Ubuntu
mesa package

[i965] X3100 Plasma Desktop gets invalidated with short regular line corruption unless Composite is disabled

Bug #272699 reported by Andres Järv
54
This bug affects 8 people
Affects Status Importance Assigned to Milestone
xf86-video-intel
Fix Released
High
mesa (Ubuntu)
Invalid
High
Unassigned

Bug Description

Sometimes when I have lots of windows open and close or move them the Plasma desktop doesn't update correctly and usually the area directly behind the closed window doesn't get repainted but instead shows random data (sometimes data from other windows, not interactable of course).

This is a pretty extreme example when I closed windows covering the whole desktop and it all got corrupt. It didn't return to being normal when I switched virtual desktops, only when I right clicked on the desktop.

http://www.hot.ee/hvc/plasma_b0rkage.png

I'm running a Intel Core 2 Duo on a laptop. The chipset is Intel's GM965 with the X3100 graphics. I'm running Intrepid with all the newest Xorg and KDE stuff installed (7.4, xserver 1.5 etc).

Revision history for this message
Andres Järv (andresjarv) wrote :

As you can see, the background of the Kopete icon in the system tray is corrupt as well. This also happens from time to time.

Revision history for this message
Jonathan Thomas (echidnaman) wrote :

Is this still an issue with the latest updates?

Changed in kdebase-workspace:
status: New → Incomplete
Revision history for this message
Andres Järv (andresjarv) wrote :

Yes.

Revision history for this message
Jonathan Thomas (echidnaman) wrote :

I've been able to do that, but it's only happened with the nouveau drivers.

Changed in kdebase-workspace:
status: Incomplete → New
Revision history for this message
Harald Sitter (apachelogger) wrote :

I get this sometimes on intel as well. A click on the desktop (triggers refresh) usually fixes. Most likely related to the KWin patch, therefore we should target a fix for 8.10 (or rather have to, considering that this doesn't happen with stock 4.1.2).

Nice fact is that I don't use 3D effects which make this regression a lot more annoying.

Changed in kdebase-workspace:
importance: Undecided → High
milestone: none → ubuntu-8.10
status: New → Confirmed
Revision history for this message
Harald Sitter (apachelogger) wrote :

The trouble causing patch was dropped because the quality was insufficient.

Please confirm that this issue is not appearing with latest packages.

Changed in kdebase-workspace:
status: Confirmed → Incomplete
Revision history for this message
Andres Järv (andresjarv) wrote :

I have had it occur a couple of times. But it's not as common as before.

Revision history for this message
Harald Sitter (apachelogger) wrote :

Maybe it was (partially) caused by X/the graphics driver.

Revision history for this message
Andres Järv (andresjarv) wrote :

Nah, it's actually happening at pretty much the same interval as before. My former impression was false.

Revision history for this message
Jonathan Thomas (echidnaman) wrote :

My dad's intel-graphics-powered laptop is doing this too with the latest Intrepid.

Changed in kdebase-workspace:
status: Incomplete → Confirmed
Revision history for this message
Andres Järv (andresjarv) wrote :

Adding the following to my xorg.conf seems to "fix" this bug.

Section "Extensions"
        Option "Composite" "Disable"
EndSection

Revision history for this message
nogac (nogacsoid) wrote :

X3100/GM965 support in intrepid is horrible. I have corruption, weird x crashes when using xv, artifacts, poor performance...

https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/252094

Revision history for this message
Jonathan Thomas (echidnaman) wrote :

Yeah, this is probably a bug in the intel driver, since basically all the reports have been in regards to that.

Changed in kdebase-workspace:
milestone: ubuntu-8.10 → none
Revision history for this message
Krinn (kr86420) wrote :

Same problem on an Eee PC 900 (GMA 900). Andres Järv's workaround does stop the random garbage appearing.

Revision history for this message
Jaime Soriano (kronoss) wrote :

Same problem on MacBook Santa Rosa (Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller (rev 03))

Revision history for this message
Robert Persson (ireneshusband) wrote :

I'm confused as to how a change to xorg.conf should make a difference if X is configured by hal now. How do you put the equivalent of Andres's fix in /etc/hal/fdi?

