[965gm] GPU lockup with screen corruption [lucid]

Please update to the current 2.6.35-6.7 kernel and see if you can reproduce the issue.

Bryce, the lucid tag you added seems wrong, is that some script needed an update?

I think Bryce adds those tags automatically, so he may not see this. However, it's partly correct:

system:
 distro: Ubuntu
 codename: maverick
 architecture: i686
 kernel: 2.6.32-21-generic

This is the Lucid kernel running on a Maverick system. I guess Bryce's script looks at the kernel to decide the Ubuntu release. Is this an upgrade that didn't go quite well?

Geir, that's correct, the script looks at the kernel version in the log to determine the distro version

We're closing this bug since it is has been some time with no response from the original reporter. However, if the issue still exists please feel free to reopen with the requested information. Also, if you could, please test against the latest development version of Ubuntu, since this confirms the bug is one we may be able to pass upstream for help.

This bug causing the display server the hang, the only way to recover is to reboot the system.
GPU: Intel GM45

These are the kernel message when problem occurred:

[drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
render error detected, EIR: 0x00000000
[drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 1398676 at 1398642)
[drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
[18701.420092] render error detected, EIR: 0x00000000
[18701.420113] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 1398678 at 1398642)
[18701.628454] iwlagn 0000:03:00.0: iwl_tx_agg_start on ra = 00:22:3f:8d:a1:42 tid = 0
[drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
render error detected, EIR: 0x00000000
[drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 1398681 at 1398642)
[drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
render error detected, EIR: 0x00000000

I'm seeing the same thing on Maverick () with GM45. Previously Lucid had been fine.

When it occurs the mouse cursor remains responsive, but everything else is locked up. Switching to a VT shows the following repeating:
[drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung

Xorg log shows:

[ 11405.592] (EE) intel(0): Failed to submit batch buffer, expect rendering corruption or even a frozen display: Input/output error.
[ 11405.916] (EE) intel(0): Failed to submit batch buffer, expect rendering corruption or even a frozen display: Input/output error.
[ 11406.888] (EE) intel(0): Failed to submit batch buffer, expect rendering corruption or even a frozen display: Input/output error.
[ 11408.209] (EE) intel(0): Failed to submit batch buffer, expect rendering corruption or even a frozen display: Input/output error.
[ 11408.905] (EE) intel(0): Failed to submit batch buffer, expect rendering corruption or even a frozen display: Input/output error.
[ 11409.334] (WW) intel(0): get vblank counter failed: Invalid argument
[ 11409.338] (WW) intel(0): get vblank counter failed: Invalid argument
[ 11409.338] (WW) intel(0): first get vblank counter failed: Invalid argument
[ 11409.642] (EE) intel(0): Failed to submit batch buffer, expect rendering corruption or even a frozen display: Input/output error.

These same 3 lines repeat indefinitely, in no particular order.

I do think, probably because I am a "victim" of this bug that the status should be critical. We are speaking of a complete freeze of the desktop with no other alternative than to reboot it...

Maybe it is related with this bug. When I try to see taskbar thumbnails, they appear empty (screenshot1).

And when I show all apps using "Showing Windows", they dissapear suddenly, showing their name only (screenshot 2)

I am seeing the identical symptoms reported above after upgrading from Lucid -> Maverick.

Output of lspci -vv is attached.

After further testing, I can reproduce this 100% of the time with the following setup:
1) Laptop with external display attached and setup with xrandr (So 2 monitors running)
2) Put laptop to sleep
3) Disconnect external monitor
4) Wake laptop

I am still experiencing this lockup at other times, but the above steps trigger it every time for me

It looks I'm encountering this issue on a new computer that I just built, though I haven't noticed it on another computer that I've had for about a year. Both computers are using the i915 driver, but the new computer has a slightly different video chipset. Let me know if there's any specific information that would be useful from either system.

