Slow performance and tiling issues on i915

Bug #349314 reported by Eric Appleman
502
This bug affects 46 people
Affects Status Importance Assigned to Milestone
Linux
Fix Released
Medium
linux (Ubuntu)
Fix Released
Low
Andy Whitcroft
Jaunty
Fix Released
Low
Andy Whitcroft

Bug Description

This problem is particularly noticeable with the Ubuntu Netbook Remix launcher, on systems such as the eeePC 900.

http://lists.freedesktop.org/archives/intel-gfx/2009-January/001187.html

===

SRU Justification Jaunty

Impact: Video performance of affected systems is very poor, watching video is not possible

Fix Description: detect and enable previously disabled tiling support IFF not already enabled by the BIOS

Patch:
    http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-jaunty.git;a=commit;h=de385c9db845645e6cc40a5355b6779e044d6afb
    http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-jaunty.git;a=commit;h=3533af638b8e149568fd3d6cdbbb4a384674f170

Risks: MCHBAR is disabled for good reason on some unknown platform, given this is a chipset level feature this is very unlikely

TEST CASE: see bug

Revision history for this message
Eric Appleman (erappleman) wrote :
description: updated
Changed in linux:
status: Unknown → Confirmed
Changed in linux:
status: New → Fix Committed
Revision history for this message
Andy Whitcroft (apw) wrote :

As this is a performance not a crashing issue, marking it low priority.

Changed in linux:
status: Fix Committed → In Progress
importance: Undecided → Low
Revision history for this message
Andy Whitcroft (apw) wrote :

@Starks -- i have built some test kernels with this patch applied for you to test. Could you try these out and let me know if they show any problems, and the subjective performance thereof. Please report back here. Kernels are at the URL below:

    http://people.ubuntu.com/~apw/lp349314-jaunty/

Changed in linux:
status: In Progress → Incomplete
Revision history for this message
Neil J. Patel (njpatel) wrote :

This also affects UNR Jaunty, as the launcher is based on GL. The performance is really bad on the eeePC (where it was very good in Intrepid).

Loïc Minier (lool)
Changed in linux:
milestone: none → ubuntu-9.04
Revision history for this message
Eric Appleman (erappleman) wrote :

I had rolled my own kernel last night with the patch applied and didn't notice any improvement until I disabled tiling.

Hopefully Andy's kernel will fare better so that I don't need to disable tiling.

Revision history for this message
Eric Appleman (erappleman) wrote :

No go.

[ 2.393704] (EE) intel(0): Failed to set tiling on front buffer: rejected by kernel

Revision history for this message
Loïc Minier (lool) wrote :

The proposed kernel .deb work great for me on EeePC 701SD and solve the slowness! Thanks

Changed in linux:
status: Incomplete → Triaged
Revision history for this message
Eric Appleman (erappleman) wrote :

I'm running a Dell Inspiron 640m.

00:02.0 VGA compatible controller [0300]: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller [8086:27a2] (rev 03)

No improvement unless tiling is disabled.

1 comments hidden view all 171 comments
Revision history for this message
Eric Appleman (erappleman) wrote :

[ 2.393704] (EE) intel(0): Failed to set tiling on front buffer: rejected by kernel

Tiling is still erroring in the logs, but the performance is back to normal, even without the patched kernel.

Did something change in the PPA intel driver?

Revision history for this message
Eric Appleman (erappleman) wrote :

*somewhat back to normal, but still less than Intrepid

Revision history for this message
Eric Appleman (erappleman) wrote :

Once again, I need to correct myself. The performance gains were from UXA being enabled.

Revision history for this message
Andy Whitcroft (apw) wrote :

@Loic -- it seems others are not seeing tiling turned on. Could you attach an Xorg.0.log so we can confirm you are getting tiling. Also can you confirm your exact intel chipset you have (may be in the same log.

Changed in linux:
status: Triaged → In Progress
Tim Gardner (timg-tpi)
Changed in linux (Ubuntu Jaunty):
assignee: nobody → apw
Revision history for this message
Robbie Williamson (robbiew) wrote :

I can confirm that Andy's kernel worked for me on my eeePC 900.

