning: None
> DRM.card0.LVDS.1:
> status: connected
> enabled: enabled
> dpms: On
> modes: 800x480
> edid-base64:
> DRM.card0.VGA.1:
> status: connected
> enabled: enabled
> dpms: On
> modes: 1680x1050 1280x1024 1280x1024 1280x960 1152x864 1024x768 1024x768
> 1024x768 832x624 800x600 800x600 800x600 800x600 640x480 640x480 640x480
> 640x480 720x400
> edid-base64:
> AP///////wBMLX4CMjJFTQ4RAQMOLx54KtUVpFVJmicUUFS/74CzAIGAgUBxTwEBAQEBAQEBfC6QoGAaHkAwIDYA2igRAAAaAAAA/QA4Sx5RDgAKICAgICAgAAAA/ABTeW5jTWFzdGVyCiAgAAAA/wBIOU5QNDAwMTk3CiAgAM8=
> Date: Wed Mar 2 15:23:13 2011
> DistUpgraded: Yes, recently upgraded Log time: 2011-02-17 19:00:58.983764
> DistroCodename: natty
> DistroVariant: ubuntu
> DumpSignature: c38b9ae8 (ESR: 0x00000001 IPEHR: 0x02000004)
> ExecutablePath: /usr/share/apport/apport-gpu-error-intel.py
> GraphicsCard:
> Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller
> [8086:2592] (rev 04) (prog-if 00 [VGA controller])
> Subsystem: ASUSTeK Computer Inc. Device [1043:82d9]
> Subsystem: ASUSTeK Computer Inc. Device [1043:82d9]
> InstallationMedia: Ubuntu 11.04 "Natty Narwhal" - Alpha i386 (20110202)
> InterpreterPath: /usr/bin/python2.7
> MachineType: ASUSTeK Computer INC. 7...

Lets leave the font corruption to a different bug, unless it can be definitively shown to be caused by the fig ppa.

So, one more yes vote from mkis62, and perhaps too early to say for David.

I think the fig ppa has fixed the issue for me...

Alright I've gotten a few more confirmations of the fix on bugs #755693 and #763259. I think we may be too late for the release unfortunately, but I can try to get it accepted as an SRU. The criteria for SRUs are stricter than development changes though. The more +1 confirmations we can gather that this change resolves freeze issues the better.

Since this bug report is the oldest of this family of bugs I'm going to make it the master for all the dupes and the primary for filing the SRU. Apologies ahead of time if this generates a lot of email traffic for everyone (you can unsub via the bug report in launchpad if you don't want the emails, but hopefully it should be over and done in a couple weeks).

@David, since you're the original reporter on this bug report, would you mind following up with your latest findings vis a vis the ppa fix?

Fix confirmations on dupe bugs:

https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/755693/comments/4
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/755693/comments/5
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/763259/comments/3
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/763259/comments/6 # Solved on 4 eeepc's
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/763259/comments/7

I really hope this makes it in the release, I was not even able to login prior to installing the fix via the terminal in recovery mode :(

Bryce, the patch is empty...

The freezes have not occurred for several days now so I would consider them fixed. The patch has not made other problems worse.

+1
I tested the ppa. Everything looks good so far - no crash since I installed the patched driver.

> Bryce, the patch is empty...

Ha, true enough. Let's try that again.

Btw, I'm pretty sure the release is already more or less in the bag, so the fix is not likely to make it to the cd. (I could be wrong; it's a severe enough bug the archive team might pull it in if there are other last minute bugs making it worthwhile to regenerate the CD iso.)

It's more likely this will go through the SRU process. That means, it'll be reviewed and approved to go into natty-proposed for people to test, where it'll love for a period from a week to several weeks until enough testing has been done to show it does not cause regression. At that point it will move to natty-updates and be generally available to all users.

