[gm45] Xserver crash

[Robbie Williamson]

Binary package hint: xserver-xorg-video-intel

[ 78755.900]
Backtrace:
[ 78755.900] 0: /usr/bin/X (xorg_backtrace+0x28) [0x4a3488]
[ 78755.900] 1: /usr/bin/X (0x400000+0x64f1d) [0x464f1d]
[ 78755.900] 2: /lib/libpthread.so.0 (0x7ffab8e66000+0xfb50) [0x7ffab8e75b50]
[ 78755.900] 3: /usr/bin/X (miDoCopy+0x55) [0x54e405]
[ 78755.901] 4: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7ffab5fe3000+0x3ae20) [0x7ffab601de20]
[ 78755.901] 5: /usr/bin/X (0x400000+0xda818) [0x4da818]
[ 78755.901] 6: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7ffab5fe3000+0x36942) [0x7ffab6019942]
[ 78755.901] 7: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7ffab5fe3000+0x37669) [0x7ffab601a669]
[ 78755.901] 8: /lib/libdrm.so.2 (drmHandleEvent+0x103) [0x7ffab6443703]
[ 78755.901] 9: /usr/bin/X (WakeupHandler+0x4b) [0x435d7b]
[ 78755.901] 10: /usr/bin/X (WaitForSomething+0x1d7) [0x45f337]
[ 78755.901] 11: /usr/bin/X (0x400000+0x30802) [0x430802]
[ 78755.901] 12: /usr/bin/X (0x400000+0x2616a) [0x42616a]
[ 78755.901] 13: /lib/libc.so.6 (__libc_start_main+0xfd) [0x7ffab7dcfd8d]
[ 78755.901] 14: /usr/bin/X (0x400000+0x25d19) [0x425d19]
[ 78755.901] Segmentation fault at address (nil)
[ 78755.901]
Caught signal 11 (Segmentation fault). Server aborting
[ 78755.901]
Please consult the The X.Org Foundation support
         at http://wiki.x.org
 for help.

ProblemType: Bug
DistroRelease: Ubuntu 10.10
Package: xserver-xorg-video-intel 2:2.11.0-1ubuntu2
ProcVersionSignature: Ubuntu 2.6.35-6.9-generic 2.6.35-rc3
Uname: Linux 2.6.35-6-generic x86_64
Architecture: amd64
DRM.card0.DisplayPort.1:
 status: disconnected
 enabled: disabled
 dpms: Off
 modes:
 edid-base64:
DRM.card0.DisplayPort.2:
 status: disconnected
 enabled: disabled
 dpms: Off
 modes:
 edid-base64:
DRM.card0.HDMI_Type_A.1:
 status: disconnected
 enabled: disabled
 dpms: Off
 modes:
 edid-base64:
DRM.card0.LVDS.1:
 status: connected
 enabled: enabled
 dpms: On
 modes: 1440x900
 edid-base64: AP///////wAwrnRAAAAAACASAQOAHRJ46tCjlltXlCYjUlkAAAABAQEBAQEBAQEBAQEBAQEB2CegjFGEGjAwIDYAH7QQAAAYNCGgjFGEGjAwIDYAH7QQAAAYAAAADwCVCjKVCigeAQAwZABVAAAA/gBMVEQxMzNFUTFCCiAgAHM=
DRM.card0.VGA.1:
 status: disconnected
 enabled: disabled
 dpms: Off
 modes:
 edid-base64:
Date: Wed Jul 7 16:19:14 2010
DkmsStatus:
 virtualbox-ose, 3.2.4, 2.6.35-2-generic, x86_64: installed
 virtualbox-ose, 3.2.4, 2.6.35-4-generic, x86_64: installed
 virtualbox-ose, 3.2.4, 2.6.35-6-generic, x86_64: installed
 virtualbox-ose, 3.2.4, 2.6.35-5-generic, x86_64: installed
EcryptfsInUse: Yes
MachineType: LENOVO 2777CTO
ProcCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.35-6-generic root=UUID=185050f9-b1be-4215-889f-803f2bacff21 ro crashkernel=384M-2G:64M,2G-:128M quiet splash
ProcEnviron:
 PATH=(custom, user)
 LANG=en_US.utf8
 SHELL=/bin/bash
SourcePackage: xserver-xorg-video-intel
dmi.bios.date: 05/15/2009
dmi.bios.vendor: LENOVO
dmi.bios.version: 6EET41WW (3.01 )
dmi.board.name: 2777CTO
dmi.board.vendor: LENOVO
dmi.board.version: Not Available
dmi.chassis.asset.tag: No Asset Information
dmi.chassis.type: 10
dmi.chassis.vendor: LENOVO
dmi.chassis.version: Not Available
dmi.modalias: dmi:bvnLENOVO:bvr6EET41WW(3.01):bd05/15/2009:svnLENOVO:pn2777CTO:pvrThinkPadX301:rvnLENOVO:rn2777CTO:rvrNotAvailable:cvnLENOVO:ct10:cvrNotAvailable:
dmi.product.name: 2777CTO
dmi.product.version: ThinkPad X301
dmi.sys.vendor: LENOVO
system:
 distro: Ubuntu
 codename: maverick
 architecture: x86_64
 kernel: 2.6.35-6-generic