Revision history for this message
Loïc Minier (lool) wrote :

Tiling is enabled, no error in Xorg.log

lspci reports this adapter:
00:02.0 VGA compatible controller: Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller (rev 04)

Revision history for this message
Robbie Williamson (robbiew) wrote :

lspci for my eeePC 900 shows the same:
00:02.0 VGA compatible controller: Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller (rev 04)

Revision history for this message
allisterbrizan (allisterbrizan) wrote :

Andy's debs [ http://people.ubuntu.com/~apw/lp349314-jaunty/ ] do the trick on my eeepc also but I still think there is value in the option to disable animations and mouseover effects

Revision history for this message
Chris Quach (quach-c) wrote :

Have the same error in Xorg.0.log, but I have a Intel 945GM.. I'm using the Jaunty beta and applied all updates which came through the update-manager but still very slow desktop effects... Do you think the patched kernel will help on my system?

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

from the compiz gconf configuration:

    sync_to_vblank = true

disable that and try again. The default is false.

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

duh, I commented on the wrong bug, sorry for the noise :)

Revision history for this message
Nafallo Bjälevik (nafallo) wrote :

This kernel is total win on my EeePC 701 running UNR. It makes things usable again. I here by give my blessings.

Revision history for this message
Sean Sosik-Hamor (sciri) wrote :

I tried Andy's 2.6.28-11.38lp349314apw1 kernel on my Mac mini Core Duo with Intel GMA950 on mdz's recommendation after following Bug #339555 and Bug #353245 but have inconclusive results.

After my initial Intrepid to Jaunty upgrade video performance was basically unusable even with all Visual Effects disabled. Just moving windows around the screen resulted in choppy refresh, and switching workspaces took 2-3 seconds for all the windows to redraw.

I eventually tested using a Jaunty Desktop Beta i386 LiveCD and performance was normal; video was just as fast, crisp, and smooth as I was used to with Intrepid.

I then reboot back to the internal disk and video performance was mysteriously better. Still not on par with Intrepid but usable. No configuration changes or package upgrades were made.

A dist-upgrade was done to pick up the latest packages but still no additional improvement.

I then installed Andy's 2.6.28-11.38lp349314apw1 kernel and performance seems better compared to my original unusable Intrepid to Jaunty upgrade, but performance doesn't come close to my test on the Jaunty Desktop Beta i386 LiveCD.

Hence, my inconclusive results.

My unscientific test is simple...create 4 workspaces (2 rows, 2 columns), open a Firefox window on each workspace, then rapidly switch between workspaces with Ctrl-Alt-Arrow. On Intrepid and the Jaunty Desktop Beta i386 LiveCD switching workspaces is smooth and fluid. On my upgraded Jaunty install it sometimes takes 2-3 seconds to switch workspaces as the windows choppily animate across the screen.

00:00.0 Host bridge: Intel Corporation Mobile 945GM/PM/GMS, 943/940GML and 945GT Express Memory Controller Hub (rev 03)
00:02.0 VGA compatible controller: Intel Corporation Mobile 945GM/GMS, 943/940GML Express Integrated Graphics Controller (rev 03)

(II) intel(0): Attempting memory allocation with tiled buffers.
(EE) intel(0): Failed to set tiling on front buffer: rejected by kernel
(EE) intel(0): Failed to set tiling on back buffer: rejected by kernel
(EE) intel(0): Failed to set tiling on depth buffer: rejected by kernel
(II) intel(0): Tiled allocation successful.

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

This bug was fixed in the package linux - 2.6.28-11.40

---------------
linux (2.6.28-11.40) jaunty; urgency=low

  [ Amit Kucheria ]

  * Disable DEVKMEM for all archs on Jaunty
    - LP: #354221

  [ Andy Whitcroft ]

  * SAUCE: md: wait for possible pending deletes after stopping an array
    - LP: #334994

  [ Brad Figg ]

  * ARM: Setting the bootloader for imx51 flavour.
    - LP: #348382
  * ARM: Add bootloader package Recomendation to iop32x and ixp4xx flavours
    - LP: #348382

  [ Tim Gardner ]

  * SAUCE: [i915] allocate MCHBAR space & enable if necessary
    - LP: #349314

  [ Upstream Kernel Changes ]

  * hpilo: open/close fix
    - LP: #353496

 -- Amit Kucheria <email address hidden> Thu, 02 Apr 2009 11:26:22 -0400

Changed in linux (Ubuntu Jaunty):
status: In Progress → Fix Released
Revision history for this message
Leo Milano (lmilano) wrote :

Has this new kernel already been uploaded? I would like to test the nightly UNR image to see it the jerky display bug is fixed in the eeepc 701. Thanks

Revision history for this message
Leo Milano (lmilano) wrote :

FIXED!. The newest UNR image works smoothly on my EEE PC 701. Thanks for the fantastic work! Has this been fixed in the kernel upstream?