Happening on an upgraded system from 9.04 -> 9.10 -> 10.04 -> 10.10

Running: 2.6.35-22-generic w/ i915 driver

Seems to happen on more graphics intensive operations such as "Desktop Grid" or "Present Windows"

I'm having the same problem, using a specific tool in paraview 3.8.1 software. I'm using 10.10 with 2.6.35-24-generic with the latest i915 driver.

The problem occurs when using the slice tool in paraview and dragging the slice plane. Error 100% reproducible also with older Ubuntu releases/drivers and with paraview 3.8.0

I encounter this problem 2 or 3 times a day. Especially when I use my Dell Latitude E6500 on the E-Port docking station which is connected to a 1920x1200 24" monitor (laptop screen disactivated).

The first time in a day if I try to boot with laptop closed using the dock power button (so only the 24" external screen is used), the system freeze at the end of the plymouth splashscreen. If I open the laptop, I have one chance on two.

Many time when I lock my screen then I try to get back : the screen stay blank. Sometime I can go to console, other time the system is completely frozen.

The problems mentioned by Joseph Maillardet also occur on my ASUS UL30A laptop which has Intel 4500MHD graphics.

I can confirm the same problem as noisymime #14. It's reproducable 25% of times when going to sleep with attached external monitor and waking up without external monitor connected.

drm:i915_hangcheck_elapsed
freezes computer, have to reeboot.

if I turn off compositing this problem is not there. (but then no desktop effects)

latitude e6410
Linux tuklah 2.6.35-26-generic #46-Ubuntu SMP Sun Jan 30 06:59:07 UTC 2011 x86_64 GNU/Linux

I seem to be having the same problem on a Lenovo X200 tablet. I get the error after using a projector (setting it up with xrandr) then putting the computer to sleep. It freezes when I try to wake it up, forcing me to reboot. The problem started after upgrading from 10.04 to 10.10.

>> uname -av

Linux xyzzy 2.6.35-25-generic #44-Ubuntu SMP Fri Jan 21 17:40:44 UTC 2011 x86_64 GNU/Linux

From kern.log:

Feb 7 10:17:55 xyzzy kernel: [32788.390039] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
Feb 7 10:17:55 xyzzy kernel: [32788.390055] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiti
ng 70 at 53)
Feb 7 10:17:56 xyzzy kernel: [32789.170035] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
Feb 7 10:17:56 xyzzy kernel: [32789.170585] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiti
ng 74 at 53)

not much to add - my situation is basically identical to david's post #22, except in this case it's a lenovo t510.

I had this same problem with my CCE notebook. COREI5 with onboard video. Using ubuntu 10.10 32bits here.

Same issue but somewhat random..
It works fine with vesa driver.

HP touchsmart tm2100e

kernel:
Linux Mobiligor5 2.6.35-27-generic-pae #48-Ubuntu SMP Tue Feb 22 21:46:58 UTC 2011 i686 GNU/Linux

VGA controller:
00:02.0 VGA compatible controller: Intel Corporation Core Processor Integrated Graphics Controller (rev 02) (prog-if 00 [VGA controller])
        Subsystem: Hewlett-Packard Company Device 1486
        Flags: bus master, fast devsel, latency 0, IRQ 45
        Memory at c0000000 (64-bit, non-prefetchable) [size=4M]
        Memory at b0000000 (64-bit, prefetchable) [size=256M]
        I/O ports at 4050 [size=8]
        Expansion ROM at <unassigned> [disabled]
        Capabilities: [90] MSI: Enable+ Count=1/1 Maskable- 64bit-
        Capabilities: [d0] Power Management version 2
        Capabilities: [a4] PCI Advanced Features
        Kernel driver in use: i915
        Kernel modules: i915

I recently switched from Ubuntu to ArchLinux (which provides a newer 2.14.0-2 version of the X.org Intel i810/i830/i915/945G/G965+ video drivers) and the problems in this bug report no longer occur for me. So I think Ubuntu should just provide a newer version of the xserver-xorg-video-intel package to solve this bug. Cheers.

