Bug#523589: xserver-xorg-video-radeon: dualhead broke with upgrade

[di dit]

2009/4/14 Alex Deucher <alexdeucher at gmail.com>:
> Both heads are probably using the same crtc, but xrandr isn't smart
> enough to assign a different one.  You can see what crtc is in use by
> running xrandr --verbose.  You can assign a different crtc to the
> other head like so:
> xrandr --output DVI-1 --crtc 1 --mode 1280x1024 --left-of DVI-0


Thanks for the suggestion. Unfortunately, it doesn't solve my problem.
Here are the results of some tests obtained using the following
minimal xorg.conf file:
Section "Screen"
        Identifier      "screen"
        SubSection "Display"
                Depth           24
                Virtual 2600 1024
        EndSubSection
EndSection

The output of xrandr --verbose at each step is given in the attached file.

1/ startx (window manager e16) results in monitors in clone mode as expected.
DVI-0 and DVI-1 do indeed use the same crtc (0).

2/ xrandr --output DVI-0 --crtc 0 --mode 1280x1024
nothing happens (as expected)

3/ xrandr --output DVI-1 --crtc 1 --mode 1280x1024
The monitor connected to DVI-0 enter power save mode.

4/ xrandr --output DVI-0 --off
nothing happens except that the following command wouldn't change
anything without it.

5/ xrandr --output DVI-0 --crtc 0 --mode 1280x1024
DVI-0 monitor goes on, DVI-1 monitor goes into power save mode

Another "funny" issue is that after running startx and without
"playing" with xrandr, if I switch off the
monitor connected to DVI-1 then the mouse pointer doesn't move on the
other monitor. This problem
doesn't exist when the monitor connected to DVI-1 is on and the other one off.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: log.radeon
Type: application/octet-stream
Size: 31347 bytes
Desc: not available
Url : http://lists.x.org/archives/xorg-driver-ati/attachments/20090415/010a2791/attachment-0001.obj