Revision history for this message
C Snover (launchpad-net-zetafleet) wrote :

X still reads preferences from the Xorg.conf if you put them there.

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

[This is an automatic notice.]

We'd like to forward your bug upstream, however upstream requires
that you first test it against their newer driver code.

To save you the effort of building the driver from source, we've built
packages for the driver and its new dependencies.

So you have a couple options:

 1. Download and test .debs for intrepid, from:
     https://edge.launchpad.net/~intel-gfx-testing/+archive

 -or-

 2. Download and test the Jaunty alpha-2 (or newer) Live CD,
     (which includes a beta of the new xserver 1.6 as well).
     See http://cdimage.ubuntu.com/releases/9.04/ for ISOs

Thanks ahead of time! You can simply reply to this email to report your
findings.

P.S., if you wish to forward your bug upstream yourself, please follow
these directions to do so:
  http://intellinuxgraphics.org/how_to_report_bug.html

Changed in xserver-xorg-video-intel:
status: Confirmed → Incomplete
Revision history for this message
C Snover (launchpad-net-zetafleet) wrote :

Tested provided updated drivers, problem still exists here (Q45 chipset, X4500, intrepid). It's worth noting the desktop does not get corrupted with Desktop Effects turned on (in Kubuntu), but corruption still occurs on the login page as well as when windows are brought back from minimisation or new windows are created. Also, 3D performance is even more abysmal (300fps in glxgears when running nothing else, down from a still extremely abysmal but not quite as bad 700fps in the current release driver), but that's another bug.

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

@Andres, I assume you're able to reproduce C Snover's findings?

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

Andres, can you please attach your /var/log/Xorg.0.log and the output of lspci -vvnn?

Also, for the 3D and performance issues, it would be worthwhile to test against the Jaunty alpha ISO.

Changed in xserver-xorg-video-intel:
status: Triaged → Incomplete
Revision history for this message
C Snover (launchpad-net-zetafleet) wrote :

Since I'm experiencing the same issues and Andres doesn't seem to be responding I will attach my own Xorg.0.log and lspci -vvnn data.

The performance issues are well documented in Bug #252094 and affect Jaunty even more severely.

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

Okay, thanks, hopefully Andres can chime in when he has a chance but we can focus on your case for upstreaming for now.

Thanks for including the needed files. (Btw, generally we like having the files each attached separately since we can then review them without having to download and unpack the tarball.)

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

Created an attachment (id=21856)
plasma_b0rkage.png

Forwarding this bug from a Ubuntu reporter:
https://bugs.edge.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/272699

[Problem]
With many windows open, sometimes KDE Plasma desktop shows corruption such as shown in the attached image (or available at http://www.hot.ee/hvc/plasma_b0rkage.png)

[Original Report]
Sometimes when I have lots of windows open and close or move them the Plasma desktop doesn't update correctly and usually the area directly behind the closed window doesn't get repainted but instead shows random data (sometimes data from other windows, not interactable of course).

This is a pretty extreme example when I closed windows covering the whole desktop and it all got corrupt. It didn't return to being normal when I switched virtual desktops, only when I right clicked on the desktop.

http://www.hot.ee/hvc/plasma_b0rkage.png

I'm running a Intel Core 2 Duo on a laptop. The chipset is Intel's GM965 with the X3100 graphics. I'm running Intrepid with all the newest Xorg and KDE stuff installed (7.4, xserver 1.5 etc).

As you can see, the background of the Kopete icon in the system tray is corrupt as well. This also happens from time to time.

Adding the following to my xorg.conf seems to "fix" this bug.

Section "Extensions"
        Option "Composite" "Disable"
EndSection

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

Created an attachment (id=21857)
Xorg.0.log

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

00:00.0 Host bridge [0600]: Intel Corporation 4 Series Chipset DRAM Controller [8086:2e10] (rev 03)
 Subsystem: Intel Corporation Device [8086:1003]

00:02.0 VGA compatible controller [0300]: Intel Corporation 4 Series Chipset Integrated Graphics Controller [8086:2e12] (rev 03)
 Subsystem: Intel Corporation Device [8086:1003]

Revision history for this message
In , Gordon Jin (gordon-jin) wrote :

The original Ubuntu reporter uses GM965, and the later reporter (with log provided) uses Q45.