[Robbie Williamson (robbiew)]

This has occurred numerous times over the past 2 days after doing an apt-get dist-upgrade. Seems to only occur when the screensaver/power management kicks in, i.e. I step away from the machine while logged in and return to find X has crashed and the user login screen is up.

[Sebastien Bacher (seb128)]

Thanks Robbie, Chris could you try to figure what is going on there?

[Robbie Williamson (robbiew)]

Had another crash when I returned to my machine this morning, except this time the screen was black (but still active). I ssh'd into my laptop and pulled the Xorg file....looks like possibly a different crash, so let me know if I should open another bug.

[Robbie Williamson (robbiew)]

Relevant backtrace from above attached Xorg.0.log.old file (and actually it may be the same crash):

[137489.471] (II) PM Event received: Capability Changed
[137489.471] I830PMEvent: Capability change
[137527.328] [mi] EQ overflowing. The server is probably stuck in an infinite loop.
[137527.328]
Backtrace:
[137527.332] 0: /usr/bin/X (xorg_backtrace+0x28) [0x4a3488]
[137527.332] 1: /usr/bin/X (mieqEnqueue+0x1f4) [0x4a2d04]
[137527.332] 2: /usr/bin/X (xf86PostMotionEventP+0xc4) [0x47cb74]
[137527.332] 3: /usr/lib/xorg/modules/input/evdev_drv.so (0x7f63ab639000+0x53ef) [0x7f63ab63e3ef]
[137527.332] 4: /usr/bin/X (0x400000+0x6f787) [0x46f787]
[137527.332] 5: /usr/bin/X (0x400000+0x11e163) [0x51e163]
[137527.332] 6: /lib/libpthread.so.0 (0x7f63af84d000+0xfb50) [0x7f63af85cb50]
[137527.332] 7: /usr/lib/xorg/modules/libfb.so (fbValidateGC+0x1c0) [0x7f63abf9f9a0]
[137527.332] 8: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7f63ac9ca000+0x39706) [0x7f63aca03706]
[137527.332] 9: /usr/bin/X (0x400000+0xd756a) [0x4d756a]
[137527.332] 10: /usr/bin/X (ValidateGC+0x24) [0x443cb4]
[137527.332] 11: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7f63ac9ca000+0x368df) [0x7f63aca008df]
[137527.332] 12: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7f63ac9ca000+0x37669) [0x7f63aca01669]
[137527.332] 13: /lib/libdrm.so.2 (drmHandleEvent+0x103) [0x7f63ace2a703]
[137527.332] 14: /usr/bin/X (WakeupHandler+0x4b) [0x435d7b]
[137527.332] 15: /usr/bin/X (WaitForSomething+0x1d7) [0x45f337]
[137527.332] 16: /usr/bin/X (0x400000+0x30802) [0x430802]
[137527.332] 17: /usr/bin/X (0x400000+0x2616a) [0x42616a]
[137527.332] 18: /lib/libc.so.6 (__libc_start_main+0xfd) [0x7f63ae7b6d8d]
[137527.332] 19: /usr/bin/X (0x400000+0x25d19) [0x425d19]

[Chris Halse Rogers (raof)]

That's curious - EQ overflow errors often indicate that the GPU has locked up and X is the first to notice. Usually the intel kernel module will catch those, though. That probably had some interesting dmesg and intel_gpu_dump output.

