Kernel Oops - BUG: unable to handle kernel paging request at 850f08bc; EIP is at kmem_cache_alloc+0x50/0x100

[gene]

after last update of the 2.6.38-13-generic #53-Ubuntu SMP Mon Nov 28 19:23:39 UTC 2011 i686 i686 i386 GNU/Linux
Radeon would crash the system to require restart.
The kernel's oops and ktrace will be attached below:

ProblemType: Bug
DistroRelease: Ubuntu 11.04
Package: linux-image-2.6.38-13-generic 2.6.38-13.53
ProcVersionSignature: Ubuntu 2.6.38-13.53-generic 2.6.38.8
Uname: Linux 2.6.38-13-generic i686
AlsaVersion: Advanced Linux Sound Architecture Driver Version 1.0.23.
AplayDevices:
 **** List of PLAYBACK Hardware Devices ****
 card 0: SB [HDA ATI SB], device 0: ALC861-VD Analog [ALC861-VD Analog]
   Subdevices: 1/1
   Subdevice #0: subdevice #0
Architecture: i386
ArecordDevices:
 **** List of CAPTURE Hardware Devices ****
 card 0: SB [HDA ATI SB], device 0: ALC861-VD Analog [ALC861-VD Analog]
   Subdevices: 1/1
   Subdevice #0: subdevice #0
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: 1673 F.... pulseaudio
Card0.Amixer.info:
 Card hw:0 'SB'/'HDA ATI SB at 0xc0400000 irq 40'
   Mixer name : 'Realtek ALC861-VD'
   Components : 'HDA:10ec0862,1179820d,00100001 HDA:11c11040,11790001,00100200'
   Controls : 19
   Simple ctrls : 12
Date: Tue Dec 27 12:16:20 2011
HibernationDevice: RESUME=UUID=5d87ad1c-529d-43e5-a5b9-d8bcc75895ce
Lsusb:
 Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
 Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
 Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
MachineType: TOSHIBA Satellite L35
PccardctlIdent:
 Socket 0:
   no product info available
PccardctlStatus:
 Socket 0:
   no card
ProcEnviron:
 LANGUAGE=en_US:en
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-2.6.38-13-generic root=UUID=8ed12067-caa1-4af3-b3b4-c97a2fe5cc3d ro radeon.modeset=1 acpi_sleep=nonvs resume=/dev/sda5 crashkernel=384M-2G:64M,2G-:128M quiet profile kexec_jump_back_entry=0xec101058
RelatedPackageVersions:
 linux-restricted-modules-2.6.38-13-generic N/A
 linux-backports-modules-2.6.38-13-generic N/A
 linux-firmware 1.52
RfKill:
 0: phy0: Wireless LAN
  Soft blocked: no
  Hard blocked: no
SourcePackage: linux
UpgradeStatus: Upgraded to natty on 2011-05-06 (235 days ago)
WpaSupplicantLog: (does not work anyways bug 788274, , why bother?)

dmi.bios.date: 10/24/2007
dmi.bios.vendor: TOSHIBA
dmi.bios.version: V3.10
dmi.board.name: Satellite L35
dmi.board.vendor: TOSHIBA
dmi.board.version: Not Applicable
dmi.chassis.type: 10
dmi.chassis.vendor: TOSHIBA
dmi.chassis.version: N/A
dmi.modalias: dmi:bvnTOSHIBA:bvrV3.10:bd10/24/2007:svnTOSHIBA:pnSatelliteL35:pvrPSL33U-03W02K:rvnTOSHIBA:rnSatelliteL35:rvrNotApplicable:cvnTOSHIBA:ct10:cvrN/A:
dmi.product.name: Satellite L35
dmi.product.version: PSL33U-03W02K
dmi.sys.vendor: TOSHIBA

[gene (eugenios)]

I got a kernel freeze tonight. No reqsys keys were functioning, power button restart.
I am going to try out other than the 38 version of the kernel and see if the issue is there.
--

#Please do not send me Microsoft Office/Apple iWork documents. Send OpenDocument instead! http://fsf.org/campaigns/opendocument

# this message is sent from Mutt and is composed in the GNU Emacs

[gene (eugenios)]

The 3.0 version of the kernel seems to be free of the bug:
uname -a
Linux 3.0.14-030014-generic #201112211655 SMP Wed Dec 21 22:03:55 UTC 2011 i686 i686 i386 GNU/Linux
uptime
 up 13 days, 23:19

[gene (eugenios)]

Had a radeon/kernel freeze on the new updated 3.0.17 kernel. I was able to ssh to the machine though.
I attach the call trace of the accident below.

