[i830] xorg hang on i830 video driver 2.8.0

Bug #403037 reported by jerrylamos
18
This bug affects 2 people
Affects Status Importance Assigned to Milestone
xf86-video-intel
Fix Released
Critical
linux (Ubuntu)
Fix Released
High
Unassigned

Bug Description

Binary package hint: xorg

karmic 2.6.31-3 with intel driver 2.8.0 on IBM Thinkpad R31 i830 video graphics. Hangs after login with brown screen and cursor. Keyboard not responsive except power off.

Karmic was running O.K. fresh install of daily build 20090217. Then did apt-get update, dist-upgrade on 20090722 which included intel driver 2.8.0.

Dual booted Intrepid (distinctly better video response!) to get .xsession-errors, Xorg.0.log, syslog, lspci -vvnn which will be attached following.

I'm attempting to get today's daily build to see if it has 2.8.0 however rsync is running slowly.

Any other data needed? Anything to try?

Thanks, Jerry

Revision history for this message
jerrylamos (jerrylamos) wrote :
Revision history for this message
jerrylamos (jerrylamos) wrote :

1-3 with intel driver 2.8.0 on IBM Thinkpad R31 i830 video graphics. Hangs after login with brown screen and cursor. Keyboard not responsive except power off.

Karmic was running O.K. fresh install of daily build 20090217. Then did apt-get update, dist-upgrade on 20090722 which included intel driver 2.8.0.

Dual booted Intrepid (distinctly better video response!) to get .xsession-errors, Xorg.0.log, syslog, lspci -vvnn which will be attached following.

I'm attempting to get today's daily build to see if it has 2.8.0 however rsync is running slowly.

Any other data needed? Anything to try?

Thanks, Jerry

Revision history for this message
jerrylamos (jerrylamos) wrote :

syslog after install, update, dist-upgrade, and hang

Revision history for this message
jerrylamos (jerrylamos) wrote :

Thinkpad R31 i830 lspci

Revision history for this message
jerrylamos (jerrylamos) wrote :

There wasn't any xorg.conf. Tried to do recovery mode, root prompt dpkg-reconfigure xserver-xorg and still no xorg.conf. I think I remember dpkg might have some problems lately?

Copied in xorg.conf from another boot, then set Option "ModeDebug" "true" and reproduced the error. Attached is the resulting .xsession-errors.

Jerry

Bryce Harrington (bryce)
Changed in xorg (Ubuntu):
status: New → Confirmed
Revision history for this message
jerrylamos (jerrylamos) wrote :

i845 intel video does not boot either. After login, brown screen and cursor. Keyboard entirely unresponsive. That's with karmic 20090722. Will try with karmic 20090723 Alpha 3 as soon as I have time, hopefully tonight.

Jerry

Revision history for this message
jerrylamos (jerrylamos) wrote :

i845 was able to install karmic A3 from grub screen install choice since it doesn't use normal gnome desktop. Any attempt at booting the installed image gets a brown screen after login. Booting in recovery mode goes to the recovery mode choices, then continuing boot results in brown screen after login. Removing splash no help.

Jerry

Revision history for this message
jerrylamos (jerrylamos) wrote :

i845 with intel driver 2.8.0 hang after login

Revision history for this message
jerrylamos (jerrylamos) wrote :

i845 Xorg.0.log after hang on A3 showing intel driver 2.8.0

Jerry

Revision history for this message
jerrylamos (jerrylamos) wrote :

Syslog I think it is trimmed down to the start - hang. Pretty verbose.

Jerry

Revision history for this message
jerrylamos (jerrylamos) wrote :

i845 lspci -vvnn. Couldn't get /proc/version which is blank at time of failure. The .iso is the A3. Installed from CD boot screen which works. CD itself fails same way as the install, however no way to get the logs from the CD. Not having any luck with ssh from another pc on the net.

Jerry

Revision history for this message
jerrylamos (jerrylamos) wrote :

A3 finally booted on i845! Hooray!

e edit the recovery boot line to add "nomodeset"

at recovery mode root prompt, copy in a xorg.conf from previous karmic since dpkg-reconfigure xserver-xorg didn't work. Maybe dpkg is broken.

nano /etc/X11/xorg.conf

add "Driver" "vesa" after Identifier

chmod 644 /boot/grub/grub.cfg

add "nomodeset" to linux line and remove splash, change timeout to 10

reboot It did! Sure looks to me like intel driver 2.8.0 is the problem.

By the way, I did a "fix broken packages" on another A3 boot image which updated xserver-xorg-video-intel to 20090724 however it still doesn't boot. I'll put nomodeset and vesa on that image too.