I have been bite by this bug too. I think it's closely related to this one

https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/560433

I reproduce this bug with exactly the same scenario than noisymime (see above).

It happens to me too (Dell Optiplex 760 desktop). The problem only occurs when I enable desktop effects.

Here's the video-related output of lspci -v:

00:02.0 VGA compatible controller: Intel Corporation 4 Series Chipset Integrated Graphics Controller (rev 03) (prog-if 00 [VGA controller])
        Subsystem: Dell Device 027f
        Flags: bus master, fast devsel, latency 0, IRQ 51
        Memory at fe800000 (64-bit, non-prefetchable) [size=4M]
        Memory at d0000000 (64-bit, prefetchable) [size=256M]
        I/O ports at ec90 [size=8]
        Expansion ROM at <unassigned> [disabled]
        Capabilities: [90] MSI: Enable+ Count=1/1 Maskable- 64bit-
        Capabilities: [d0] Power Management version 2
        Kernel driver in use: i915
        Kernel modules: i915

00:02.1 Display controller: Intel Corporation 4 Series Chipset Integrated Graphics Controller (rev 03)
        Subsystem: Dell Device 027f
        Flags: bus master, fast devsel, latency 0
        Memory at fe700000 (64-bit, non-prefetchable) [size=1M]
        Capabilities: [d0] Power Management version 2

$ uname -a
Linux sybren.chess.nl 2.6.35-28-generic #50-Ubuntu SMP Fri Mar 18 18:42:20 UTC 2011 x86_64 GNU/Linux

$ apt-cache policy xserver-xorg-video-intel
xserver-xorg-video-intel:
  Installed: 2:2.12.0-1ubuntu5.2
  Candidate: 2:2.12.0-1ubuntu5.2
  Version table:
 *** 2:2.12.0-1ubuntu5.2 0
        500 http://nl.archive.ubuntu.com/ubuntu/ maverick-updates/main amd64 Packages
        100 /var/lib/dpkg/status
     2:2.12.0-1ubuntu5 0
        500 http://nl.archive.ubuntu.com/ubuntu/ maverick/main amd64 Packages

Might be sth similar:
https://bugs.launchpad.net/bugs/775936
I have Intel GMA 4500MHD.

Same thing here.

This problem is affecting me since i started using a Dual Head schema (Laptop Monitor + LCD monitor).

I'm having this problem since Ubuntu 9.10 and it's really annoying!!!! I use dual monitors everyday at work, and when i leave work i put my laptop in sleep mode hoping that when i get in class or home, all my documents, programs and setup is going to be running as it should.

Sometimes i get lucky and Ubuntu wakes just fine, but most times it crashes and i end up loosing all my setup and need to open all my programs again.

This is bad because the system becomes untrusty... I keep saving my things like crazy hoping that in one of these crashes i do not miss anything unsaved...

My setup is a Dell Vostro 1400 (1440x900) + LG LCD 17' Monitor (1280x800).

I have tried to solve this problem by switching to a Dell Vostro 1520 (1280x800) but it did'nt seem to work any better :(

Every new version of Ubuntu i try to quikly update to check if this problem is fixed but it's not... I noticed that is not fixed even on 11.04... So i wont update, for now. I'm going to wait for 11.10.

I'm about to give up my Dual Head setup...

Another idea is to switch distro as Suraj have suggested that a newer version of Xorg apparently fixes this. I think i'm going back to Gentoo. It's complicated, and uses a lot of machine power to compile things, but at least i can use newer versions of base software not depending on Ubuntu.

Sorry for my bad english, i'm in a hurry.

Victor.

This bug is still present on a Ubuntu Lucid 10.04.2 installation.

This error is triggered by the screensaver, but only seldom.