[gene (eugenios)]

I built 3.0.17 kernel from git source with a custom config. It hung on me in a few hours without any traces in the logs. So I wonder if the buggy patches applied between 3.0.14 and 3.0.17 were those for radeon. I will try to figure this out by building it again with more generic config. Not sure how to do it with git yet, though. It would be much easier just to be able to apply only the security patches.

Kernel freezes on this machine now become usual.
--

[gene (eugenios)]

As it turned out there have been only two patches (directly) related to radeon between the stable for me 3.0.14 and buggy 3.0.17 versions.
I reverted the corresponding commits:
-------------------------
commit 77a10a203dc228dc9e1d40075b3c61c3e0882db2
Author: gene <gene@jenshen.(none)>
Date: Wed Jan 18 11:44:15 2012 -0600

    Revert "drm/radeon/kms: bail on BTC parts if MC ucode is missing"

    This reverts commit 77e00f2ea94abee1ad13bdfde19cf7aa25992b0e.

commit 06ca5dda377dae29ce50032b6a47849fd7354ec1
Author: gene <gene@jenshen.(none)>
Date: Wed Jan 18 11:41:23 2012 -0600

    Revert "drm/radeon/kms/atom: fix possible segfault in pm setup"

    This reverts commit 4376eee92e5a8332b470040e672ea99cd44c826a.

commit e9d23be2708477feeaec78e707c80441520c1ef6
Author: Greg Kroah-Hartman <email address hidden>
Date: Thu Jan 12 11:54:29 2012 -0800

    Linux 3.0.17

commit 6e474bc4d5c0d94c070a18901e00ff33de0c1be1
Author: Xi Wang <email address hidden>
Date: Mon Dec 12 21:55:52 2011 +0000
----------------------------
Then I used the generic ubuntu config with one finishing touch disabling the M$ hyperv drivers build:
diff:
------------------
3c3
< # Linux/i386 3.0.17-nonrad Kernel Configuration
---
> # Linux/i386 3.0.14-030014-generic Kernel Configuration
5369c5369,5374
< # CONFIG_HYPERV is not set
---
> CONFIG_HYPERV=m
> CONFIG_HYPERV_STORAGE=m
> CONFIG_HYPERV_BLOCK=m
> CONFIG_HYPERV_NET=m
> CONFIG_HYPERV_UTILS=m
> CONFIG_HYPERV_MOUSE=m
------------------

Finally, the binary turned out to be 5 MB larger than the generic one!? This might be because my version of gcc is newer.

OK, the thing is that it seems to be stable. Have had no freezes yet.
So QED and it might be one of the removed patches causing the problems???

--

[gene (eugenios)]

Changed the subject since the problems occur on 3.* versions as well. I am not sure if the cause is the same.
The afore mentioned 3.0.17 build is still pretty stable. AS far as the the deb files difference are considered the buid options might be different I build my kernels with the command:

~$ fakeroot make-kpkg --initrd --append-to-version=-custom kernel-image
and using the ccache package.

The contents of both generic and my own packages are a little different different.

[gene (eugenios)]

I finally got a kernel freeze. When suing one of he compiz plugins (show all workspaces) the screen turned dark with only backlight on. No response to keyboard action no working ssh, no trace in the logs.
So where do I go now?

--

[penalvch (penalvch)]

gene, thank you for reporting this bug and helping make Ubuntu better. If you could also please test the latest upstream kernel available that would be great. It will allow additional upstream developers to examine the issue. Refer to https://wiki.ubuntu.com/KernelMainlineBuilds . Once you've tested the upstream kernel, please remove the 'needs-upstream-testing' tag. This can be done by clicking on the yellow pencil icon next to the tag located at the bottom of the bug description and deleting the 'needs-upstream-testing' text. Please let us know your results.

Thanks in advance.

[Launchpad Janitor (janitor)]

[Expired for linux (Ubuntu) because there has been no activity for 60 days.]