If there's another update to 2.8.0 I'll try that. Just comment out the "vesa" line and e the linux line at boot to remove nomodeset.

All this doesn't work on the i830 since booting with driver vesa doesn't work on jaunty or intrepid on the i830 either.

Jerry

Timo Aaltonen (tjaalton)
affects: xorg (Ubuntu) → xserver-xorg-video-intel (Ubuntu)
Revision history for this message
In , Bryce Harrington (bryce) wrote :

Forwarding this bug from Ubuntu reporter jerrylamos:
http://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/403037

[Problem]
Hangs after login with brown screen and cursor. Regression with the 2.8.0 driver.

[Original Description]
karmic 2.6.31-3 with intel driver 2.8.0 on IBM Thinkpad R31 i830 video graphics. Hangs after login with brown screen and cursor. Keyboard not responsive except power off.

Karmic was running O.K. fresh install of daily build 20090217. Then did apt-get update, dist-upgrade on 20090722 which included intel driver 2.8.0.

Dual booted Intrepid (distinctly better video response!) to get .xsession-errors, Xorg.0.log, syslog, lspci -vvnn which will be attached following.

Alpha3 booted on my i845 by passing the "nomodeset" parameter and booting "vesa" driver. But this doesn't work on the i830 since booting with driver vesa doesn't work on jaunty or intrepid on the i830 either.

Revision history for this message
In , Bryce Harrington (bryce) wrote :

Created an attachment (id=28066)
xsession-errors.modedebug

Revision history for this message
In , Bryce Harrington (bryce) wrote :

Created an attachment (id=28067)
syslog after hang

Revision history for this message
In , Bryce Harrington (bryce) wrote :

Created an attachment (id=28068)
Xorg.0.log.0722

Bryce Harrington (bryce)
tags: added: karmic
summary: - karmic xorg hang on i830 video driver 2.8.0
+ xorg hang on i830 video driver 2.8.0
Revision history for this message
In , Gordon Jin (gordon-jin) wrote :

How about "nomodeset" with intel 2.8.0 driver?

Bryce Harrington (bryce)
summary: - xorg hang on i830 video driver 2.8.0
+ [i830] xorg hang on i830 video driver 2.8.0
Bryce Harrington (bryce)
Changed in xserver-xorg-video-intel (Ubuntu):
importance: Undecided → High
Revision history for this message
Bryce Harrington (bryce) wrote :

Jerry, I've forwarded your bug upstream to https://bugs.freedesktop.org/show_bug.cgi?id=22982 - please subscribe to this bug in case upstream needs further information or wishes you to test something.

I'm focusing this bug report only on the i830 issue so as not to complicate investigation. Upstream prefers separate bugs for different hardware.

Changed in xserver-xorg-video-intel (Ubuntu):
status: Confirmed → Triaged
Revision history for this message
Bryce Harrington (bryce) wrote :

Also, please obtain a gpu batchbuffer dump after reproducing the hang. Directions on how to do this are available at:
https://wiki.ubuntu.com/X/Troubleshooting/Freeze#How%20to%20Get%20a%20Batchbuffer%20Dump%20(-intel%20only)

Revision history for this message
jerrylamos (jerrylamos) wrote :

Bryce, thanks for the dump directions.

Installed openssh-server from recovery root prompt
installed intel-gpu-tools as in How to Get A Batchbuffer Dump from recovery root prompt
rebooted test system up to the point of login
ssh into root of test system, ssh working O.K.
on test system did user login to reproduce hang
test system hung
ssh does not respond to any key, I don't know how get the dump. ssh is hung as well.

Any ideas?

I am keeping the i830 up to date because from recovery root prompt it will do update & dist-upgrade. It's triple booted with a 2.6.30-9 karmic and an intrepid as well so I can get logs.

I'll enter a bug for the i845 and try to get the dump on it. It fails boot after login also however may not be the same cause.

Thanks,

Jerry

Changed in xserver-xorg-video-intel:
status: Unknown → Confirmed
Revision history for this message
In , jerrylamos (jerrylamos) wrote :

Created an attachment (id=28204)
intel batchbuffer dump

Attached is the batchbuffer dump as requested taken after the hang using ssh.

It's a bit of a bear to ssh since as hung, the mouse pointer has to be moved after every keystroke or copy in order to get any response.

Let me know if there's a problem with the dump and it needs to be re-done.

Thanks, Jerry

Revision history for this message
jerrylamos (jerrylamos) wrote :

Attached is the batchbuffer dump as requested taken after the hang using ssh.

