[Bug 72346] New: xf86-video-intel-2.99 SNA mode gives broken rendering

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=72346

          Priority: medium
            Bug ID: 72346
          Assignee: chris at chris-wilson.co.uk
           Summary: xf86-video-intel-2.99 SNA mode gives broken rendering
        QA Contact: intel-gfx-bugs at lists.freedesktop.org
          Severity: normal
    Classification: Unclassified
                OS: All
          Reporter: tiwai at suse.de
          Hardware: Other
            Status: NEW
           Version: git
         Component: Driver/intel
           Product: xorg

Reported by Jan Engelhardt on openSUSE bugzilla:
    https://bugzilla.novell.com/show_bug.cgi?id=853066

Here is the quote:

Rendering of certain bitmappery is broken: http://inai.de/mah04467a.mp4
So far, I have only been able to observe this with vncviewer.

xorgserver  xf86intel  mode           result
1.13        2.20       default(uxa?)  ok
1.13        2.99       default(sna)   crap
1.13        2.99       uxa            ok
1.14        2.20       default(uxa?)  ok
1.14        2.99       default(sna)   crap
1.14        2.99       uxa            ok

Using either the 2.20 driver, or switching to UXA in 2.99 resolves the issue.
It seems to be independent of the Xorg server version.

I read that SNA stands for SandyBridge New A..something. However, I do have a
Pineview Atom with GMA 3150, so defaulting to SandyBridgey-specific rendering
would seem hazardous.

(in comment #12)
In the video, I used the Window Rollup feature of the local window manager
(xfce) to roll the VNC window up, and then down again, so as to cause a repaint
of the local X window, nothing more. Moving windows inside the VNC is also one
way to cause remote repaints and subsequently local repaints, which is also
shown.

A potential cure seems to be to reinitialize the rendering pipes, for example
by turning off at least one (either LVDS1 or VGA1) and then going back to the
desired setup:

  xrandr --output LVDS1 --off
  xrandr --auto


***

The problem still happens with git 2.99.906-78-g6c37ab3ddc59.  It's PineView
x86-64.

Interestingly, we couldn't reproduce the issue here with another machine with
the same chip.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20131205/8dd5ab8a/attachment.html>