Video driver not working for Savage chipset

Bug #1083032 reported by Richard Hollabaugh
66
This bug affects 11 people
Affects Status Importance Assigned to Milestone
X.Org X server
Fix Released
Medium
xorg-server (Ubuntu)
Fix Released
Undecided
Unassigned
Raring
Won't Fix
Undecided
Unassigned
xserver-xorg-video-savage (Ubuntu)
Fix Released
Medium
Unassigned
Raring
Won't Fix
Medium
Unassigned

Bug Description

[Impact]
For xserver-xorg-video-savage: Display corruption in X on all Savage-equipped laptops such as Thinkpad T22, making the system unusable.
For xorg-server: System lockup on boot.

[Test Case]
Attempt to boot a system with one of the affected Savage cards; the Thinkpad T22 is at least one such system.

The display will be corrupted beyond usability.

Install xserver-xorg-video-savage and xserver-xorg-core from proposed, and you should have a usable X session.

[Regression Potential]
The X server patch touches only the DRI1 initialisation codepath; this is not used for the vast majority of systems, so breakage is low-impact. Breakage here would most likely manifest as an Xserver crash on startup. (Most likely any driver using DRI1 is broken without this patch, but since DRI1 is mostly used for mesa and our mesa does not support DRI1 any longer, this is not very visible. However, the savage driver uses DRI1 for 2D acceleration as well.)

The savage patch switches the default acceleration method to EXA (from XAA) and adds some fixes to EXA support. We don't build XAA support into the X server any more, so switching away from it means enabling EXA acceleration(*), which can cause pretty much arbitrary problems, but is unlikely to. In any case, problems would be restricted to users of xserver-xorg-video-savage.

*) The build-without-XAA fallback to no acceleration is broken, so EXA is the best way out of this. EXA has been reported to work well, and is now enabled by default upstream (the patches in this report are from upstream). The user can still disable (EXA) acceleration in xorg.conf, which should work out fine when done this way.

[Original Report]

When installing 12.10 on our Thinkpad T22 laptops the display does not work. 12.04 works fine.
xserver-xorg-video-s3 is version 1:0.6.5-0ubuntu1
xserver-xorg-video-savage is version 1:2.3.6-0ubuntu1
I don't know if these are the packages involved but they are the ones that I suspect may be involved.
This T22 model (and many other thinkpads from that era) use the "S3 Savage IX8" chipsets.

Using the info from this link makes the display work.
=====
http://www.physics.wustl.edu/~alford/thinkpad/T23_F17.html

Xorg :1 -configure
# ignore Configuration Failed message
cp xorg.conf.new /etc/X11/xorg.conf

Edit the /etc/X11/xorg.conf file: uncomment Option "DisableTile" in the Section "Device" that contains the Savage driver options.
============

Related branches

Revision history for this message
Brad Figg (brad-figg) wrote : Missing required logs.

This bug is missing log files that will aid in diagnosing the problem. From a terminal window please run:

apport-collect 1083032

and then change the status of the bug to 'Confirmed'.

If, due to the nature of the issue you have encountered, you are unable to run this command, please add a comment stating that fact and change the bug status to 'Confirmed'.

This change has been made by an automated script, maintained by the Ubuntu Kernel Team.

Changed in linux (Ubuntu):
status: New → Incomplete
Revision history for this message
Joseph Salisbury (jsalisbury) wrote :

Would it be possible for you to test the latest upstream kernel? Refer to https://wiki.ubuntu.com/KernelMainlineBuilds . Please test the latest v3.7 kernel[0] (Not a kernel in the daily directory) and install both the linux-image and linux-image-extra .deb packages.

If this bug is fixed in the mainline kernel, please add the following tag 'kernel-fixed-upstream'.

If the mainline kernel does not fix this bug, please add the tag: 'kernel-bug-exists-upstream'.

