Bug 20470

Summary: UploadToScreen causes corruption
Product: xorg Reporter: Thomas Wendt <thoemy>
Component: Driver/radeonhdAssignee: Luc Verhaegen <lverhaegen>
Status: RESOLVED WONTFIX QA Contact: Xorg Project Team <xorg-team>
Severity: normal    
Priority: medium    
Version: unspecified   
Hardware: Other   
OS: All   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
Some pictures of mouse cursors
none
Firefox buttons
none
Xorg log
none
xorg.conf none

Description Thomas Wendt 2009-03-04 13:52:11 UTC
Created attachment 23533 [details]
Some pictures of mouse cursors

Hi,

after the DFS and UTS hooks were enabled I experienced corruption. The
corruption affects the mouse cursor, sometimes window decorations,
nearly always the buttons in firefox, sometimes the desktop background,
sometimes launchers in the gnome-panel and other things.

Attached are screenshots that demonstrate how the corruption looks. The
corruption of the mouse cursor is not visible on screenshots so I took them with a digital camera.

The stripes appear only on the right of the cursor and it changes if I move the cursor into another text input box.

The corruptions are not very serious. I didn't complained until yet
because I was busy and I hoped that the problem gets fixed. 

Disabling the UTS hook fixes the corruptions. Unfortunately the whole
desktop is then very unresponsive.

The first bad commit was f1b9c19683090272a29a3e4c9e2c054b4a32b47c
(R6xx/R7xx EXA: add accelerated UTS/DFS hooks)


Some information about my system:

X.Org X Server 1.5.99.902 (1.6.0 RC 2)
radeonhd commit 9c8ab2dfbe61120298c4b46a2b49245c6779dbc2
drm commit 189cd825a8a2cc2c0a28023589a5f18a4122410b

01:00.0 VGA compatible controller: ATI Technologies Inc Radeon HD 3870
(prog-if 00 [VGA controller])
        Subsystem: PC Partner Limited Device e620
        Flags: bus master, fast devsel, latency 0, IRQ 25
        Memory at c0000000 (64-bit, prefetchable) [size=256M]
        Memory at ff1f0000 (64-bit, non-prefetchable) [size=64K]
        I/O ports at b000 [size=256]
        Expansion ROM at ff1c0000 [disabled] [size=128K]
        Capabilities: [50] Power Management version 3
        Capabilities: [58] Express Legacy Endpoint, MSI 00
        Capabilities: [a0] Message Signalled Interrupts: Mask- 64bit+ Queue=0/0
Enable-
        Capabilities: [100] Vendor Specific Information <?>
        Kernel driver in use: radeon
        Kernel modules: fglrx


Any other things I can provide?
Comment 1 Thomas Wendt 2009-03-04 13:52:56 UTC
Created attachment 23534 [details]
Firefox buttons
Comment 2 Darin McBride 2009-03-04 14:01:07 UTC
Actually, I'm getting other, similar mouse corruptions from time to time, though I've not narrowed down anything enough to figure out what causes it.

X.Org X Server 1.5.3
Release Date: 5 November 2008
X Protocol Version 11, Revision 0
Build Operating System: Linux 2.6.27-gentoo-r7 x86_64
Current Operating System: Linux naboo 2.6.27-gentoo-r8 #1 SMP Tue Mar 3 12:06:12 MST 2009 x86_64
Build Date: 08 February 2009  08:04:58AM

DRM kernel module from the r6xx-r7xx tree as of 03032009, radeonhd X11 driver from master as of 02272009.
Comment 3 Alex Deucher 2009-03-04 14:06:25 UTC
Ugh AGP.  Please attach your xorg log and config.

you can disable UTS and DFS like so:
Option "ExaNoDownloadFromScreen"
Option "ExaNoUploadToScreen"

You might also try forcing the bustype to PCIE to use the on-chip GART interface rather than AGP (i.e., return RHD_CARD_PCIE in rhdGetCardType()).
Comment 4 Thomas Wendt 2009-03-04 14:59:19 UTC
Created attachment 23535 [details] [review]
Xorg log

Actually my card is a PCIe one.
Comment 5 Thomas Wendt 2009-03-04 15:00:32 UTC
Created attachment 23536 [details]
xorg.conf
Comment 6 Thomas Wendt 2009-03-26 14:49:22 UTC
Either switching the kernel from 2.6.27 to 2.6.29 or to a more recent XServer version (from 2:1.4.2-11 to 2:1.5.99.902-1 in Debian) fixed that issue for me. I think the latter one is more likely.
Comment 7 Thomas Wendt 2009-03-31 14:10:02 UTC
I just forgot to compile the drm and radeon module. So problem persists.
Comment 8 Pierpaolo Follia 2009-06-05 00:06:42 UTC
I can confirm that using:

Option "ExaNoDownloadFromScreen"

solve this problem on my Radeon Mobility X1400
Comment 9 Pierpaolo Follia 2009-06-12 00:21:15 UTC
I add this: disabling download from screen solves the problem, but now I noticed that the system sometimes becomes unusable, freezing for 1 or 2 seconds during some operations (like changing the desktop background or during fade in/out of windows in compiz).
Comment 10 Jeremy Huddleston Sequoia 2011-10-16 15:57:45 UTC
Does this issue occur with the preferred ati driver (xf86-vide-ati)?  If so, please move this to the Driver/Radeon component.  

Development of radeonhd has pretty much halted and development focus is on the ati driver.  Please see http://www.x.org/wiki/radeonhd

If the issue does not exist in the ati driver (or if there is no response to this message), this bug will be closed as WONTFIX unless someone contributes a patch.
Comment 11 Thomas Wendt 2011-10-22 06:25:41 UTC
(In reply to comment #10)
> Does this issue occur with the preferred ati driver (xf86-vide-ati)?  If so,
> please move this to the Driver/Radeon component.  
> 
> Development of radeonhd has pretty much halted and development focus is on the
> ati driver.  Please see http://www.x.org/wiki/radeonhd
> 
> If the issue does not exist in the ati driver (or if there is no response to
> this message), this bug will be closed as WONTFIX unless someone contributes a
> patch.

Please close as WONTFIX. Not sure if the problem still exists with the radeonhd driver but the xf86-video-ati doesn't show the corruption.

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.