Revision history for this message
Robbie Williamson (robbiew) wrote :

I also confirm the -40 kernel fixes this on my Eee PC 900.

Revision history for this message
CassieMoondust (cassie-lx) wrote :

I confirm the 40 kernel fixes the error on my eeePC 701 4G. Have now smooth animations wirh the netbook-launcher!
Great, thanks!

Revision history for this message
Matt Zimmerman (mdz) wrote :

According to the kernel changelog, this patch has been reverted:

linux (2.6.28-11.41) jaunty; urgency=low
[...]
  * Revert "SAUCE: [i915] allocate MCHBAR space & enable if necessary"
    Appears to cause hard locks in some cases.
    - LP: #349314

Changed in linux (Ubuntu Jaunty):
status: Fix Released → Triaged
Revision history for this message
Scott Evans (vk7hse) wrote : Re: [Bug 349314] Re: [i915] allocate MCHBAR space & enable if necessary

Looks like I need to pin to #40 as #41 is useless here! this has
regressed to what it was like prior to this current patch :(

On Wed, 2009-04-08 at 13:05 +0000, Matt Zimmerman wrote:

> According to the kernel changelog, this patch has been reverted:
>
> linux (2.6.28-11.41) jaunty; urgency=low
> [...]
> * Revert "SAUCE: [i915] allocate MCHBAR space & enable if necessary"
> Appears to cause hard locks in some cases.
> - LP: #349314
>
>
> ** Changed in: linux (Ubuntu Jaunty)
> Status: Fix Released => Triaged
>

Scott Evans VK7HSE
Phone: +61362291658
Mobile: +61417586157
Skype: vk7hse
<email address hidden>
http://www.vk7hse.hobby-site.org
PGP/GPG Key ID 74639624

--
This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed.
If you have received this email in error please notify the system manager. This message contains confidential information and is intended only for the
individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail. Please notify the sender immediately
by e-mail if you have received this e-mail by mistake and delete this e-mail from your system. If you are not the intended recipient you are notified
that disclosing, copying, distributing or taking any action in reliance on the contents of this information is strictly prohibited.

Revision history for this message
Robbie Williamson (robbiew) wrote : Re: [i915] allocate MCHBAR space & enable if necessary

I hope this was not reverted for bug 356951, as the reporter was on #38 when he saw the problem.

Revision history for this message
Cesar Eduardo Barros (cesarb) wrote :

No, I think it was reverted because of http://bugs.freedesktop.org/show_bug.cgi?id=20520 (mentioned on http://bugzilla.kernel.org/show_bug.cgi?id=12950, which is linked to by this bug).

For me it was also a performance regression, so I also reverted to .40 and put a hold on the package.

Revision history for this message
Jason Anton (rev-jt-anton) wrote :

The .41 was a regression on my 900 as well. The .40 worked so well! Now I can't find it!...argh!

Revision history for this message
Matt Zimmerman (mdz) wrote : Re: [Bug 349314] Re: [i915] allocate MCHBAR space & enable if necessary

On Thu, Apr 09, 2009 at 03:09:45PM -0000, Jason Anton wrote:
> The .41 was a regression on my 900 as well. The .40 worked so well!
> Now I can't find it!...argh!

https://edge.launchpad.net/ubuntu/+source/linux/2.6.28-11.40 (scroll down to
the list of builds and select i386 or amd64).

--
 - mdz

Revision history for this message
CassieMoondust (cassie-lx) wrote : Re: [i915] allocate MCHBAR space & enable if necessary

Back to the darkness, the .41 kernel breaks my 701 and the netbook-launcher is useless for me again. This is a showstopper for the jaunty netbook remix release. Looking forward for a fix soon...

Revision history for this message
Paul Larson (pwlars) wrote :

Based on comment 29 in http://bugs.freedesktop.org/show_bug.cgi?id=20520, it looks like the problems seen were probably unrelated to this patch.

Revision history for this message
Andy Whitcroft (apw) wrote :

Backing out that patch definitivly fixed the affected users. It is suspected that there is a general tiling bug which is tripped by enabling this on those platforms. The patch in question has since been recinded and replaced by a different interim patch and work is still continuing to solve the underling gem issues.

