[lucid] HD3200 intermittently does not start, resulting in low-graphics mode

[Ben Blout]

Binary package hint: xorg

I have been running Lucid for about three weeks. Roughly five days ago, on a reboot, X did not start properly, and I was presented with a recovery screen, which allowed me to use a low-graphics mode for that session only.

However, it turns out that X starts perfectly roughly three out of every four boots. I have no idea what might be causing this intermittent behavior. I doubt it is a hardware problem - I have been using this machine continuously with no problems for about a year.

The xorg log attached automatically by ubuntu-bug shows what happens when I am thrown into low-graphics mode. The 'old' xorg log attached by ubuntu-bug is from my previous boot - it shows a successful startup.

Please let me know what additional information I might supply to help troubleshoot this odd (at least to me!) problem. In case it is helpful, I have attached the section of my syslog from this startup.

Also starting about five days ago, I lose all video during the boot - for a length of time my monitor displays 'no signal'. So, for example, if a fsck is run during the boot, I have no idea what is happening, as there is no video output during that stage of boot.

(II) [drm] DRM interface version 1.0
(EE) [drm] Could not set DRM device bus ID.
(EE) RADEON(0): [dri] DRIGetVersion failed to open the DRM
[dri] Disabling DRI.
(EE) RADEON(0): Kernel modesetting setup failed
(II) UnloadModule: "radeon"
(EE) Screen(s) found, but none have a usable configuration.

ProblemType: Bug
Architecture: i386
Date: Fri Feb 12 22:15:25 2010
DistroRelease: Ubuntu 10.04
DkmsStatus: Error: [Errno 2] No such file or directory
InstallationMedia: Error: [Errno 13] Permission denied: '/var/log/installer/media-info'
MachineType: ECS A780GM-A
Package: xorg 1:7.5+1ubuntu2
ProcCmdLine: BOOT_IMAGE=/vmlinuz-2.6.32-13-generic root=/dev/mapper/new--lvm-root ro quiet splash
ProcEnviron:
 PATH=(custom, user)
 LANG=en_US.utf8
 SHELL=/bin/bash
ProcVersionSignature: Ubuntu 2.6.32-13.18-generic
RelatedPackageVersions:
 xserver-xorg 1:7.5+1ubuntu2
 libgl1-mesa-glx 7.7-0ubuntu8
 libdrm2 2.4.17-0ubuntu2
 xserver-xorg-video-intel 2:2.9.1-1ubuntu4
SourcePackage: xorg
Symptom: display
Uname: Linux 2.6.32-13-generic i686
dmi.bios.date: 05/20/2008
dmi.bios.vendor: American Megatrends Inc.
dmi.bios.version: 080014
dmi.board.name: A780GM-A
dmi.board.vendor: ECS
dmi.board.version: 1.0
dmi.chassis.type: 3
dmi.chassis.vendor: ECS
dmi.modalias: dmi:bvnAmericanMegatrendsInc.:bvr080014:bd05/20/2008:svnECS:pnA780GM-A:pvr1.0:rvnECS:rnA780GM-A:rvr1.0:cvnECS:ct3:cvr:
dmi.product.name: A780GM-A
dmi.product.version: 1.0
dmi.sys.vendor: ECS
system:
 distro: Ubuntu
 architecture: i686kernel: 2.6.32-13-generic

[Ben Blout (bdb-new)]

The situation has gotten worse.
About 80% of the time, the system will boot, and I just have a black screen.
The other 20% of the time, I get low-graphics mode.

If I revert to VESA drivers, the situation is NOT improved.

If I boot with the 2.6.32-10 kernel, I have no problems.

[Ben Blout (bdb-new)]

It seems this may be related to bug #387013, but I have a different card, and I am not using the DVI connection.

[Pauli (paniemin)]

Please try daily mainline-ppa https://wiki.ubuntu.com/KernelTeam/MainlineBuilds and you will need a special firmware http://xorg.freedesktop.org/wiki/radeonBuildHowTo#TroubleshootingExtraFirmwareforR600.2BAC8-R700

This sounds like null pointer deference in init failure path that was fixed in 2.6.33-rc7 or rc8. Also firmware loading problem might be fixed in newer kernel.

[Ben Blout (bdb-new)]

I installed the daily kernel from mainline-ppa. First time I've done it, but I followed instructions. I also installed the firmware.

uname -a reports 2.6.33-999-generic #201002201003 SMP Sat Feb 20 10:58:34 UTC 2010 i686 GNU/Linux

I have rebooted once so far, and no problems! I'll be able to reboot a few more times tomorrow, which will be a better test, since the problem was intermittent.

[Ben Blout (bdb-new)]

Ten reboots total, and a smooth boot every time. The newer kernel does seem to fix this bug.

If it is helpful, from the descriptions, the people reporting bugs 413686 and 495973 may be seeing the same problem.

[Bryce Harrington (bryce)]

Looks like something is wrong with the kernel DRM. We're likely to pull in the 2.6.33 DRM which may well fix this. You can test it by installing one of the mainline kernel builds.

Meanwhile, popping this over to the kernel team.

[Ben Blout (bdb-new)]

Bryce Harrington wrote:
>You can test it by installing one of the mainline kernel builds.

As I alluded to in #5, I did try a mainline build, and it fixed this problem completely.
I am now using
2.6.33-999-generic #201002201003
I am happy to try other kernels if that would help.

[Ben Blout (bdb-new)]

I had another ATI bug, #526181.
To help troubleshoot that bug, I switched to a different kernel, with a DRM update.
https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-ati/+bug/526181/comments/8

This bug is also fixed with that kernel, as well as with 2.6.33-999 as I described in comment #8.

[Aries K (outatime0001)]

Hi, I am using an Nvidia 7050. I too am having this problem, will the fix for this help Nvidia users as well or should I file a separate bug for it? Thanks in advance.

[Ben Blout (bdb-new)]

I believe you should file a separate bug, and in your description mention this bug. If a developer concludes that it is the same root cause, they will mark your (newly reported) bug as a duplicate of this one. But if it turns out your problems have a different cause, it is better to have a separate bug report.

[Aries K (outatime0001)]

Ok thanks I filed a bug. https://bugs.launchpad.net/ubuntu/+bug/562565

[Leann Ogasawara (leannogasawara)]

Hi Ben,

Is this still an issue for you on Lucid? Reason I ask is because Lucid should have been pulling periodic DRM stack updates from 2.6.33. I assume if you also upgraded to Maverick or Natty it would also be resolved as those ship with a newer 2.6.35 kernel for Maverick and a 2.6.38 kernel for Natty.

[Ben Blout (bdb-new)]

Leann,

You are indeed correct; I've not had problems with graphics during boot for the last six months or so (very rough estimate), so I think this bug is indeed solved.

[Leann Ogasawara (leannogasawara)]

Closing Fix Released per comment #13 from Ben.