X fails to start on Radeon HD 2400 Mobility (Ubuntu 8.10 RC)

Bug #289026 reported by michael mestre
36
This bug affects 4 people
Affects Status Importance Assigned to Milestone
xserver-xorg-video-ati (Ubuntu)
Fix Released
High
Unassigned
Nominated for Hardy by mitkotaichi
Intrepid
Invalid
Undecided
Unassigned
Jaunty
Fix Released
High
Unassigned

Bug Description

On Ubuntu 8.10 RC desktop :

I have a "Radeon HD 2400 Mobility" graphics card on an Asus Z99S laptop.
When trying to install Ubuntu 8.10RC, it tries to launch X, the graphical display turns on and I see funny shapes (just like what happens when I start X normally with Ubuntu 8.04 in the first few seconds), and then I see a black screen with a blinking cursor.

After 1mn, I see the error message "X failed to start after 60s", and I get a console screen.

I would like to point out that I use Ubuntu 8.04 on the same hardware without any issues.
I have included the outputs from lspci -vvnn and dmesg.

Thanks in advance for your help.
[lspci]
00:00.0 Host bridge [0600]: Intel Corporation Mobile PM965/GM965/GL960 Memory Controller Hub [8086:2a00] (rev 03)
     Subsystem: ASUSTeK Computer Inc. Unknown device [1043:1517]
01:00.0 VGA compatible controller [0300]: ATI Technologies Inc Mobility Radeon HD 2400 [1002:94c9] (prog-if 00 [VGA controller])
     Subsystem: ASUSTeK Computer Inc. Unknown device [1043:15b2]

Revision history for this message
michael mestre (michael-mestre) wrote :
Revision history for this message
Luke Faraone (lfaraone) wrote :

Thank you for your bug report. Could you please boot up in single-user-mode ("recovery mode" from GRUB) and select "Fix X" (or something similar) and see if this solves your issue?

Do you have any proprietary drivers installed?

Changed in xorg:
status: New → Incomplete
Revision history for this message
Luke Faraone (lfaraone) wrote :

(the following was received in a email mistakenly sent to me, rather than to the bug, and I've replied inline)
On Sat, Oct 25, 2008 at 18:41, Michael Mestre <email address hidden> wrote:
> Thanks for your reply. Actually, the problem occurs for the installer
> ; in this case, would selecting "fix X" do any good ? (since the
> version I am trying to install has not been installed yet, by
> definition :o)

Ah, sorry, I missed that. This _is_ quite a problem, then.

> Also, I have an installed ubuntu 8.04 that works very well and I don't
> want to break it by.
>
> I don't know about proprietary drivers selected by the installer, but
> I didn't select any on purpose.

Changed in xorg:
status: Incomplete → New
status: New → Confirmed
status: Confirmed → Triaged
Revision history for this message
michael mestre (michael-mestre) wrote :

Hi,

Little update :

I tried the 8.10 final release today, and this bug affects it too in the same manner.
(probably not so surprising since it hasn't been fixed yet).

Michael

Revision history for this message
ajw107 (alex-wood-bcs) wrote :

I have a Toshiba Satellite Pro p200 with the same graphics card, and I too can not get x server to start. Looking at xorg.0.log (attached) it does warn that the R600 chip set support is experimental, direct rendering is disabled, hardware acceleration is disables, some other vendor specific warnings, then finally it fails some ParseTable thing. Using the Vesa driver also fails, and just brings you to the command prompt, this time the driver seems to work, but it fails due to no screens being found (a generic one is set up, as per default, so no idea why it can;t find one) and fails with a fatal error also. Son in short, no matter what I do, I can;t get the thing to even boot to a gui. Even wireless is acting up for some reason, so I can run apt-get update, aghhhhh! This is tghe same on a distro upgrade, live CD and now a fresh install (from both normal and Alternative CDs). What on earth can I do, as even if they do create a bug fix, unless they bring out a new cd image I won't be able to apply it!!!

Revision history for this message
ajw107 (alex-wood-bcs) wrote :

Minutes after I posted, and after a whole night of looking last night, I eventually found the answer.