Revision history for this message
Matt Zimmerman (mdz) wrote : Re: [Bug 349314] Re: [i915] allocate MCHBAR space & enable if necessary

On Thu, Apr 09, 2009 at 10:47:58PM -0000, Andy Whitcroft wrote:
> Backing out that patch definitivly fixed the affected users. It is
> suspected that there is a general tiling bug which is tripped by
> enabling this on those platforms. The patch in question has since been
> recinded and replaced by a different interim patch and work is still
> continuing to solve the underling gem issues.

What is the bug number for the regression caused by this patch?

--
 - mdz

Revision history for this message
Ulrik Mikaelsson (rawler) wrote : Re: [i915] allocate MCHBAR space & enable if necessary

.40 did not resolve the problem for my Intel Mac Mini.

In dmesg, I read the following when starting X.
[ 220.727730] [drm:i915_setparam] *ERROR* unknown parameter 4
[ 220.727772] [drm:i915_getparam] *ERROR* Unknown parameter 6
[ 221.627995] [drm:i915_getparam] *ERROR* Unknown parameter 6

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

Ulrik, those errors are actually completely innocuous and can be ignored.

Revision history for this message
Daniel Veillette (drv) wrote :

Confirming that .41 was a major regression performance-wise on my EEE 900. Sticking with .40 until fixed.

Steve Langasek (vorlon)
Changed in linux (Ubuntu Jaunty):
milestone: ubuntu-9.04 → jaunty-updates
Matt Zimmerman (mdz)
summary: - [i915] allocate MCHBAR space & enable if necessary
+ Slow performance and tiling issues on i915
description: updated
Andy Whitcroft (apw)
Changed in linux (Ubuntu):
status: Triaged → In Progress
Andy Whitcroft (apw)
description: updated
Andy Whitcroft (apw)
Changed in linux (Ubuntu):
status: In Progress → Fix Committed
status: Fix Committed → In Progress
Changed in linux (Ubuntu Jaunty):
status: Triaged → Fix Committed
Martin Pitt (pitti)
Changed in linux (Ubuntu):
milestone: jaunty-updates → none
91 comments hidden view all 171 comments
Revision history for this message
mobile (carlleach) wrote :

Adding Option "Tiling" "False" to xorg.conf works for me too on Asus EeePC 701 with 2.6.28-13.44 kernel.

Revision history for this message
Knef (knef) wrote :

Setting "Tiling" to "False" in xorg.conf didn't work for me (I have an EeePC 900).

Revision history for this message
Tommy Trussell (tommy-trussell) wrote :

There has been yet another kernel update (second update to kernel 2.6.28-13-generic) but this patch still isn't included. Now I'm a bit confused whether there was a regression with the patch -- didn't someone decide the regression was not due to this patch?

The NBR performance is much better when using the patched 2.6.28-11 kernel on this eeepc 900. HOWEVER, now the unofficial patched 2.6.28-11-generic kernel linked in this bug report is missing out on several Ubuntu kernel updates.

Does the patch apply cleanly to the 2.6.28-13 kernel?

Revision history for this message
Gabriel C (jgcastell) wrote :

I can confirm too that using Setting "Tiling" to "False" in xorg.conf didn't work in my EeePC 900

Revision history for this message
Martin Pitt (pitti) wrote :

Accepted linux into jaunty-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

tags: added: verification-needed
Revision history for this message
Jordi Guillaumes (jguillaumes) wrote : Re: [Bug 349314] Re: Slow performance and tiling issues on i915

2009/7/8 Martin Pitt <email address hidden>:
> Accepted linux into jaunty-proposed, the package will build now and be
> available in a few hours. Please test and give feedback here. See

Hi there. Forgive me if I am asking an obvious question. Could you
tell me which is the name of the package which is going to fix the
issue? I've just taken a look into jaunty-proposed and i've just found
a package which I think could be relevant to this issue (Xorg sever
for intel chipsets). I've installed it and it still does not work, so
I'm back with the "unofficial" 2.6.28-11 patched kernel...

Could be posible that the updated kernel is not yet in the proposed repository?

Thanks in advance.

Revision history for this message
Martin Pitt (pitti) wrote :

Jordi Guillaumes [2009-07-08 11:09 -0000]:
> Hi there. Forgive me if I am asking an obvious question. Could you
> tell me which is the name of the package which is going to fix the
> issue?

The kernel took a while to build and needed archive admin processing
(NEW). I did that now. linux-image-2.6.28-14-generic should be
available for testing in about two hours on archive.ubuntu.com.

Revision history for this message
Andrew Manson (real_ate) (real-ate) wrote :

2 hours have passed... and in fact 2 days have passed but still no kernel available. I have tried enabling the proposed repo and also searching packages.ubuntu.com and no luck.