It's a bit of a bear to ssh since as hung, the mouse pointer has to be moved after every keystroke or copy in order to get any response.

Let me know if there's a problem with the dump and it needs to be re-done.

Thanks, Jerry

Revision history for this message
In , Carl Worth (cworth) wrote :

(In reply to comment #5)
> Created an attachment (id=28204) [details]
> intel batchbuffer dump
>
> Attached is the batchbuffer dump as requested taken after the hang using ssh.
>
> It's a bit of a bear to ssh since as hung, the mouse pointer has to be moved
> after every keystroke or copy in order to get any response.

Thanks for your bug report Jerry. The symptom of needing to move the mouse pointer to get a keystroke response via ssh is certainly an interesting one. I'm not aware of having heard that before.

I'll try getting an i830 booting recent Ubuntu here to see if I can replicate the bug.

-Carl

Revision history for this message
Timothy G. Rundle (tgrundle) wrote :

I am experiencing this issue as well, but on a i845. I updated to the latest 2.8.0 driver from Bryce, but still seeing the issue. I am attaching the bacthbuffer dump.

Changed in xserver-xorg-video-intel:
status: Confirmed → In Progress
Revision history for this message
vmc (vmclark) wrote :

I also have the same issue. I have Intel i865 chip. Mine freezes so right now I can't get the dump to work. Once I can I will attach the dump file

Geir Ove Myhr (gomyhr)
tags: added: freeze i830
Revision history for this message
Geir Ove Myhr (gomyhr) wrote :

Timothy and vmc, we don't need any more dumps for this issue. If it turns out that you have the same issue, the fix for this bug will also fix your problem and I would encourage you to follow this and the upstream bug report and test once a fix becomes available.

Revision history for this message
jerrylamos (jerrylamos) wrote :

Not a fix but at least karmic A4 boots at all on i830! This is using 2.6.31-6. A3 never did boot on i830, kept up to date in recovery mode.

I've been using A2 or intrepid. Intrepid is faster than karmic on all of my four pc's on video graphics benchmark GtkPerf from Synaptic.

On A4:

nomodeset on the linux line on grub boot menu. Otherwise karmic boots in 48 line mode like modeset=1 which then hangs up later.

Driver "vesa" finally works on i830!

Let me try the default driver.....

Jerry

Revision history for this message
jerrylamos (jerrylamos) wrote :

This launchpad bug is still very much there exactly as described with the A4 default driver xserver-xorg-video-intel 2:2.8.0-0ubuntu2 which doesn't work on this Thinkpad i830, even with nomodeset in the grub boot linux line. The last karmic that would run with 2:2.8.0 plus PPKA is A2.

Jerry

Revision history for this message
In , Carl Worth (cworth) wrote :

Eric Anholt posted a kernel patch yesterday that fixes several hangs for
pre-9xx chipsets:

http://lists.freedesktop.org/archives/intel-gfx/2009-September/004122.html

Jerry, it would be very useful if you could test with that patch and report
here whether or not it fixes your bug.

Thanks,

-Carl

Revision history for this message
In , Carl Worth (cworth) wrote :

(In reply to comment #6)
> Thanks for your bug report Jerry. The symptom of needing to move the mouse
> pointer to get a keystroke response via ssh is certainly an interesting one.
> I'm not aware of having heard that before.

By the way, I later found out that this was a kernel bug that has also been
fixed since. So hopefully a newer kernel will fix all your problems. :-)

-Carl

Revision history for this message
Axel Pospischil (apos) wrote :

Hi,

i like to add, that running Alpha 6 as of today (18/9/2009) is running fine on my Thinkpad R31/i830. This is without (!) nomodset. I have no problems with hanging mouse cursor, total system hang or else any more but, "top" gives me:

Tasks: 142 total, 3 running, 139 sleeping, 0 stopped, 0 zombie
Cpu(s): 98.0%us, 1.0%sy, 1.0%ni, 0.0%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st
Mem: 1018420k total, 470112k used, 548308k free, 37940k buffers
Swap: 1534640k total, 0k used, 1534640k free, 214444k cached

  PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
 1579 root 20 0 170m 32m 7456 R 97.5 3.3 5:15.24 Xorg

So Xorg uses allmost all resources on the PC!

I am running with this xorg.conf:
Section "Device"
 Identifier "i830"
 Driver "intel"
 Option "DRI" "false"
EndSection

Thanks for your work up to this point!
Please tell me, what kind of information you need to solve the problem.
Or should I file a new bug?