1) download the radeonhd driver to a usb drive:
http://packages.ubuntu.com/intrepid/xserver-xorg-video-radeonhd

2) boot to the command line (like you have a choice) and mount the drive:
mkdir /media/USBDRIVE
mount /dev/sdX1 /media/USBDRIVE
where the X will vary, but when you plug in the drive a message should pop up and tell you that something was shoved in sb(something) and this is it

3) install the deb package and reboot
sudo dpkg -i /media/USBDRIVE/xserver-xorg-video(blah blah blah).deb
sudo reboot
obviously don;t type the blah part, just whatever the full file name is (tab is your friend here)

And there you go, very quickly you should have a login GUI screen, i;ve included the deb file as an attachment in case the version above upgrades and things break (how could that happen, ah the irony).

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

please try booting jaunty alpha2 livecd.

Changed in xserver-xorg-video-ati:
status: Triaged → Incomplete
Revision history for this message
William Cattey (wdc-mit) wrote :

ajw107: If I understand how upstream is going, the radeon-hd driver replacement may
solve your problem for now, but will not be the stable long term solution.

I believe the world is trying to get all the latest and greatest stuff into the "radeon" code base.

Timo: Even if jaunty fixed the problem, it's from a whole X universe that's rather irrelevant to solving the problem for Intrepid.

Is there a newer upstream radeon driver that could be tried with Intrepid?

Bryce Harrington (bryce)
Changed in xserver-xorg-video-ati:
status: Incomplete → New
Revision history for this message
Bryce Harrington (bryce) wrote :

Often the Xorg Edgers site will have newer drivers - https://edge.launchpad.net/~xorg-edgers/+archive

However, the nature of this bug sort of sounds like it might need something more than just -ati to fix, so testing against Jaunty really would be the most desirable. Then, if the issue still exists, we can more easily report it upstream.

Changed in xserver-xorg-video-ati:
status: New → Incomplete
status: Incomplete → New
status: New → Incomplete
Bryce Harrington (bryce)
Changed in xserver-xorg-video-ati:
status: Incomplete → New
status: New → Incomplete
Revision history for this message
William Cattey (wdc-mit) wrote :

I burned a Jaunty Alpha DVD from from the 13-Jan-2009 snapshot, booted it, and it came up just fine. I'm typing this comment on that system.

Attached is the Xorg.0.log file of the successful run. (Just in case someone wants it.) It shows that the Radeon HD 2400 XT chip is correctly detected, that the radeon driver is correctly started, that the EDID transfer of the monitor is correctly performed, and that the expected 1680x1050 resolution is set. (Although it is a mystery to me why it probes modes SO man times.)

So, as expected, testing the exact same hardware with the "completely re-designed for Jaunty X configuration" code does NOT manifest the problem.

What more do we need to investigate getting an update to Radeon and the Interpid X server so that it will work?

Revision history for this message
William Cattey (wdc-mit) wrote :

Using the Intrepid Live CD, I booted, and indeed the X server does not come up.
A temporary kludge is to add the line:
            Driver "vesa"
to the Device section.

I have attached the Xorg.0.log file of the failed start-up under Intrepid to this comment.

Revision history for this message
William Cattey (wdc-mit) wrote :

Now have Xorg.0.log output for identical hardware with Success under Jaunty Alpha and Failure under Intrepid.

Changed in xserver-xorg-video-ati:
status: Incomplete → Confirmed
Bryce Harrington (bryce)
description: updated
Revision history for this message
William Cattey (wdc-mit) wrote :

Bryce Harrington:

I went to the xorg-edgers archive.
There were no ATI driver updates labeled Intrepid. In fact the stuff there was almost exclusively jaunty and hardy.

I'm PRETTY sure this is nothing more than getting a little more recent upstream version of the radeon driver. Red Hat Enterprise 5.3 doesn't have this problem. SURELY Ubuntu Intrepid can get a driver as good as what they are shipping.

Revision history for this message
William Cattey (wdc-mit) wrote :

So a few days have gone by with no further comments. I think the correct way to address this problem
is to take a more recent epoch of the upstream, open source ATI driver.

Can someone point me to the relevant information so that I can begin making this happen, if not for everyone using Intrepid, at least for MIT until Jaunty is released? The hardware that stopped working at all is what
MIT has standardized upon for all desktop purchases for the past year. So this is important to us.

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

Hi William,

Well, if nothing else you can just grab the source package for xserver-xorg-video-ati and build it against the Intrepid xserver to produce the debs you need.

I'm subscribing tormod to this bug; he may be willing to help you in doing this.