    [11480.256008] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
    [11480.256016] render error detected, EIR: 0x00000000
    [11480.256031] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 259791 at 259790)

When this error is detected, the current user get logged out and the system goes back to GDM.

Bitten by this problem on Natty on a thinkpad T41 with an intel. Can't reliably reproduce but freezes at least once a week.

lspci -v

00:00.0 Host bridge: Intel Corporation Mobile PM965/GM965/GL960 Memory Controller Hub (rev 0c)
 Subsystem: Lenovo T61
 Flags: bus master, fast devsel, latency 0
 Capabilities: [e0] Vendor Specific Information: Len=0a <?>
 Kernel driver in use: agpgart-intel

xserver-xorg-video-intel 2:2.14.0-4ubuntu7.1

dmesg
[42280.508107] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
[42280.509520] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -11 (awaiting 1394761 at 1394760, next 1394776)

I'll try x-org edgers PPA because I can't accept hard freezes on my desktop, it's linux after all! ;)

Similar problem on a Intel DH67CF motherboard with a i5-2500K CPU.

Jul 31 21:53:53 Spencer kernel: [21587.865167] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
Jul 31 21:53:53 Spencer kernel: [21587.873118] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -11 (awaiting 6150422 at 6150418, next 6150423)
Jul 31 21:53:53 Spencer kernel: [21587.904117] show_signal_msg: 27 callbacks suppressed
Jul 31 21:53:53 Spencer kernel: [21587.904121] compiz[2972]: segfault at 0 ip 00007f82184febe8 sp 00007fff7648f780 error 6 in i965_dri.so[7f821848d000+ac000]
Jul 31 21:53:59 Spencer kernel: [21594.310599] compiz[4853]: segfault at 0 ip 00007f4bd12e0acc sp 00007fff63177070 error 6 in i965_dri.so[7f4bd12c1000+ac000]
Jul 31 21:54:02 Spencer kernel: [21596.825503] compiz[4860]: segfault at 0 ip 00007f7bdab29acc sp 00007fff160273d0 error 6 in i965_dri.so[7f7bdab0a000+ac000]
Jul 31 21:54:04 Spencer kernel: [21599.343086] compiz[4865]: segfault at 0 ip 00007fa1c91fcacc sp 00007fff13a6e8b0 error 6 in i965_dri.so[7fa1c91dd000+ac000]

Same here on a Dell Inspiron 710m with driver i915 and Linux kernel 2.6.32-33-generic.
Sometimes, and for a reason that I haven't identified yet, if I open Thunderbird or Firefox I get a black screen then go back to the log in screen; I have to reboot to get rid of that malfunctioning. It happens at least once a week and started within the last 7 days (I would say).

kern.log:
[ 406.277038] ipw2200: Firmware error detected. Restarting.
[ 504.616250] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
[ 504.616269] render error detected, EIR: 0x00000000
[ 504.616304] [drm:i915_do_wait_request] *ERROR* i915_do_wait_request returns -5 (awaiting 15459 at 15442)
[ 746.270794] ipw2200: Firmware error detected. Restarting.

same here since i started using an external monitor on my laptop, happens on waking up from suspend, sometimes when starting vlc.

the error message i get:

Failed to submit batch buffer, expect rendering corruption or even a frozen display: Input/output error.

Intel GM45 Chipset
3.0.6-030006-generic #201110050043 SMP Wed Oct 5 02:07:23 UTC 2011 i686 GNU/Linux
ubuntu 10.10

Re-closing as there's not been further information from the original reporter.

The error message is a generic one that is written whenever the GPU is locked up. Just because you see this error message does not mean you have the same root bug as someone else.

In general, upgrading your kernel is the proper way to solve the bug. If you want to see a fix in lucid, you'd need to do a bisection search to find the relevant patch to be backported, but I'll leave that to the kernel team to explain. :-)

Fwiw, in Precise the vast bulk of these kinds of freezes are gone (there's still a few but they are much less common and more esoteric than in older kernels.)