Did apport not trigger a crash report for the initial crash? If not, I'll need to check the Xorg apport integration patch. It would be really nice to have the retracer grab a symbolic stacktrace there.

[Robbie Williamson (robbiew)]

no crash report from apport. I'll turn my screensaver/powermanagement stuff back on later today and get you some more debug info...assuming I can reproduce it.

[Robbie Williamson (robbiew)]

recreated crash with -dbg packages installed and drm debugging enabled. The hang is more random, and probably just a symptom of whatever is causing the original crash.

[Geir Ove Myhr (gomyhr)]

If the GPU locks up, the output of intel_error_decode is often more interesting than intel_gpu_dump (unless the kernel hasn't noticed yet). If X is able to restart properly, it indicates that the GPU is working which may be because its state has been reset. dmesg is most interesting with drm.debug=0x02 (which turns on the drm driver debugging. Not sure if that is what you used.

[Sebastien Bacher (seb128)]

Chris, is there any news of the issue there?

[Chris Halse Rogers (raof)]

I've accidentally run into a similar problem while debugging a -radeon crash. Checking out the equivalent code in -intel leads me to believe that this is fixed in xserver-xorg-video-intel 2.12, which we've got in Maverick. Have you been able to reproduce this crash recently?

[Mark Fraser (launchpad-mfraz)]

I'm not sure if what I am experiencing here is the same, but if after I first turn this computer on I open Konsole and do an upgrade X.org will crash and chuck me out of the desktop. If I try a second time, it works fine.

Xorg.0.log.old
[ 1253.412]
Backtrace:
[ 1253.413] 0: /usr/bin/X (xorg_backtrace+0x3b) [0x80e83bb]
[ 1253.413] 1: /usr/bin/X (0x8048000+0x5da8d) [0x80a5a8d]
[ 1253.413] 2: (vdso) (__kernel_rt_sigreturn+0x0) [0xda440c]
[ 1253.413] 3: /usr/bin/X (0x8048000+0x25d3e) [0x806dd3e]
[ 1253.413] 4: /usr/bin/X (0x8048000+0x1a5ba) [0x80625ba]
[ 1253.413] 5: /lib/libc.so.6 (__libc_start_main+0xe7) [0x573ce7]
[ 1253.413] 6: /usr/bin/X (0x8048000+0x1a191) [0x8062191]
[ 1253.413]
Caught signal 3 (Quit). Server aborting
[ 1253.413]
Please consult the The X.Org Foundation support
         at http://wiki.x.org
 for help.
[ 1253.413] Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[ 1253.413]

[Chris Halse Rogers (raof)]

On Sun, 2010-08-29 at 10:14 +0000, Mark Fraser wrote:
> I'm not sure if what I am experiencing here is the same, but if after I
> first turn this computer on I open Konsole and do an upgrade X.org will
> crash and chuck me out of the desktop. If I try a second time, it works
> fine.

No, you're seeing an unfortunate Plymouth/X interaction that I thought
we'd squashed in Lucid. You're not seeing a crash in intel_drv.

[Mark Fraser (launchpad-mfraz)]

I tried again last night and it seems that typing anything in to Konsole causes the crash and if I don't go into the Konsole until half an hour has passed since I booted it will still crash.

[Mark Fraser (launchpad-mfraz)]

The latest upgrade to x.org has got rid of the problems I was having.

[Chris Halse Rogers (raof)]

I'm going to mark this as Fix Released. I'm confident that the original crash was in the DRI2 SwapBuffers codepath fixed in 2.12, and no one has reported this as affecting xserver-xorg-video-intel 2.12.

Feel free to reopen if I'm wrong, and someone is still seeing this.

[chreko (chreko)]

Hi,

I am experiencing some strange artifacts with X, and I am wondering now if it is maybe related to this bug. I am using Ubuntu 10.10 AMD64 with xserver-xorg-video-intel 2.12. Here the problems:

1.) Some areas in the GUIs of my wine applications appear as black areas. This bug has been reported here (https://bugs.freedesktop.org/show_bug.cgi?id=30157) and here (http://bugs.winehq.org/show_bug.cgi?id=24430) and appear to be related to Xorg's Intel driver (2.12).

2.) The pull-down menus of the main panel or any window's menu fail to appear after a first click. It usually needs a second click to show up. I will need to confirm that this still happen when changing my video card.