If a lot of people test -proposed and give it a +1, that will help accelerate getting the fix into the release. If people test it and find any regressions, that will significantly delay it going in until those issues can be investigated and resolved. So, hopefully lots of people give +1's and no one gives -1's, and we'll see this live for natty for people to update to post-release.

With some system not even able to boot normally that wouldn't be very user friendly.

It's looking more and more like Lucid and i8xx all over again :(

Accepted xserver-xorg-video-intel into natty-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

I installed the package from natty proposed replacing the fig package yesterday and since then there have been no freezes.

This issue should probably be mentioned in the release notes.

On Thu, Apr 28, 2011 at 03:32:06AM -0000, AJenbo wrote:
> This issue should probably be mentioned in the release notes.

It is; I added it there earlier today.

I am using the package xserver-xorg-video-intel from natty-proposed for several days, freezes are less frequent. It is better than before, unity is able (sometimes) to run for several hours. Still I have to reboot machine several times a day.

What package would that be?

On Wed, Apr 27, 2011 at 9:14 PM, David Coggins <email address hidden>wrote:

> I installed the package from natty proposed replacing the fig package
> yesterday and since then there have been no freezes.
>
> --
> You received this bug notification because you are a direct subscriber
> of a duplicate bug (769862).
> https://bugs.launchpad.net/bugs/727594
>
> Title:
> SRU: [i915gm] GPU lockup (ESR: 0x00000001 IPEHR: 0x02000004)
>
> Status in X.org xf86-video-intel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Invalid
> Status in “xserver-xorg-video-intel” package in Ubuntu:
> Fix Committed
> Status in “linux” source package in Natty:
> Invalid
> Status in “xserver-xorg-video-intel” source package in Natty:
> Fix Committed
>
> Bug description:
> [Impact]
> Severe GPU lockup affecting the i915/i945 family of Intel chips, resulting
> in unrecoverable freeze of graphics, black screen and/or corruption,
> requiring a hard reboot to reset. The issue is widespread amongst these
> cards as evidenced by the large number of dupes; the hardware is common.
> Most users report this is a regression in behavior from maverick.
>
> [Background]
> Subsequent to the code included in maverick, upstream introduced an
> optimization to relax fencing on Intel hardware. This change reduced the
> amount of memory allocated for video buffers. However on older (pre-G33)
> hardware such as i915/i945 this results in increased chances of GPU lockups.
>
> [Fix for Development Version]
> Upstream has opted to disable the relaxed fencing optimization for their
> driver release, and the change is still present in their active upstream git
> tree. Thus, we will be pulling this fix when we update X in oneiric.
>
> [Fix for Stable Version]
> For natty, the attached patch is a backport of the patch that went
> upstream. This patch makes relaxed fencing into an xorg.conf option that
> can be set, and makes it disabled by default for gen < 33 chipsets.
>
> [Steps to Reproduce]
> The freezes typically occur intermittently after some period of use. For
> some people it occurs right at boot, others after minutes or hours of usage.
> Some users find that certain activities such as web browsing makes the
> issue more likely to occur, but others do not find it correlated to any
> particular action.
>
> However, in all cases once the system is frozen, the file
> /sys/kernel/debug/dri/0/i915_error_state will contain an error code
> for the IPEHR value. The exact value appears to vary greatly from hw
> to hw, but common values tend to be either in the 0x02xxxxxx or
> 0x7xxxxxxx range.
>
> With this patch applied, gpu freezes should either go away entirely,
> or become much less frequent. Freezes which still occur but have
> IPEHR values outside these two ranges may be unrelated bugs.
>
> [Regression Potential]
> The patch itself is relatively small and unlikely to introduce regression.
>
> However, this switches optimization paths within the driver. It
> restores us to an older codepath so presumably this will at least be
> as stable as maverick, however the amount of testing this path has
> received on natty i...

My machine boots into a blank screen, except the times it gives me the option to boot into recovery mode at boot up. Then if I run failsafex graphic from the menu I get a decent GUI. Maybe the answer is to just have a way to always boot in failsafe mode.