If you are unable to test the mainline kernel, for example it will not boot, please add the tag: 'kernel-unable-to-test-upstream'.
Once testing of the upstream kernel is complete, please mark this bug as "Confirmed".

Thanks in advance.

[0] http://kernel.ubuntu.com/~kernel-ppa/mainline/v3.7-rc6-raring/

Changed in linux (Ubuntu):
importance: Undecided → Medium
Revision history for this message
Richard Hollabaugh (haytown) wrote :

I downloaded and installed via dpkg the linux-image and linux-image-extra packages for 3.7-rc6-raring and 3.7.0-999. Both cause the display to not work as in the kernel for 12.10.
I renamed the xorg.conf file just before rebooting into the test kernel.

tags: added: kernel-bug-exists-upstream
Changed in linux (Ubuntu):
status: Incomplete → Confirmed
Revision history for this message
Peter Emerik (nothin) wrote :

Is there any progress on this? Please let me know if you need any aditional information. Thanks.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

I am not able to reproduce this on my Savage TwisterK, so I wonder if it is limited to certain cards. Can everybody affected please post the output from:
 lspci -d 5333: -nn

I am not running 12.10 on my poor laptop though, but it has 12.04 and kernel 3.7.4 and xf86-video-savage from latest git. It would be nice if someone affected could help to narrow this down by for instance:
- start off with 12.04
- install a newer kernel (from 12.10 or newer) to see if it that breaks
- install a newer savage driver (from 12.10, maybe needs rebuilding from source package, or from git)

Revision history for this message
Michael Müller (michael-mueller-3) wrote :

- Starting from 12.04. Installing 3.5 kernel with "sudo apt-get install linux-generic-lts-quantal". Graphics works fine.
- Installing new xorg server with "sudo apt-get install xserver-xorg-lts-quantal": Graphics distorted
- Creating xorg.conf with Option "DisableTile": Graphics works again

So it is not the kernel. If you want me to do more tests, I'll be happy to help

Revision history for this message
Tormod Volden (tormodvolden) wrote :

Michael, thanks for the testing. This is very helpful. Can you please also attach your Xorg.0.log?

I can not promise to get to this the next few weeks, but the first thing I will do is to build a debug build of the savage driver (add -DTRACEON to CFLAGS in debian/rules) both before and after upgrading to xserver-xorg-lts-quantal and see if something obvious is different in the logs. If anyone else want to beat me to this, please do :)

affects: linux (Ubuntu) → xserver-xorg-video-savage (Ubuntu)
Revision history for this message
Tormod Volden (tormodvolden) wrote :

BTW, did anyone try Option "AccelMethod" "EXA"?

Revision history for this message
baltasarq (baltasarq) wrote :

Hi, there!

Laptop ACER Aspire 1300 with Savage chipset, Lubuntu 12.04.

I have NOT changed my kernel, just updated it to the last version in the standard channels (i.e., just sudo apt-get update && sudo apt-get upgrade).

I've changed my xorg.conf file to:

Section "Device"
Identifier "device0"
Option "DisableTile"
Option "AccelMethod" "EXA"
EndSection

Now the xorg.conf does not mention the VESA driver, while the laptop does not show problems of any kind.
The problem is solved for me, this way, unless I am missing something big. Am I?

There are some glitches, however. Input text entries (for example, at login) are surrounded by a thick green border (?), while they are still functional, and the lxconsole program shows some artifacts while moving the window... these artifacts disappear when you press enter.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

baltasarq, I meant EXA without DisableTile. We know already that the latter works.

Anyway, I got the chance to look at this and was able to reproduce using the lts-quantal stack. Trying EXA just after XAA had messed up caused a total lock-up, but after reboot EXA seems to work, although things look slow.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

I talked too soon. EXA wasn't really running because the savage kernel module had not been loaded (new kernel after I rebooted, since the xorg-lts upgrades the kernel as well) for some reason. After loading the kernel module and restarting (with EXA) I got a full lock-up again.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

I added a possible fix (taken from upstream 1.14) to the xorg-server package in my PPA:
https://launchpad.net/~tormodvolden/+archive/ppa/+packages?field.series_filter=quantal

Can somebody please test it on 12.10?