Bryce, when you forward the bug report, could you please elaborate the driver versions? And are you able to reproduce with the upstream driver?

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

I've forwarded this bug upstream for you at https://bugs.freedesktop.org/show_bug.cgi?id=19494. Could you please subscribe yourself to this bug, in case they need further information or wish you to test something. Thanks ahead of time!

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

Focus on the Q45 issue please. The original reporter did not provide all required info so we are focusing on the Q45 reporter who said he had the same issue.

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

The upstream bug has questions for either of you...

Revision history for this message
In , Bugs-freedesktop-org-zetafleet (bugs-freedesktop-org-zetafleet) wrote :

Hi,

I am the reporter of the Q45 issue. The driver versions tested were the current Ubuntu release driver (2.4.1-0ubuntu7~intrepid), and a driver 2.5.1-1ubuntu5~intrepid which was provided by the Ubuntu Intel Graphics Driver Testing Team. I do not have a build environment in place currently to pull in the latest upstream driver for testing, but I will try to cobble something together and let you know what happens with it.

To also make sure my experience is properly echoed here, I am using kwin (4.1.3-0ubuntu1~intrepid1) as my window manager, and the desktop corruption occurs for pretty much any window that gets drawn over the desktop and then removed. Unlike with the original reporter, there do not need to be very many of them; simply opening and closing the KDE Menu, for example, will show graphics corruption (of the type illustrated in the original screenshot) in the area where the KDE Menu window used to be.

This problem does *not* occur when Desktop Effects are enabled within KDE. Additionally, clicking on the desktop seems to force a redraw which removes any graphical corruption that is present.

Regards,

Revision history for this message
C Snover (launchpad-net-zetafleet) wrote :

Sorry Bryce, I thought those questions were directed at you. I will respond upstream.

Bug Watch Updater (bug-watch-updater)
Changed in xserver-xorg-video-intel:
status: Unknown → Confirmed
Revision history for this message
In , Eric Anholt (eric-anholt) wrote :

Does this go away with UXA? The zero-copy texture-from-pixmap code in EXA is flawed and will fail when the rendering for a single frame is too large.

Revision history for this message
C Snover (launchpad-net-zetafleet) wrote :

Bryce,

Upstream is asking for testing with UXA, but from what I can tell, UXA doesn’t exist until driver version 2.5.0 (and that version was pretty bad the last time I tried it from the intel-gfx-testing PPA). Is it possible to get a backported 2.6.0 driver for use with intrepid, or would you recommend that I just try running a Jaunty livecd?

Let me know,

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

Hi C Snover,

2.6.0's UXA support requires xserver and kernel changes not available in intrepid, so I could recommend that you try running against a Jaunty Alpha-5 livecd.

Revision history for this message
RoDoLFo TX (rodolfo-inec) wrote :

This issue is still there, I can confirm. Not to mention that on Jaunty RC (9.04), if you leave the notebook (ACER Aspire 5315 - powered by intel X3100)for a few seconds and start typing something, the screen blinks from time to time. This is the first version of ubuntu that I see this weird behavior. Anyone can confirm this?

Revision history for this message
In , Eric Anholt (eric-anholt) wrote :

Feedback timeout (4 months).

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

Upstream has closed the bug as expired due to no response from reporter.
The issue is believed fixed with Karmic.

Changed in mesa (Ubuntu):
status: Triaged → Invalid
Bug Watch Updater (bug-watch-updater)
Changed in xserver-xorg-video-intel:
status: Confirmed → Fix Released
Bug Watch Updater (bug-watch-updater)
Changed in xserver-xorg-video-intel:
importance: Unknown → High
Bug Watch Updater (bug-watch-updater)
Changed in xserver-xorg-video-intel:
importance: High → Unknown
Bug Watch Updater (bug-watch-updater)
Changed in xserver-xorg-video-intel:
importance: Unknown → High
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

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