On Sun, May 01, 2011 at 03:15:29AM -0000, mic wrote:
> I am using the package xserver-xorg-video-intel from natty-proposed for
> several days, freezes are less frequent. It is better than before, unity
> is able (sometimes) to run for several hours. Still I have to reboot
> machine several times a day.

You could have another bug at work (we've seen this before), or an
underlying condition that exacerbates the problem.

What this fix does is increase the size of the memory pool the gpu uses.
It's sort of like instead of requiring it to hit the bullseye on a
target we're just asking that it hit anywhere on the target or the hay
bales behind it. So whatever causes your GPU to be such a bad shot
still exists, but it's not getting kicked out of the competition so
much.

In general, the two things needed for GPU bug reports are the 'dmesg'
and /sys/kernel/debug/dri/0/i915_error_state. Look at the IPEHR value
especially, as that seems to be a roughly good indicator of dupes. If
it's still the same values as this bug (0x02xxxxxx or 0x7xxxxxxx) then
your system may just be sensitive. If it's a very different value, then
might be worth handling it as a second, unrelated bug and handling it
separately.

On Sun, May 01, 2011 at 01:00:49PM -0000, Jamison Lofthouse wrote:
> What package would that be?

xserver-xorg-video-intel from natty-proposed.

(It's essentially the same as the fig ppa, just different version number
and changelog entry.)

On Sun, May 01, 2011 at 06:04:06PM -0000, Patrick M wrote:
> My machine boots into a blank screen, except the times it gives me the
> option to boot into recovery mode at boot up.

Heh, you have to explain a lot more than that. Did the issue only occur
after installing the fix? Do you have GPU lockups matching this bug
prior to the black screen? Or do you see GPU error codes matching this
bug when it is black screened? What is your graphics card?

In general, I would suggest to all, unless you KNOW you have this exact
bug, or if you have found a regression which is traceable specifically
to this fix, let's handle your issue on separate bug reports, not here.
That should help spare people from lots of extraneous email...

> Then if I run failsafex
> graphic from the menu I get a decent GUI. Maybe the answer is to just
> have a way to always boot in failsafe mode.

It's not a bad idea. I've filed this as bug #775093, feel free to
subscribe if you'd like to follow it. I plan on spending some time
reworking X diagnostic tools during oneiric, and this would fit with
those plans.

This bug was fixed in the package xserver-xorg-video-intel - 2:2.14.0-4ubuntu7.1

---------------
xserver-xorg-video-intel (2:2.14.0-4ubuntu7.1) natty-proposed; urgency=low

  * Add 119_disable_relaxed_fencing.patch: The relaxed fencing
    optimization is suspected as the cause for various i915/945 gpu lockup
    issues. This disables the optimization by default but adds an
    xorg.conf parameter to let people experiment with it turned on.
    (LP: #727594, #761143, #761632, #755693)
 -- Bryce Harrington <email address hidden> Fri, 22 Apr 2011 19:12:55 -0700

Copied to oneiric as well.

Has this happened in the released version (released recently) not in the
beta2 version?

On Mon, May 2, 2011 at 8:53 AM, Martin Pitt <email address hidden> wrote:

> Copied to oneiric as well.
>
> ** Changed in: xserver-xorg-video-intel (Ubuntu Oneiric)
> Status: Fix Committed => Fix Released
>
> ** Changed in: xserver-xorg-video-intel (Ubuntu Oneiric)
> Milestone: natty-updates => None
>
> --
> You received this bug notification because you are a direct subscriber
> of a duplicate bug (769862).
> https://bugs.launchpad.net/bugs/727594
>
> Title:
> SRU: [i915gm] GPU lockup (ESR: 0x00000001 IPEHR: 0x02000004)
>
> Status in Release Notes for Ubuntu:
> New
> Status in X.org xf86-video-intel:
> Confirmed
> Status in “linux” package in Ubuntu:
> Invalid
> Status in “xserver-xorg-video-intel” package in Ubuntu:
> Fix Released
> Status in “linux” source package in Natty:
> Invalid
> Status in “xserver-xorg-video-intel” source package in Natty:
> Fix Released
> Status in “linux” source package in Oneiric:
> Invalid
> Status in “xserver-xorg-video-intel” source package in Oneiric:
> Fix Released
>
> Bug description:
> [Impact]
> Severe GPU lockup affecting the i915/i945 family of Intel chips, resulting
> in unrecoverable freeze of graphics, black screen and/or corruption,
> requiring a hard reboot to reset. The issue is widespread amongst these
> cards as evidenced by the large number of dupes; the hardware is common.
> Most users report this is a regression in behavior from maverick.
>
> [Background]
> Subsequent to the code included in maverick, upstream introduced an
> optimization to relax fencing on Intel hardware. This change reduced the
> amount of memory allocated for video buffers. However on older (pre-G33)
> hardware such as i915/i945 this results in increased chances of GPU lockups.
>
> [Fix for Development Version]
> Upstream has opted to disable the relaxed fencing optimization for their
> driver release, and the change is still present in their active upstream git
> tree. Thus, we will be pulling this fix when we update X in oneiric.
>
> [Fix for Stable Version]
> For natty, the attached patch is a backport of the patch that went
> upstream. This patch makes relaxed fencing into an xorg.conf option that
> can be set, and makes it disabled by default for gen < 33 chipsets.
>
> [Steps to Reproduce]
> The freezes typically occur intermittently after some period of use. For
> some people it occurs right at boot, others after minutes or hours of usage.
> Some users find that certain activities such as web browsing makes the
> issue more likely to occur, but others do not find it correlated to any
> particular action.
>
> However, in all cases once the system is frozen, the file
> /sys/kernel/debug/dri/0/i915_error_state will contain an error code
> for the IPEHR value. The exact value appears to vary greatly from hw
> to hw, but common values tend to be either in the 0x02xxxxxx or
> 0x7xxxxxxx range.
>
> With this patch applied, gpu freezes should either go away entirely,
> or become much less frequent. Freezes which still occur but have
> IPEHR values outside these two ranges may be unrelated...

Ithink this is related to my Minecraft bug on dh55hc mobo

I have the patched version and am still getting freezes in both unity-2d and gnome classic

@Richard, file a new bug report, this SRU is complete.

See:
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/727594/comments/40
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/727594/comments/42

Confirming that this is fixed on my G915. I had very often lockups with the natty beta's, since 1 week (when this new driver and a new kernel arrived), not even one.

Thanks to all working on this.

Hello, please see bug 768986. It is perhaps what Richard (comment #47) has. As he notes, I also have the patched version *7.1 and I still have freezes, but upon Kubuntu login. (I don't use GNOME so I can't say about that.)

I am dealing with an issue (Bug #882893 ) that may be related.

In general the PC (with Intel 82865G) is running OK, with Ubuntu 11.10 + Unity-2D, and latest updates.

Only problem: GPU (or CPU) crashes at 'screen OFF', on screensaver timeout.

( I will try: RevertingIntelDriverTo2.4 )

My coworker and I who are running the same model of machine are experiencing random lock ups. The holy hint is a message in the syslog and kern about the GPU hanging. I've included some data in the attached file.

It started recently for me within the last two weeks, definitely after some round of updates. Which ones I'm not certain of.

I have an old machine with this intel graphics card:
$ lspci -v|grep -i vga
00:02.0 VGA compatible controller: Intel Corporation 82865G Integrated Graphics Controller (rev 02) (prog-if 00 [VGA controller])
I am using Kubuntu natty since a while now, and I never had a problem whatsoever.
I only learned about this bug as I got the scary warning while trying to upgrade to oneiric.
Should I be worried? Is there something useful I can do?
I have arrived to natty via an upgrade, and I always kept all desktop effects deactivated, I guess that may be why I never had any issues.