The bug is also seen on Gentoo with the 1.13 xserver: http://lists.x.org/archives/xorg/2013-March/055511.html

Revision history for this message
Tormod Volden (tormodvolden) wrote :

22 hours and no response, what a useless testing crowd. I uploaded a patched xorg-server-lts-quantal to my PPA as well and tested it on my Precise setup. It now works better when using EXA (kernel module gets loaded and DRI works). But after a server restart (logout/login) it locks up again. Without EXA (thus "Falling back to shadowfb") it is all messed up.

Revision history for this message
Richard Hollabaugh (haytown) wrote : Re: [Bug 1083032] Re: Video driver not working for Savage chipset

Tormod,

Perhaps you can provide test instructions? I looked at your PPA files.
Do I install all of them? Just some? How do I uninstall after testing?
I guess at this point you have already tested. I am willing to test but
need a little help to get started.

Rich
On 3/7/2013 4:58 PM, Tormod Volden wrote:
> 22 hours and no response, what a useless testing crowd. I uploaded a
> patched xorg-server-lts-quantal to my PPA as well and tested it on my
> Precise setup. It now works better when using EXA (kernel module gets
> loaded and DRI works). But after a server restart (logout/login) it
> locks up again. Without EXA (thus "Falling back to shadowfb") it is all
> messed up.
>

Revision history for this message
Rubin (rubin-xs4all) wrote :

Hey Tormod,

I'm setting up an old Acer Aspire 1300 with the S3 Savage chip in it. I added your PPA to my sources.list to test and installed your version of the Xorg server + drivers. I rebooted single user, ran Xorg -configure and I edited xorg.conf to only contain an entry for the savage driver so as to avoid accidental "it works!" while you're unconsciously running vesa or something else.

I specified only Option "AccelMethod" "EXA" and not DisableTile. I can confirm the graphics work. Without an Xorg.conf, graphics won't work and are gibberish. I'm not familiar with these cards experience wise, but it seems slow, draging windows is jerky and scrolling too.

If you want me to check a few things, let me know!

Revision history for this message
pradyot (pradyotsn) wrote :

hey everyone,

I am a newbie at this. I have

 Processor: Intel(R) Pentium(R) 4 CPU 1.60GHz
Memory: 512MB RAM
Hard Disk: 160 GB

Display Devices
---------------
Card name: S3 Graphics Inc. Savage4
Manufacturer: S3 Graphics Inc.
Chip type: S3 Savage4

I am getting this weird coloured bars.

http://ubuntuforums.org/showthread.php?t=2128715

Revision history for this message
In , Musikolo (musikolo) wrote :

Hi guys,

I'm writing to you because I have found the latest version of savage driver is not working. The problem is caused because the xxa module is not found and, in turn, DRI support cannot be enabled. In the meantime, I have had no other choice than using vesa driver until the issue is sorted out. I'm using ArchLinux and the maintainer of the package of this driver has told me to contact you to find a solution.

Here you can find:
- The content of my Xorg.0.log file: http://pastebin.com/qdp7datT
- The thread open to discuss the problem in ArchLinux https://bugs.archlinux.org/task/34527
- And the code to build the package:https://projects.archlinux.org/svntogit/packages.git/tree/trunk?h=packages/xf86-video-savage

Any help to solve this issue would be really appreciated.

Best regards,
Carlos.

Revision history for this message
In , Bugzi11-fdo-tormod (bugzi11-fdo-tormod) wrote :

Thanks for your report. This broke in xserver 1.13 and was partly fixed in 1.14. Try Option "AccelMethod" "EXA". A workaround is to turn off acceleration with Option "DisableTile".