Any update? a link to the package on packages.ubuntu.com would be very appreciated once it comes available.

Revision history for this message
Tommy Trussell (tommy-trussell) wrote :

I have installed the 2.6.28-14-generic kernel, and so far it is working very well on my ASUS Eee PC 900. SO if anyone is needing confirmation that the kernel works and it fixes the graphics slowdowns in the NBR launcher, YES! It works!

I did have some difficulties installing the package (as Andrew Manson expressed above) but I finally was able to get the package installed. My apologies to everyone who already knows this but it had me stumped for days. SO I'm writing this out hoping it helps someone else.

1) Activate the jaunty-proposed archive as described in https://wiki.ubuntu.com/Testing/EnableProposed -- basically add the following line to the end of /etc/apt/sources.list --

deb http://archive.ubuntu.com/ubuntu/ jaunty-proposed restricted main multiverse universe

2) Update your local package cache using your favorite method (such as "sudo aptitude update").

Note: No matter what I did I did NEVER saw the new kernel proposed as an update to the existing kernel, whether using Synaptic, Aptitude or the "apt-cache policy" command. The only way I could even SEE that the new kernel was available was by searching for it. For example, the terminal command "aptitude search linux-image-generic" showed it was available.

3) Install the kernel explicitly. The easiest method is to use this terminal command:

    sudo apt-get install linux-image-2.6.28-14-generic

4) After installing the kernel, you can undo step 1 -- disable the jaunty-proposed archive by commenting out the line you added in /etc/apt/sources.list --

# deb http://archive.ubuntu.com/ubuntu/ jaunty-proposed restricted main multiverse universe

4) When you reboot, press the Esc key to display the grub boot menu and make sure the new 2.6.28-14 kernel will be the version that gets booted.

Note: Since I have been using experimental kernels I use the "default saved" option in grub so it automatically boots from the last kernel selected. For some reason, after I installed the -14 kernel, it defaulted to the -13 kernel rather than using the -14 version I just installed or the -11 kernel I booted before.

I hope that helps someone who was having as much trouble as I was.

Revision history for this message
Scott Brown (scotty-b-brown) wrote :

Thanks Tommy - worked a treat!

Revision history for this message
Tim Clevenger (tjclevenger) wrote :

Same problem on an Asus eeePC 701 4GB. Installed the -14 kernel per Tommy's instructions and it works great. Thanks guys!

Martin Pitt (pitti)
tags: added: verification-done
removed: verification-needed
Revision history for this message
mobile (carlleach) wrote :

Confirmed linux-image-2.6.28-14-generic fixed problem on Asus Eee PC 701SD. Thanks.

Revision history for this message
Jordi Guillaumes (jguillaumes) wrote :

Same thing here. 14-generic fixed the problem on an eeePC 900. Good job!

Revision history for this message
DrewTardif (uplink59) wrote :

I use the eee900 like many and have the same problem, sorry for being stupid, but how EXACTLY do I download and install the patch? I click on the patch and It just brings me to another page.

Revision history for this message
Conan Barnes (barneco2000) wrote :

Another confirmation that 14-generic resolves the issue on my eee 900.

@DrewTardif: No need to download and install a patch. The fix is in the 14 kernel that's available from the proposed archive. Simply enable the proposed archive in your software sources, then utilize the familiar apt-get instruction to retrieve and install 14 and you should be good to go. For a step-by-step script of how to do this, look about 6 posts up and Tommy Trussell has outlined it perfectly.

Revision history for this message
Anton¡o Sch¡fano (skiantoz) wrote :

The 14-generic kernel seems to work also on mi eeepc 1000 hd, i.e. tiling is enabled and performance is better.
One strange thing I am investigating: supertux2 is too slow unless run by root. Any ideas?

Revision history for this message
Zack Evans (zevans23) wrote :

Re the supertux2 problem: Could be related to permissions on /dev/dri/video0, which sometimes has permissions for root and the "video" group only. Have a look at permissions and also try putting your user into the "video" group.

I'm not sure if there's an underlying bug in udev that causes this - seems to me like this ought to be writeable by any user sitting at the console by default...

Revision history for this message
Martin Pitt (pitti) wrote :

Zack Evans [2009-07-20 16:28 -0000]:
> I'm not sure if there's an underlying bug in udev that causes this -
> seems to me like this ought to be writeable by any user sitting at the
> console by default...

