xserver starts but display is corrupt (i830M,845G)

[Dirk Niggemann]

Binary package hint: xserver-xorg-video-intel

When attempting to use the new xserver-xorg-video-intel on a Fujitsu Stylistic 4110P with i830 graphics, I receive screen corruption (blank screen with vertical or horizontal banding which corresponds roughly to currently displayed screen in brightness). The X server does not crash and appears to continue working whilst it fails to display correctly (in other words, it appears to be attempting to display the log-in prompt).

xserver-xorg-video-i810 works correctly but I am trying to change over because it appears impossible to configure it to use more than 8m video memory and to support dual displays.

[Jamie Lokier (jamie-shareable)]

This might be the same as bug #124946. I downgraded to an older driver version, 2:2.0.0-1feisty1, which I think comes from "deb http://www.burtonini.com/debian/ feisty/". You might try adding that to your /etc/apt/sources.list, then downgrading to that version in Synaptic (using "Package -> Force Version" followed by "Apply" followed by "Package -> Lock Version"), and see if it makes your display work.

[Bryce Harrington (bryce)]

Can you test this against Gutsy-final? We fixed several issues relating to intel screen corruption, so want to rule out that this has already been addressed.

[Dirk Niggemann (dirk-niggemann)]

Hi , Bryce,

Can't test before middle of next week, I'm afraid
The machine i run this on is used only for lighting control and is in my
loft at the mo.

Let me see what I can do.

Dirk

Bryce Harrington wrote:
> Can you test this against Gutsy-final? We fixed several issues relating
> to intel screen corruption, so want to rule out that this has already
> been addressed.
>
> ** Changed in: xserver-xorg-video-intel (Ubuntu)
> Status: Triaged => Incomplete
>
>

[furry_gizmo (stephen-pountney)]

Hi Guys,

I too have a Fujitsu Scenic with integrated Intel graphics, and I have a similar problem with the release version of 7.10. It was a clean install of the OS, not an upgrade or dual boot.

The symptoms are; From power up, the Ubuntu splash screen and progress bar are displayed ok. Once the progress bar reaches the end and the splash screen closes, there is a brief delay then the display corrupts with vertical green, white, blue and black lines. The circular mouse cursor with the rotating dots is also displayed but the dots and cursor freeze. The display then flashes once to black then the vertical lines are displayed again, in a different pattern. The PC freezes at this point and I have to power off to restart it.

I can get round this by tapping the F8 key several times as soon as the splash screen closes, and the PC boots normally. I don't know if the F8 key does anything specific or if any F key could be used.

I also think that if you reboot the PC, it restarts as normal. It is only when you power up from cold that this happens, but I haven't confirmed this yet.

I don't have the exact details of the PC to hand, but I noticed the graphics driver was marked as 'experimental'.

I hope this helps. Let me know if you want the exact specs of the machine or more info.

Regards

Steve

[Peter Clifton (pcjc2)]

furry_gizmo: Could you post the config and log files from:
/etc/X11/xorg.conf
/var/log/Xorg.0.log (Ideally from when running with the corruption)
or if you get corruption, then reboot (and a new X server has started), /var/log/Xorg.0.log.old

[furry_gizmo (stephen-pountney)]

Hi Peter,

Here is the file you asked for. There are also 3 other log files, Xorg.0.log.old from yesterday, Xorg.20.log and Xorg.20.log.old from Monday. Do you want any of these files too?

Steve

[furry_gizmo (stephen-pountney)]

Peter,

Here is the Xorg.conf file.

Steve

[Peter Clifton (pcjc2)]

If any of the other log files end in a crash report / backtrace, attach them, otherwise I don't think we'll learn anything new. If you have a digital camera, could you take a picture of the screen corruption and attach that please?

[Peter Clifton (pcjc2)]

Can you try with the following driver:

http://www2.eng.cam.ac.uk/~pcjc2/ubuntu/xserver-xorg-video-intel_2.1.1-0ubuntu10~pcjc2.2_i386.deb

This encorporates all other unreleased fixes I've been working on for the intel driver. Chances are it'll have no effect, but its worth ruling out the possibility the bug is triggered in a similar way to one of those fixed already.

Please let me know if this changes anything.

[furry_gizmo (stephen-pountney)]

Hi Peter,

Here is a photo of the corruption.

I've just installed your driver, so I'll try it out in a few moments.

