[Bug 79488] New: Checkerboard tearing on Radeon card

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=79488

          Priority: medium
            Bug ID: 79488
          Assignee: dri-devel at lists.freedesktop.org
           Summary: Checkerboard tearing on Radeon card
          Severity: normal
    Classification: Unclassified
                OS: All
          Reporter: dan.doel at gmail.com
          Hardware: Other
            Status: NEW
           Version: unspecified
         Component: DRM/Radeon
           Product: DRI

Created attachment 100211
  --> https://bugs.freedesktop.org/attachment.cgi?id=100211&action=edit
dmesg while the machine is in the 'tearing' state

For several months now, I've occasionally run into situations where my screen
displays a strange artifact or tearing pattern. It is only visible when things
are changing, of course, and it looks like the monitor is divided into small
squares, and two different 'frames' are drawn alternately. Here is a video of
the problem I took back at the end of February:

  https://www.youtube.com/watch?v=e2T8BeE0j74

I've just encountered this again, and it reminded me to finally post a bug.

You'll notice that the video was recorded on a cell phone. I tried using a
desktop recorder (the one built into gnome-shell) and the resulting video did
not display the artifacts. So it would seem to be happening at a level below
what capture can detect.

I'll attach dmesg and xorg logs from when the problem is happening. You'll see
that there are lots of page flip errors in both. However, I think that actually
might be a red herring. Doing a tail -f on the logs, I can cause artifacts to
display (for instance, selecting a different channel in my irc client, causing
highlighting to change, which briefly displays checkerboard highlights) without
a corresponding xorg error. And I also frequently see those errors accompanied
by ordinary tearing. But I can't definitively say that they aren't related,
either.

I'm using:

  mesa from git as of about a week ago (rev 63069)
  latest glamor (as of 2014-03-13)
  xf86-video-ati 7.3.0
  kernel 3.14.4

But this has been happening for significantly longer than just those versions.

I have a Radeon HD 7870 (pitcairn).

I'm not certain that I've assigned this bug to the right project, although at
least some of the visible errors involve drm. Apologies if it's the fault of
something else, though.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/dri-devel/attachments/20140531/18b1a683/attachment.html>