It is. In karmic that's done by udev, but in previous releases it was
done by hal's automatic ACL magic.

Revision history for this message
Andy Whitcroft (apw) wrote :

This fix hit mainline in v2.6.31-rc1 and therefore hit Karmic in 2.6.31-1 kernel. I am therefore closing the linux task Fix Released.

Changed in linux (Ubuntu):
status: In Progress → Fix Released
Revision history for this message
Anton¡o Sch¡fano (skiantoz) wrote :

Regarding the supertux2 problem, I think the permissions are ok (see below), as my user is in the group video and the system sets the ACL correctly. Note that the device is /dev/dri/card0 though (not video0).

uid=1000(antadm) gid=1000(antadm) groups=4(adm),20(dialout),24(cdrom),29(audio),30(dip),44(video),46(plugdev),104(fuse),106(lpadmin),112(netdev),115(pulse-access),116(pulse-rt),121(admin),122(sambashare),1000(antadm),1001(eeeuser)

antadm@kicker:~$ getfacl /dev/dri/card0
getfacl: Removing leading '/' from absolute path names
# file: dev/dri/card0
# owner: root
# group: video
user::rw-
user:antadm:rw-
group::rw-
mask::rw-
other::---

Revision history for this message
Steve Beattie (sbeattie) wrote :

Antonio: do you see the supertux2 issue with only the -14 kernel, or does it occur with the -13 kernel as well? WewWould like to get some indication whether this is an issue strictly introduced by the kernel in -proposed. Thanks for testing!

Revision history for this message
Steve Beattie (sbeattie) wrote :

Antonio: do you see the supertux2 issue with only the -14 kernel, or does it occur with the -13 kernel as well? We would like to get some indication whether this is an issue strictly introduced by the kernel in -proposed. Thanks for testing!

Revision history for this message
Anton¡o Sch¡fano (skiantoz) wrote :

Steve Beattie wrote:
> Antonio: do you see the supertux2 issue with only the -14 kernel, or
> does it occur with the -13 kernel as well? We would like to get some
> indication whether this is an issue strictly introduced by the kernel in
> -proposed. Thanks for testing!
>
>
With the -13 kernel, the performance is poor for every user, root included.
Kernel -11 lp349314apw5 behaves like the -14: slow performance, except
if run as root.

Revision history for this message
uncajesse (uncajesse) wrote :

I have a 900 and whatever change with the packages (i use release and proposed) that just happened today ended up making this bug come back, especially with the Netbook Remix launcher.

Revision history for this message
Bruno Girin (brunogirin) wrote :

@uncajesse: can you open a terminal and post the output of the command "uname -r" please? If you are using the correct kernel, the output should be "2.6.28-14-generic". If not, we'll have to find out why you're not using that version.

Revision history for this message
Chris N (slothdog) wrote :

Just downloaded and installed the latest kernel update, and am still seeing the slowdowns. Tested on an Eee PC 900 using the World of Goo demo.

uname-r shows:
2.6.28-14-generic

Revision history for this message
Bruno Girin (brunogirin) wrote :

OK, I can see what happened now. I had the 2.6.28-14 kernel from the jaunty-proposed repository and it worked fine (including World of Goo). Then I fired up Update Manager. As part of the update was a *-14 kernel from the jaunty-updates repository. That *-14 kernel is not the same as the *-14 kernel in jaunty-proposed and does not include the patch, thus reverting to the original problem.

Can we do a *-15 kernel with that patch in and make sure it gets to jaunty-updates one way or another and doesn't get clobbered again please?

So, to reproduce the issue that uncajesse and Chris N see:
1. make sure you have the -14 kernel from jaunty-proposed installed
2. remove jaunty-proposed from the sources
3. fire up Update Manager
4. update, including kernel -14

Expected result: no change, all works well

Actual result: the -14 kernel in jaunty-updates does NOT include the patch that was in the -14 kernel in jaunty-proposed and the bug re-appears.

Revision history for this message
uncajesse (uncajesse) wrote :

2.6.28-14-generic
here too. And yes, I forced the version from jaunty-proposed and it doesn't have the bug.

Revision history for this message
Scott Brown (scotty-b-brown) wrote :

I can confirm that the mainstream 2.6.28-14 kernel still has the bug in Ubuntu NBR on my 701.

How do i force a reinstall of the packages from jaunty-proposed???

Revision history for this message
Martin Pitt (pitti) wrote :

Indeed there has been a security update in between which became -14.
-15 was just accepted into jaunty-proposed which re-includes this fix.
It will build now and be available in a few hours.

