Bug 35466 - Screen corruption with 2.6.38 and i855
Summary: Screen corruption with 2.6.38 and i855
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: x86 (IA32) Linux (All)
: medium normal
Assignee: Chris Wilson
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2011-03-20 08:02 UTC by Cesare Leonardi
Modified: 2017-07-24 23:05 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments

Description Cesare Leonardi 2011-03-20 08:02:29 UTC
With 2.6.38 my screen shows many corruptions, as i already described here:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=619022

Here you can find also an attachment that shows the problem.

Some information not reported in Debian bug report:
xserver-xorg 1:7.6+4
xserver-xorg-core 2:1.9.4-90
xserver-xorg-video-intel 2:2.14.0-4
libdrm-intel1 2.4.23-3

Let me know if you need more info.

Cesare.
Comment 1 Chris Wilson 2011-03-20 08:06:12 UTC
Update your userspace to include:

xf86-video-intel commit d21d781466785c317131a8a57606925867265dc8
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Tue Feb 22 18:31:44 2011 +0100

    Fix relaxed tiling on gen2
    
    Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

[2.14.901], but watch out for bug 35460.
Comment 2 Cesare Leonardi 2011-03-20 08:20:57 UTC
(In reply to comment #1)
> [2.14.901], but watch out for bug 35460.

Thank you very much for the really fast reply, Chris.
So, if i have understood correctly:
- 2.6.38 require at least an intel driver >= 2.14.901
- even if the driver is updated there are still corruption problems (#35460) on which you are working on.

So while waiting for updated intel driver, for an i855 user should be wise to use an older kernel.

Am i correct?

Cesare.
Comment 3 Chris Wilson 2011-03-20 08:28:08 UTC
Or an older userspace. I introduced lazy fencing as I developed the kernel patches, about 6 months ago. It was only with the general release of the kernel enabling patches that the gen2 bug in the userspace code was discovered.

If I had sufficient foresight I would have also added a module option to disable relaxed fence allocation in case of emergencies such as this. :(


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.