Greets Axel

Revision history for this message
Axel Pospischil (apos) wrote :

Hi,

now I tested deaktivating my Xorg.conf !!!

I got a better result: CPU is "only" about 80-90% CPU usage (intel PIII 1 GHz) when starting e.g. firefox. It goes down to appr. 15% in normal usage after appr. half a minute. That is not acceptable - but system works ;)

All in all performance is very slow, especiallly when starting programms, subscreens.

glxinfo | render gives me:
 direct rendering: Yes
 OpenGL renderer string: Software Rasterizer

Greets Axel

Revision history for this message
Timothy G. Rundle (tgrundle) wrote :

I am still having issues, more severe issues. PC locks before login screen. I can't even SSH in to get a batch bufferdump. I upgraded from Jaunty to karmic alpha 6. I am going to try a fresh install next.

Revision history for this message
Axel Pospischil (apos) wrote :

After a day trying again, the PC locks up again with black screen before login like Timothy wrote.

I tried booting with the i915.modset=0 kernel parameter and the system went through the login screnn, even from the live cd. The only strange thing ist that the boot process stats "Unknown boot option 'i915.modset=0': ignoring" ???

But as I mentioned before, the system performance is eaten up by the Xorg server process with appr. 15%-90%.

Also there is another process called "sreadahead" with takes up the rest of the CPU usage with disappears after a certain amount of time (minutes) after starting the live CD/ installed system.

Greets Axel

Revision history for this message
Axel Pospischil (apos) wrote :

No change in Karmic Beta.

Revision history for this message
Bryce Harrington (bryce) wrote :

[According to the upstream bug, a kernel fix solves this issue, so refiling to linux]

affects: xserver-xorg-video-intel (Ubuntu) → linux (Ubuntu)
Changed in linux (Ubuntu):
status: Triaged → New
tags: added: xorg-needs-kernel-fix
Revision history for this message
In , Eric Anholt (eric-anholt) wrote :

Feedback timeout. If you later find that the bug isn't fixed by that patch, please re-open.

Changed in xserver-xorg-video-intel:
status: In Progress → Fix Released
Revision history for this message
Axel Pospischil (apos) wrote :

As of update of today evening (ubuntu beta)

- xserver-xorg-video-int 2:2.9.0-1ubuntu1
- kernel 2.6.31-13-generic
- i830M Chipset (Thinkpad R31)

and using an empty xorg.conf the system boots now without problems into the graphical user interface.

Using i915.modset=0 does not deliver kernel messages any more and lets start the system smoothly.
Not using it, the system also starts into the graphical (gnome) user interface, but gives some artifacts when the ubuntu login screen appears.

There is no difference in performance whether using i915.modset or not. The performance issues due to my statement in #23 and #25 are gone away completely ;)

Thanks very much for fixing these issues that would have made an update to 9.10 not possible on this notebook.

Greets Axel

Revision history for this message
Leann Ogasawara (leannogasawara) wrote :

Karmic should have the patch mentioned upstream. Marking this Fix Released for the kernel.

ogasawara@yoji:~/ubuntu-karmic$ git show e4fd873b7882285e4bcd76486267003b2f755082

commit e4fd873b7882285e4bcd76486267003b2f755082

Author: Eric Anholt <email address hidden>

Date: Thu Sep 10 17:48:48 2009 -0700

    agp/intel: Fix the pre-9xx chipset flush.

    commit e517a5e97080bbe52857bd0d7df9b66602d53c4d upstream.

Changed in linux (Ubuntu):
status: New → Fix Released
Revision history for this message
jerrylamos (jerrylamos) wrote :

Karmic RC does not hang on Thinkpad R31 i830 or ThinkCentre A30 i845.

Xorg is 2.9.0 and kernel is 2.6.31-14.

Problem solved for me at the moment, now for the next update....

Thanks much, Jerry

p.s. It's another problem, i830 pretty slow on GtkPerf:
Karmic takes 63.13 seconds vs. Intrepid 36.04 seconds. Karmic almost twice as slow.

Major problem is drawing circles:
Karmic takes 33.36 seconds vs. Intrepid much faster at 4.19 seconds. Wow.

Revision history for this message
Petr Dlouhý (petr-dlouhy) wrote :

Xorg still hangs after boot with latest Intel driver on Asus A6L (Intel 82852/855GM). Should I send the logs, or it is not expected to be working yet?

Changed in xserver-xorg-video-intel:
importance: Unknown → Critical
Changed in xserver-xorg-video-intel:
importance: Critical → Unknown
Changed in xserver-xorg-video-intel:
importance: Unknown → Critical
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.