Revision history for this message
Bruno Girin (brunogirin) wrote :

I can confirm that the -15 kernel in jaunty-proposed resolves the problem on my EeePC 701. I also did the "World of Goo" test: it works great :-) To install it, follow the instructions given by Tommy here: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/349314/comments/140 and replace the last command with:

sudo apt-get install linux-image-2.6.28-15-generic

Martin, is there a chance that this fix will make it into jaunty-updates soon-ish and before it gets clobbered by another important update?

Revision history for this message
Launchpad Janitor (janitor) wrote :
Download full text (7.2 KiB)

This bug was fixed in the package linux - 2.6.28-15.48

---------------
linux (2.6.28-15.48) jaunty-proposed; urgency=low

  [ Andy Whitcroft ]

  * SAUCE: pnp: add PNP resource range checking function
    - LP: #349314
  * SAUCE: i915: enable MCHBAR if needed
    - LP: #349314

  [ Brad Figg ]

  * SAUCE: Add information to recognize Toshiba Satellite Pro M10 Alps
    Touchpad
    - LP: #330885

  [ Colin Ian King ]

  * Input: atkbd - add forced release keys quirk for Samsung Q45
    - LP: #347623

  [ Manoj Iyer ]

  * SAUCE: Added quirk to enable the installer to recognize NetXen NIC.
    - LP: #389603

  [ Stefan Bader ]

  * SAUCE: input: Blacklist digitizers from joydev.c
    - LP: #300143

  [ Tim Gardner ]

  * Revert "SAUCE: md: wait for possible pending deletes after stopping an
    array"
    - LP: #334994

  [ Upstream Kernel Changes ]

  * bonding: Fix updating of speed/duplex changes
    - LP: #371651
  * net: fix sctp breakage
    - LP: #371651
  * ipv6: don't use tw net when accounting for recycled tw
    - LP: #371651
  * ipv6: Plug sk_buff leak in ipv6_rcv (net/ipv6/ip6_input.c)
    - LP: #371651
  * netfilter: nf_conntrack_tcp: fix unaligned memory access in tcp_sack
    - LP: #371651
  * xfrm: spin_lock() should be spin_unlock() in xfrm_state.c
    - LP: #371651
  * bridge: bad error handling when adding invalid ether address
    - LP: #371651
  * bas_gigaset: correctly allocate USB interrupt transfer buffer
    - LP: #371651
  * USB: EHCI: add software retry for transaction errors
    - LP: #371651
  * USB: fix USB_STORAGE_CYPRESS_ATACB
    - LP: #371651
  * USB: usb-storage: increase max_sectors for tape drives
    - LP: #371651
  * USB: gadget: fix rndis regression
    - LP: #371651
  * USB: add quirk to avoid config and interface strings
    - LP: #371651
  * cifs: fix buffer format byte on NT Rename/hardlink
    - LP: #371651
  * b43: fix b43_plcp_get_bitrate_idx_ofdm return type
    - LP: #371651
  * Add a missing unlock_kernel() in raw_open()
    - LP: #371651
  * x86, PAT, PCI: Change vma prot in pci_mmap to reflect inherited prot
    - LP: #371651
  * security/smack: fix oops when setting a size 0 SMACK64 xattr
    - LP: #371651
  * x86, setup: mark %esi as clobbered in E820 BIOS call
    - LP: #371651
  * dock: fix dereference after kfree()
    - LP: #371651
  * mm: define a UNIQUE value for AS_UNEVICTABLE flag
    - LP: #371651
  * mm: do_xip_mapping_read: fix length calculation
    - LP: #371651
  * vfs: skip I_CLEAR state inodes
    - LP: #371651
  * net/netrom: Fix socket locking
    - LP: #371651
  * kprobes: Fix locking imbalance in kretprobes
    - LP: #371651
  * netfilter: {ip, ip6, arp}_tables: fix incorrect loop detection
    - LP: #371651
  * ALSA: hda - add missing comma in ad1884_slave_vols
    - LP: #371651
  * SCSI: libiscsi: fix iscsi pool error path
    - LP: #371651
  * SCSI: libiscsi: fix iscsi pool error path again
    - LP: #371651
  * posixtimers, sched: Fix posix clock monotonicity
    - LP: #371651
  * sched: do not count frozen tasks toward load
    - LP: #371651
  * spi: spi_write_then_read() bugfixes
    - LP: #371651
  * powerpc: Fix data-corrupting bug in __futex_atomic_op
    - LP...