See also https://bugs.launchpad.net/bugs/1083032

Changed in xorg-server:
importance: Unknown → Medium
status: Unknown → Confirmed
Revision history for this message
In , Musikolo (musikolo) wrote :

Hi Tormod,

Thanks for your prompt response and assistance.

I have tried the suggested options, but I'm still having the same issue, no change at all.

- http://pastebin.com/yAgshj2V

Is there any other work around? When is it expected to have a fix for it?

Regards.

Revision history for this message
In , Bugzi11-fdo-tormod (bugzi11-fdo-tormod) wrote :

Simply using EXA (and a 1.14 fix backported to 1.13) worked for me. Maybe more is broken in 1.14. What exactly happens in your case? Please be more specific than "does not work".

There is no schedule for such bug fixes, most developers do not care about hardware which is more than a few years old. I will look closer into it sometime in the next weeks, although no guarantees given. There seems to be more continued use of the savage driver than I thought :)

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 77694
My scrambled screen

Revision history for this message
In , Musikolo (musikolo) wrote :

Hi Tormod,

Thanks once again for your prompt response.

When say "it doesn't work", I mean the screen is scrambled and nothing is distinguishable. As the telling goes, a picture is worth more than a thousand words, so please, have a look at the attached picture of the my screen.

Kindly, let me know if there is any patch that can be used to fix this issue.

Best regards,
Carlos.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

For Quantal please try the packages from my PPA https://launchpad.net/~tormodvolden/+archive/ppa/+packages?field.series_filter=quantal

You'll need both the new xorg-server 2:1.13.0-0ubuntu6.3~bug1083022 and savage 1:2.3.6-0ubuntu1.1~tormod~q packages. You may add the PPA for simplicity, there are only bug fixes for Quantal anyway. The packages are queued for building while I write this, so please wait until they are built before trying (estimated 2 hours from now).

No xorg.conf should be needed.

Revision history for this message
In , Bugzi11-fdo-tormod (bugzi11-fdo-tormod) wrote :

I have committed a couple of fixes to the savage driver repository. One problem was that the driver, even when built for an xserver without XAA support, would internally default to XAA instead of EXA nonetheless. Secondly there was an EXA function that needed an update for recent xservers.

This works well for me. The only issue I see now is that console switching does not work, the text consoles are unusable. But I am not sure this is savage-specific since for instance the vesa driver locks up the machine completely if I try to switch consoles.

Revision history for this message
Richard Hollabaugh (haytown) wrote :

I added your PPA to my system and downloaded the changes. Renamed the xorg.conf file and rebooted. It Works! Thanks!

Revision history for this message
In , Musikolo (musikolo) wrote :

Hi Tormod,

Awesome mate!

What version will this fix be included in?

Regarding the console issue, I think this problem is being there for a long time. It would be nice to find a fix it too, but it's not a critical issue at all.

Thanks a lot for your effort, I appreciate it a lot.

Best regards,
Carlos.

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

For those of you who are already on Raring (Ubuntu 13.04), I have uploaded PPA packages (builds in 6 hours from now). To get this fixed in the official Ubuntu repositories, it is a longer journey. First it must be fixed in current development version Saucy.

Attached is a debdiff for Saucy (xorg-server).

Revision history for this message
Tormod Volden (tormodvolden) wrote :
Revision history for this message
Ubuntu Foundations Team Bug Bot (crichton) wrote :

The attachment "debdiff for Saucy (xorg-server)" seems to be a debdiff. The ubuntu-sponsors team has been subscribed to the bug report so that they can review and hopefully sponsor the debdiff. If the attachment isn't a patch, please remove the "patch" flag from the attachment, remove the "patch" tag, and if you are member of the ~ubuntu-sponsors, unsubscribe the team.

[This is an automated message performed by a Launchpad user owned by ~brian-murray, for any issue please contact him.]

tags: added: patch
Revision history for this message
Tormod Volden (tormodvolden) wrote :

Here are SRU patches for Raring, once we get that far...