Regards

Steve

[furry_gizmo (stephen-pountney)]

Peter,

Sorry to say that the new driver hasn't solved the problem, however, there is a slight difference in what happens.

Previously, once the splash screen progress bar reached the end, the display would go black, the round cursor with the spinning dots would appear, then after a few seconds the cursor would freeze. The display would then corrupt. Then after another second or so, the screen would change to black again then the corruption would re-appear.

This time, everything happens the same with the exception that once the screen goes black after the initial corruption, it stays black.

I've left the PC running to see if it continues to the login screen, but it never gets there (you don't hear the login drums).

I've attached another log file for you which seems to have some errors listed.

I hope this helps.

Regards

Steve

[furry_gizmo (stephen-pountney)]

Here is a log file with details of the video modes etc.

Steve

[Peter Clifton (pcjc2)]

I think the most intersting / telling section of the log is:

(WW) intel(0): PRB0_CTL (0x0001f001) indicates ring buffer enabled
(WW) intel(0): PRB0_HEAD (0x00000154) and PRB0_TAIL (0x00001490) indicate ring buffer not flushed
(WW) intel(0): Existing errors found in hardware state.

furry_gizmo, I'm wondering either if this is left over state from the framebuffer, or the BIOS. Are any BIOS updates available for your laptop?
Does the corruption occur if you disable the splash screen? (Try booting with vga=normal or just with the vga=??? boot command removed).

[Jamie Lokier (jamie-shareable)]

You might also try this package (which fixed a different and much less severe problem which the package suggested earlier in this bug doesn't fix), just in case it works: http://people.ubuntu.com/~bryce/Testing/intel/xserver-xorg-video-intel_2.1.1~git20071004-0ubuntu1_i386.deb

[furry_gizmo (stephen-pountney)]

Hi guys,

Sorry, I have been away for a while, and I'll be away again shortly.

There is a BIOS update for my motherboard, but I haven't installed it yet as it needs 'Windows'. I'll try and get it done in another week or two. I haven't tried the boot with no splash screen yet either.

As for the package Jamie suggests, I tried to install it, but the PC seems I have a later version already installed.

I'll let you know how I get on in another couple of weeks.

Steve

[Curaga (lauri-kasanen)]

I have this too, with a 855G.

Setting depth to 16bits had no effect, but disabling acceleration ("noaccel" "true") fixed it..

So I must think it's in the acceleration code..

[Curaga (lauri-kasanen)]

Spoke too soon - it only removed the banding, but didn't affect distorted colors.

I didn't notice that, as plain X is b&w..

[Curaga (lauri-kasanen)]

Just a report - using the i810 driver works

[Bryce Harrington (bryce)]

Can someone test this against Hardy Alpha-1? It's got the new -intel 2.2 driver.

[Rolla Selbak (rolla-n-selbak)]

Just a quick note about the upstream direction for Intel gfx...

The -intel driver has support for i810 and forward chips. Some laptops with i845 chips will need a fairly recent driver update to get support for their panels, but at least some people have reported that these work today.

There are no plans to keep up with -i810. All pre-855 support should be present in -intel (and therefore any bugs found should be filed against -intel to get correct attention).

[Tobey (tobey-chao)]

 Bryce Harrington:

  I´ve got the new intel driver on Hardy Alpha-2.

It gives me the right resolution. But ¨screen and graphics¨ shows bothing.

And in ¨desktop resolution¨, no other resolution can be achieved except the highest--1400*1050

Another problem is that with compiz enabled, my mouse scrolling becomes slow and hard to control.

But with i810, I can only got the lowest resolution--640*480.
Since with i810 in Gutsy, compiz works much smoother than intel in Hardy, I consider that this is a intel driver problem.
PS: intel doesn´t work in Gutsy here.

[Bryce Harrington (bryce)]

Curaga, Tobey, and any one else commenting on this bug; please make sure to include your Xorg.0.log, xorg.conf, and a photo of the corruption you're seeing. Thanks!

[Bryce Harrington (bryce)]

Dirk, your issue has been reported as upstream bug 14065: https://bugs.freedesktop.org/show_bug.cgi?id=14065

furry_gizmo, your issue has been reported as upstream bug 14066: https://bugs.freedesktop.org/show_bug.cgi?id=14066

Curaga, as you've provided insufficient information, upstream has refused to look into your issue.