Read more...

Changed in linux (Ubuntu Jaunty):
status: Fix Committed → Fix Released
Revision history for this message
Bruno Girin (brunogirin) wrote :

I just updated my EeePC 701 with the -15 kernel that is now available in jaunty-updates and I can confirm the bug fixed with that version of the kernel too.

Revision history for this message
Purdy (jason-purdy) wrote :

New UNR user here & I have the same problem (screen refreshes, online flash video staggers and netbook-launcher menu changes are painfully slow [w/ high cpu rates]) after a new install. I'm using a Dell Inspiron Mini 12. My graphics card has something different, so maybe that's the cause.

$ lspci
...
00:02.0 VGA compatible controller: Intel Corporation System Controller Hub (SCH Poulsbo) Graphics Controller (rev 07)
...

$ uname -a
Linux mini-jason 2.6.28-15-generic #49-Ubuntu SMP Tue Aug 18 18:40:08 UTC 2009 i686 GNU/Linux

I installed the 2.6.28-11.43~lp349314apw5 kernel and that didn't change. I tried the force low graphics setting in gconf-editor and that didn't work. I also tried the "Options" "Tiling" "False" in xorg.conf and that also didn't work.

Let me know if I can provide any more information...

Thanks!

Revision history for this message
Purdy (jason-purdy) wrote :

I tried re-compiling the kernel (quite a slow process here on the netbook ;)) and these two warnings seemed relevant:

  Building modules, stage 2.
  MODPOST 2292 modules
WARNING: drivers/video/vesafb.o(.exit.text+0x57): Section mismatch in reference from the function vesafb_remove() to the (unknown reference) .init.data:(unknown)
The function __exit vesafb_remove() references
a (unknown reference) __initdata (unknown).
This is often seen when error handling in the exit function
uses functionality in the init path.
The fix is often to remove the __initdata annotation of
(unknown) so it may be used outside an init section.

WARNING: drivers/video/vesafb.o(.exit.text+0x62): Section mismatch in reference from the function vesafb_remove() to the variable .init.data:vesafb_fix
The function __exit vesafb_remove() references
a variable __initdata vesafb_fix.
This is often seen when error handling in the exit function
uses functionality in the init path.
The fix is often to remove the __initdata annotation of
vesafb_fix so it may be used outside an init section.

Perhaps my issue should be refiled as a different bug since the hardware is not the i915?

Revision history for this message
Bartek (tschew) wrote :

Hi Purdy,

I don't have a poulsbo machine but a quick search brought this up: https://wiki.ubuntu.com/HardwareSupportComponentsVideoCardsPoulsbo

Maybe that will resolve your issues? Not sure whether X will automatically load the relevant driver though after you've installed the packages...

-Bartek

Revision history for this message
Purdy (jason-purdy) wrote :

Hi Bartek,

Thanks for the pointer! That did the trick. I had to add the ppa source in /etc/apt/sources.list.d and then install the psb drivers and now my video is spotless and the performance is great.

Now on to the sound. ;)

affects: linux (Ubuntu Jaunty) → linux-meta (Ubuntu Jaunty)
Martin Pitt (pitti)
affects: linux-meta (Ubuntu) → linux (Ubuntu)
Revision history for this message
cbartle (cliff-bartle) wrote :

hi andy, i have just installed ubuntu on my eeepc and video is very poor would this fix be in the latest ubuntu download

Revision history for this message
Bruno Girin (brunogirin) wrote :

@cbartle: it should resolve the problem indeed. Go to Update Manager, make sure you install all the latest updates, reboot when asked and it should work fine. If it doesn't, can you add a further comment to this thread with the exact model of EeePC you have please?

Tman (t-mac1-y-ming11)
Changed in linux (Ubuntu):
status: Fix Released → Fix Committed
Revision history for this message
Steve Langasek (vorlon) wrote :

Please do not change bug status without explanation.

Changed in linux (Ubuntu):
status: Fix Committed → Fix Released
tags: added: iso-testing
Russell Bate (russwbate)
Changed in linux (Ubuntu):
status: Fix Released → Fix Committed
Martin Pitt (pitti)
Changed in linux (Ubuntu):
status: Fix Committed → Fix Released
Changed in linux:
status: Confirmed → Fix Released
Changed in linux:
importance: Unknown → Medium
Displaying first 40 and last 40 comments. View all 171 comments or add a comment.
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.