Revision history for this message
Tormod Volden (tormodvolden) wrote :
Revision history for this message
Maarten Lankhorst (mlankhorst) wrote :

Could you create a new upstream release for savage so I can pull this to saucy?

http://www.x.org/wiki/Development/Documentation/ReleaseHOWTO

Revision history for this message
Launchpad Janitor (janitor) wrote :

This bug was fixed in the package xorg-server - 2:1.13.3-0ubuntu9

---------------
xorg-server (2:1.13.3-0ubuntu9) saucy; urgency=low

  * Cherry-pick patch from 1.14 to fix DRI1 regression. (LP: #1083032)
    - dri1-fix-dri1-startup-since-459c6da0f907ba33d733c7e6.patch
 -- Tormod Volden <email address hidden> Sun, 12 May 2013 11:30:46 +0200

Changed in xorg-server (Ubuntu):
status: New → Fix Released
Revision history for this message
Tormod Volden (tormodvolden) wrote :

> Could you create a new upstream release for savage so I can pull this to saucy?

Maarten, I will also try to sort out some other issues before I make a new upstream release. Distros can make git snapshots at any point. I would like you to commit these patches to Saucy because I want to have them SRU'ed. I can imagine that SRU'ing a new version is almost impossible, for bureaucratic reason (I mean, never mind that the current driver is pretty much useless).

Revision history for this message
Tormod Volden (tormodvolden) wrote :

Maarten, by the way, I do intend to package a snapshot for Debian soon, so you can just sync it to Saucy later (I don't think the single Ubuntu patch for PCI instead of AGP is needed any longer). But before that, it would simplify SRU if the above patch has been wetted in Saucy.

Revision history for this message
In , Musikolo (musikolo) wrote :

Hi Tormod,

I have tried a new build of the savage package supposedly including your changes (https://bugs.archlinux.org/task/34527), but it doesn't solve my problem. I'm still getting my screen scrambled, just multi-color lines everywhere. I have tried a configuration forcing to use EXA module as well as a brand new xorg.conf file generated through X -configure command.

Please, let me know if there anything I can you to help you find the problem.

Best regards,
Carlos.

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

This bug was fixed in xserver-xorg-video-savage 1:2.3.6-0ubuntu2:

---------------
xserver-xorg-video-savage (1:2.3.6-0ubuntu2) saucy; urgency=low

  * Fix EXA and use it by default (LP: 1083032)
   - 110-exa-Use-exaGetPixmapFirstPixel-instead-of-devPrivate.patch
   - 120-savage-fix-default-accel-arch-when-built-with-no-XAA.patch
   - 130-Really-fix-up-default-accel-arch-when-built-with-no-.patch
 -- Tormod Volden <email address hidden> Tue, 07 May 2013 22:47:50 +0200

(not sure why the janitor did not send this out, missing "#"?)

Changed in xserver-xorg-video-savage (Ubuntu):
status: Confirmed → Fix Released
Revision history for this message
In , Bugzi11-fdo-tormod (bugzi11-fdo-tormod) wrote :

Musikolo, please attach (no pastebin) your new Xorg.0.log.

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80797
xorg.conf.new

File generated using X -configure command

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80798
xorg.conf_S3

Version I used to have working before the driver broke

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80799
xorg.conf_VESA

Only configuration currently working based on vesa driver.

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80818
Xorg.0.log_NEW

Log for xorg.conf.new configuration without patch

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80819
Xorg.0.log_NEW_WITH_PATCH

Log for xorg.conf.new configuration with patch

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80820
Xorg.0.log_S3

Log for xorg.conf_S3 configuration without patch

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80821
Xorg.0.log_S3_WITH_PATCH

Log for xorg.conf_S3 configuration with patch

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80822
Xorg.0.log_VESA

Log for xorg.conf_VESA configuration without patch

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80823
Xorg.0.log_VESA_WITH_PATCH

Log for xorg.conf_VESA configuration with patch

Revision history for this message
In , Bugzi11-fdo-tormod (bugzi11-fdo-tormod) wrote :

Have you tried without any xorg.conf? You have two device sections in your configuration, I don't understand why. The last working configuration from comment 11 simply switches to EXA, which is the default now.

