Bug 28426 - hardware cursor corruption with radeon+kms
Summary: hardware cursor corruption with radeon+kms
Status: RESOLVED MOVED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Radeon (show other bugs)
Version: unspecified
Hardware: x86-64 (AMD64) Linux (All)
: medium normal
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-06-07 09:16 UTC by Francesco R
Modified: 2019-11-19 08:13 UTC (History)
2 users (show)

See Also:
i915 platform:
i915 features:


Attachments
various log and information on the system (50.82 KB, application/zip)
2010-06-07 09:20 UTC, Francesco R
no flags Details
VGA-0-cursor-broken.jpeg (777.25 KB, image/jpeg)
2010-06-07 09:20 UTC, Francesco R
no flags Details
DVI-0-cursor-broken.jpeg (360.87 KB, image/jpeg)
2010-06-07 09:21 UTC, Francesco R
no flags Details
cursor-gimp.jpeg (608.90 KB, image/jpeg)
2010-06-07 09:21 UTC, Francesco R
no flags Details

Description Francesco R 2010-06-07 09:16:41 UTC
please see also bug #19215, may be related

After some days the xorg server is up and running the cursor get corrupted.
This is a kde4 desktop on a dual 1920x1080 screen powered by an ATI Radeon HD 4670.
The bug happens often, having played a game like Aquaria (in windowed mode) help a lot, but it does not guarantees an immediate trigger.
This time xorg was up and running for about four days (not totally sure), yesterday I've played with the already mentioned Aquaria.

please feel free to request further information

since I don't know where to start I've tried to be broad on covering the system information, here is what the attached zip file contain:

 - "cursor-gimp.jpeg" 
    cursor as it should be (after a reboot)
 
 - "VGA-0-cursor-broken.jpeg"
    cursor as it's shown in the left monitor VGA-0
 
 - "DVI-0-cursor-broken.jpeg"
    cursor as it's shown in the right monitor DVI-0
 
 - "Xorg.0.log"
    log for the server on vt7 whit kde running
 
 - "Xorg.1.log"
    tentative to start another X server on vt8, it also had a broken cursor
 
 - "xorg.conf"
    "SWcursor" is commented out
 
 - "radeontool--debug-regs.log"
    `radeontool debug regs` output
 
 - "installed-software.log"
    radeon and x11 related software installed on this pc
 
 - "modinfo-radeon.log"
    `modinfo radeon` output
 
 - "lsmod.log"
    `lsmod` output
 
 - "xrandr--verbose.log"
    `xrandr -q --verbose` output
 
 - "lspci-vvv.log"
    `lspci -vvv` output
 
 - "lshw.log"
    `lshw` output, it's an hardware lister
 
 - "dmesg.log"
    `dmesg` output
Comment 1 Francesco R 2010-06-07 09:20:20 UTC
Created attachment 36105 [details]
various log and information on the system
Comment 2 Francesco R 2010-06-07 09:20:52 UTC
Created attachment 36106 [details]
VGA-0-cursor-broken.jpeg
Comment 3 Francesco R 2010-06-07 09:21:24 UTC
Created attachment 36107 [details]
DVI-0-cursor-broken.jpeg
Comment 4 Francesco R 2010-06-07 09:21:59 UTC
Created attachment 36108 [details]
cursor-gimp.jpeg
Comment 5 Daniel Fort 2010-08-09 10:52:24 UTC
got absolutely the same problems with HD 4870.
Comment 6 Roger Luethi 2011-08-17 03:06:17 UTC
Attachment 36107 [details] (DVI-0-cursor-broken.jpeg) looks like what I am seeing.

Radeon HD 4650
Radeon driver 6.14.2
no composite

I have not found a way to trigger it, it just happens after a few hours working on a desktop.

1) the cursor corruption seems to occur when the cursor crosses the boundary between the two displays of a dual-head setup

2) restarting X does not get back the regular cursor; it stays corrupted until the machine is rebooted

3) I did _not_ see this problem with a Radeon X300SE on an almost identical software setup
Comment 7 Jürg Billeter 2011-11-02 00:50:41 UTC
I have the same issue with a Radeon HD 4770 and two 1920x1200 monitors connected via DVI. I'm currently running Linux 3.0.6 and xf86-video-ati 6.14.2, but I've been seeing this for a long time.
Comment 8 Michel Dänzer 2011-11-02 03:53:49 UTC
Kernel 3.0.7 has one HW cursor fix, 3.1 has more and other HW cursor changes. Do any of those help?
Comment 9 Francesco R 2011-11-02 12:39:09 UTC
sorry, I've switched to nvidia, cannot test
Comment 10 Jürg Billeter 2011-11-03 00:25:59 UTC
I've now updated to Linux 3.1. It's looking fine so far but it usually takes a couple days or so until the cursor corruption starts. I will report back.
Comment 11 Jürg Billeter 2011-11-09 08:51:09 UTC
My Linux 3.1 system has been running for over 6 days now without issues. Seems like it may finally be fixed for good :) Thanks!
Comment 12 Roger Luethi 2011-11-10 00:24:39 UTC
Like Jürg Billeter, I haven't seen the problem since switching to Linux 3.1.

I don't currently use the affected system as much as I usually do, so it may be too early to celebrate, but I guess the bug could be closed as resolved and reopened later if need be.
Comment 13 Michel Dänzer 2011-11-10 09:46:42 UTC
(In reply to comment #12)
> Like Jürg Billeter, I haven't seen the problem since switching to Linux 3.1.

So far, so good. Thanks for the updates, guys.

> I don't currently use the affected system as much as I usually do, so it may be
> too early to celebrate, but I guess the bug could be closed as resolved and
> reopened later if need be.

Sounds like a plan. :)
Comment 14 Kai Arif 2012-01-26 05:39:19 UTC
I'm having the exact same problem and I'm using kernel 3.2 with a Radeon HD 4350 on Debian sid. I had it with 3.0 an 3.1 too. Right now I have a vertical line on my left screen and the rainbow square on my right screen. It happens randomly and usualy goes away after a while. If it doesn't I have to reboot.
Comment 15 Roger Luethi 2012-01-26 06:21:17 UTC
(In reply to comment #14)
> I'm having the exact same problem and I'm using kernel 3.2 with a Radeon HD
> 4350 on Debian sid. I had it with 3.0 an 3.1 too. Right now I have a vertical
> line on my left screen and the rainbow square on my right screen. It happens
> randomly and usualy goes away after a while. If it doesn't I have to reboot.

Is this a dual-head system, too, or do you have just one display?

For what it is worth, I have not experienced the problem since my earlier report that Linux 3.1 seems to fix it.

So the bug appears to be better hidden now, which tends to make fixing harder.

Maybe some driver developer can suggest a way out?
Comment 16 Michel Dänzer 2012-01-30 07:41:01 UTC
(In reply to comment #14)
> It happens randomly and usualy goes away after a while.

Given that this bug really seems fixed for others, and comment #7 says only a reboot helped, your problem sounds more similar to bug 33183 than this one.
Comment 17 Martin Peres 2019-11-19 08:13:26 UTC
-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/amd/issues/131.


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.