Thanks for testing that the problem no longer occurs for jaunty; I'll rejigger this bug to indicate the problem is intrepid-only.

Changed in xserver-xorg-video-ati:
status: Confirmed → Fix Released
status: New → Confirmed
Revision history for this message
Tormod Volden (tormodvolden) wrote :

I have backported drivers in my PPA: https://launchpad.net/~tormodvolden/+archive/ppa

Revision history for this message
William Cattey (wdc-mit) wrote :

Wow! Thank you very much Tormod! We'll test with the newer ATI driver and see if it fixes
the problem for us.

Revision history for this message
Florian Sesser (floriansesser) wrote :

Hi!

Tormod: Your backported radeonhd driver fixes the issue for me. Thank you!

(I had the same / a similar problem of 8.04 working but 8.10 not w/ my Radeon HD 2600 (RV630PRO) connected to dual DVI dislpays. Neither radeonhd nor fglrx worked for me, both kind of crashed at Xorg start.)

Thanks!

Revision history for this message
William Cattey (wdc-mit) wrote :

Florian: I was under the impression that the stock radeonhd driver would also resolve the issue.

I've been advocating for an updated radeon driver so that we continue to travel in the direction of the code base that is regarded as canon.

(I still have the action item to try the back-ported radeon driver. Sigh.)

Revision history for this message
William Cattey (wdc-mit) wrote :

Ok, Here is what I did:

Booted the Intrepid Live CD.
Watched as X failed to configure. (See Xorg.0.log output attached above.)
Hit <ctrl><alt><f1> to get a TTY
Logged in, Configured the Network by hand (cuz I'm not registered for DHCP).

Fetched Tormod's backport:
xserver-xorg-video-radeon_6.10.0.99+git20090114.0e69068a-0ubuntu0tormod~intrepid_i386.deb
(from http://ppa.launchpad.net/tormodvolden/ubuntu/pool/main/x/xserver-xorg-video-ati/)

did dpkg -i to install solely that file.

Hit <ctrl><alt><f7> to get the X connection.
Hit <ctrl><alt><backspace> to restart X.

Watched as X configured and started up just fine.

Attached is the Xorg.0.log file from the successful start-up, changing ONLY the rev of
xserver-xorg-video-radeon

Revision history for this message
Evan Broder (broder) wrote :

So this seems like something that would be ideal to get into an SRU, but figuring out the minimal patch to fix the package seems non-trivial:

remus:~/src evan$ debdiff xserver-xorg-video-ati_6.9.0+git20081003.f9826a56-0ubuntu2.1.dsc xserver-xorg-video-ati_6.10.0-1ubuntu1.dsc | diffstat | tail -n 1
 75 files changed, 60043 insertions(+), 1850 deletions(-)

Additionally, the Ubuntu-specific changes to the package don't seem to be in the git repo listed under Vcs-Git in the control file, so you couldn't very easily try to bisect the problem.

As far as I can tell, there are 2 approaches to getting this fixed in a way that works for the MIT environment:

1. Attempt to find the minimal changeset and develop an SRU.

  Is the git repo (or not-git repo) for Ubuntu available anywhere so that we could attempt to bisect the problem?

2. Get 1:6.10.0-1ubuntu1 into intrepid-backports.

  I'm not sure how hard this would be to do - the newer version depends on a version of xserver-xorg-core that's newer than the one in intrepid

(1) is clearly a better option than (2), but we can't afford to block on (1) getting fixed if it's not going to. Should we continue to pursue this ticket, or should we start on the backporting effort?

Revision history for this message
BCoelho2000 (bcoelho2000) wrote :

Hi everyone!

I'm trying to install Kubuntu 8.10 like this:
1. Insert the Kubuntu8.10 cd
2. choose the "Install Kubuntu" option
3. X failed to start and send me to the shell
4. Solved the bug reported here using the solution provided by Tromod Volden
5. Sucessfully started x using "startx" command

Problem: How do I finish my Kubuntu 8.10 installation? It seems that it's running from the cd...

Thank you for your time.

Revision history for this message
Alex Valavanis (valavanisalex) wrote :

Intrepid Ibex reached end-of-life on 30 April 2010 so I am closing the
report. The bug has been fixed in newer releases of Ubuntu.

Changed in xserver-xorg-video-ati (Ubuntu Intrepid):
status: Confirmed → Invalid
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.