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80873
Xorg.0.log_NO_CONF

Log having no xorg.conf configuration without patch

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80874
Xorg.0.log_NO_CONF_WITH_PATCH

Log having no xorg.conf configuration with patch

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80876
xorg.conf_S3_1_DEVICE

Same configuration as xorg.conf_S3 but having just 1 device

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80877
Xorg.0.log_S3_1_DEVICE

Log for xorg.conf.S3_1_DEVICE configuration without patch

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80878
Xorg.0.log_S3_1_DEVICE_WITH_PATCH

Log for xorg.conf.S3_1_DEVICE configuration with patch

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80884
lspci output (lspci -vv -d 5333 -nn)

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80885
lsmod output

Revision history for this message
In , Musikolo (musikolo) wrote :

Created attachment 80886
lsmod output

description: updated
description: updated
Revision history for this message
Brian Murray (brian-murray) wrote :

I've accepted this upload to raring-proposed although the bug number in the changelog was improperly formatted. It should be of the the form (LP: #1083032). Subsequently, once the SRU verification process has completed some one will manually need to let a member of the Ubuntu Stable Release updates team know so that they can release the package to -updates.

Changed in xserver-xorg-video-savage (Ubuntu Raring):
status: New → Fix Committed
importance: Undecided → Medium
tags: added: verification-needed
Revision history for this message
Launchpad Janitor (janitor) wrote :

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in xorg-server (Ubuntu Raring):
status: New → Confirmed
Revision history for this message
qman (cs-cracker) wrote :

Installing xserver-xorg-video-savage 1:2.3.6-0ubuntu1.1 from raring-proposed resolves this problem for me on a Thinkpad T23.

Revision history for this message
In , Bugzi11-fdo-tormod (bugzi11-fdo-tormod) wrote :

Can you please try savage driver version 2.3.7? It looks like it is already in Arch Linux.

Revision history for this message
In , Musikolo (musikolo) wrote :

Hi Tormod,

Sure, I will do it, but I'm not able to right now. Unfortunately, I'm traveling for business and we'll not go back home until beginning of November where my personal laptop is. I know it's quite a lot of time, but take for granted that I'll let you know as soon as I test it out.

Thanks a lot for your amazing effort and work.

Best regards,
Carlos.

Revision history for this message
Tormod Volden (tormodvolden) wrote :

Can anyone (other than qman) confirm that the raring-proposed package fixes the issue?

Revision history for this message
In , Musikolo (musikolo) wrote :

Hi Tormod,

I'm happy to tell you that the problem is now fixed. I'm writing from the affected computer. The latest version (xf86-video-savage-2.3.7-1 + savage-dri-7.11.2-5) has fully sort it out now. So, I owe you a million thanks:

1000000 x Thanks!

Let me tell you that I appreciate the effort and time you've spent in the resolution of this issue.

Best regards,
Carlos

Changed in xorg-server:
status: Confirmed → Fix Released
Revision history for this message
Steve Langasek (vorlon) wrote :

It seems that this package was accepted into the -proposed queue with a syntax error in the bug reference, so it's not quite on the SRU radar. Can anyone verify that the SRUed package works?

Revision history for this message
Rolf Leggewie (r0lf) wrote :

raring has seen the end of its life and is no longer receiving any updates. Marking the raring task for this ticket as "Won't Fix".

Changed in xserver-xorg-video-savage (Ubuntu Raring):
status: Fix Committed → Won't Fix
Rolf Leggewie (r0lf)
Changed in xorg-server (Ubuntu Raring):
status: Confirmed → Won't Fix
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.