Curaga and Tobey, since upstream apparently wishes to treat each of your issues separately, if you'd like your issues investigated, you must file separate bugs either here on launchpad or directly upstream. Please see http://wiki.ubuntu.com/X/Debugging for directions on what files and data are necessary to include in bug reports.

[Axel Pospischil (apos)]

I can confirm this bug on my Thinkpad R31 with an Intel Corporation 82830 CGC (i830M) Chipset when starting the Hoary (8.04 LTS) Alpha3 Desktop CD.

1. The screen get's completely white after some seconds of being spotted and patchy with white, red, blue and yellow dots.

2. It is possible to switch to the vga console via <Strg+Alt+F[1-6|8]. When doing this, the screen is scrambled for a few seconds, but the console becomes completely useable after this short period of time.

3. Gdm seams to be running: "ps x" shows the according desktop programs (panel, applets, deskbar,...).

4. It is not possible to get X11 run by modprobing e.g. i810 an alter the xorg.conf. If I do this, and e.g. restarting gdm or <Strg+Alt+BAckspace>, I am getting the curser "X", and a little (scrambled) info "ubuntu is running in low graphics mode" after 3 "auto-attempts" to start xorg. The screen gets immediately unusable, frozen, and scrambled again like in [1.].

I included a file with two screenshots showing the white screen, and a second one after switching aroud. The second one shows a "frozen" and scrambled sreenshot of the desktop (I only took a picture of the upper left corner) which becomes white after a short period of time as described in [1.].

Other quick infos about the system:

# dpkg -l xserver-xorg-video-intel
ii xserver-xorg-video-intel 2:2.2.0+git20080107-1ubun X.Org X server -- Intel i8xx, i9xx display driver

# lsmod | grep agp
intel_agp 25364 1
agpgart 34760 1 intel_agp
( drm is NOT loaded automaticaly)

[Axel Pospischil (apos)]

Addendum:
Forgot to mention, that the BIOS checksum failure in Xorg.0.log is just due to a wrong setting in the brightness setting after a battery change. The system works with my major gutsy installation and "drm/i810/nosplash", and also with a second SuSE 10.3 installation.

Kernel is on hoary/alpha3 is linux-image-generic, 2.6.22.14.21.
I am booting with "irqpoll" and "i8042.nomux=1", but double blind ( ;) ) tested without these settings - same problems.

Always (!) had touble with this chipset and x11/xorg > 6.8. But since 7.2 things are getting really worse and make the system almost unusable.

With the hope this will ever be able to be fixed.
Cheers Axel

[Axel Pospischil (apos)]

Attached logs for my hoary/alpha3 system.

[Axel Pospischil (apos)]

SOOORRRY - Naturally I am comenting on the HARDY/alpha 3, not the hoary ubuntu !!!!
Unfortunately one cannot change any the comments - or is there a way?

[Bryce Harrington (bryce)]

Dirk, would you be able to compile the -intel from git HEAD at freedesktop? It sounds like they believe the issue is fixed there, so would be good to confirm.

Alex, it would also be interesting to hear if current -intel git head solves your problem, if you're able to compile it? If it does not fix it, please report your issue as a separate bug.

[Axel Pospischil (apos)]

Bryce, thanks for your answer. Due to time issues i am not able to test the current intel git HEAD (http://www.x.org/wiki/IntelGraphicsDriver?highlight=%28intel%29). I will do this at mid of Feburary.

So long.

By the way. After an update-manager upgrade (xserver-xorg) the same(https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/114331/comments/26) happened on feisty/7.10. I downgraded back to the original 7.10 installation CD/DVD packages of "xserver-xorg-video-*".

[Axel Pospischil (apos)]

I mean gutsy / 7.10. Really seam to had to much coffee ;)

[Bryce Harrington (bryce)]

Dirk,

Can you please join the discussion at http://bugs.freedesktop.org/show_bug.cgi?id=14065 to get your bug resolved? Otherwise they will be rejecting the bug.

[Colin Macdonald (cbm755)]

Alex's issue is fixed in xserver-xorg-video-intel 2:2.2.0+git20080107-1ubuntu1:

I have the same hardware as Alex (thinkpad r31) and I was experiencing the same issue. X was using the vesa driver (poorly: fwiw toggling screen expansion fn-f8 would make it usable at 800x600).

I updated today to xserver-xorg-video-intel 2:2.2.0+git20080107-1ubuntu1 and now X works correctly using the intel driver (didn't try i810 as intel was autodetected).

I can't tell if this is the same issue as Dirk so I haven't commented on the bug 14065 upstream.

Now X works fine (modulu exa as in #177492: I changed my xorg.conf to use XAA and now X works as well as it does on my Gentoo partition, and as well as it did in Gutsy).

[Colin Macdonald (cbm755)]

sorry I think I meant in xserver-xorg-video-intel - 2:2.2.0+git20080107-1ubuntu2. (still getting used to these debian packaging schemes).

At any rate: the one released approx 20080129.

[Bryce Harrington (bryce)]

Stephen Pountney, you reported a problem you're having in the comments to https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/114331. I've reported your issue upstream to the -intel driver developers and they believe they have a fix for your problem, in the form of a patch to xserver. Are you able to build xserver and apply patches? If not, let me know and I can produce a patched version for you to test next week. Please also join the discussion at the upstream bug at http://bugs.freedesktop.org/show_bug.cgi?id=14066.

[Axel Pospischil (apos)]

Hi again,

just downloaded Ubuntu alpha4 and booted the live CD on my Thinkpad R31 with my Intel Corporation 82830 CGC (i830M) - see above. That showed a running xserver-xorg-video-intel - 2:2.2.0+git20080107-1ubuntu2 installed (like Collin telled us above).

Can confirm that Xorg works fine now.

@Collin
  Could not try https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/177492 on my machine, but I hope this will stabilize my sloppy 3D acceleration. >>>> Could you please upload your xorg.conf and /etc/environment?<<<< Thanks in advance.

[Colin Macdonald (cbm755)]

@Alex: I haven't played with #177492 yet. I haven't modified /etc/environment and my xorg.conf is stock except for adding the following to Section "Device"
 Driver "intel"
 Option "AccelMethod" "XAA"
Feel free to contact me out-of-band if you want to discuss it further (I doubt if devs like launchpad bugs being filled with chat!).

[VincentRC (vreneco1)]

Hi evebody,

Here is a copy of what I have written here in https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/178837 :

( https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/178837/comments/13 )

Hi everybody,

I have a problem whch seems to be the same.

I have an old computer with an old computer with an Intel 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device, and the X server won't launch.

It tries and tries again, but my CRT screen won't turn on.

I have read all the comments and they seem to say that it has been fixed, but it still does not work for me in Hardy Alpha 4. It only works when I boot in safe graphic mode.

The only distribution that work without any problem (out of the box) for me is Mandriva 2008, and yes, 16bit depth is automatically forced, but everything works fine, compiz-fusion too.

Here is my xorg.conf file from Mandriva 2008 (see original post for the file).

[Curaga (lauri-kasanen)]

Sorry Bryce, I no longer have that machine, so I cannot provide more information.

[Bryce Harrington (bryce)]

furry_gizmo, could you join the discussion on the upstream bug report for your issue? They have a patch they'd like you to test if you could.

> furry_gizmo, your issue has been reported as upstream bug 14066: https://bugs.freedesktop.org/show_bug.cgi?id=14066

[unggnu (unggnu)]

This might connected to Bug #151044 and Bug #192088.

[Bryce Harrington (bryce)]

furry_gizmo, the patch has been committed upstream and they've closed the bug as fixed, so without further feedback we'll assume your problem has been taken care of now.

[unggnu (unggnu)]

The Hardy Heron Alpha series is currently under development and contains an updated version of the -intel driver. You can download and try the new Hardy Heron Alpha release from http://cdimage.ubuntu.com/releases/hardy/ . You should be able to test the new -intel driver using the LiveCD. If you can, please verify if this bug still exists or not and report back your results. General information regarding the release can also be found here: http://www.ubuntu.com/testing/ .
Thanks.

[Bryce Harrington (bryce)]

Alright, I think we can probably close this one. Several fixes have gone in upstream which I believe are all included in the version we have in Hardy (released feb 23rd.) For some of the other cases the reporters are unfortunately unable to test, but hopefully the upstreamed fixes cover them too. There's been no recent comments, which I guess is a good sign.

Anyway, if anyone is still experiencing this issue, please feel free to reopen this bug (although maybe it'd be better to open a new bug). Include your Xorg.0.log and a photo or screenshot showing the corruption.