Dear radeon-driver developers,



your line
	"(WW) RADEON(0): R500 support is under development. Please report any issues to xorg-driver-ati@lists.x.org"
in the X server configuration log file encouraged me to send this mail.



I'm trying to get a dual-head set-up on a Dell Latitude D531 laptop
with an LG Flatron W2242T external monitor.

In my current set-up, the cursur flickers when it is close to a screen
x coordinate that is a multiple of 256 pixels.

When I switch to a software cursor, this problem disappears.

In order to isolate the problem, I used an X root image that allows
measuring x,y coordinates.  Using an optical pocket lens, I tried to
find the flickering areas as good as my weak eyes allowed.  I include
a manually edited screen shot showing the results.

It seems that every pixel of the hardware cursor crossing a vertical
line (x == n * 256 - 8 for some int n>=1) induces a "flickering
shadow" line to its right, until (and including) x == n * 256 - 1.



Attached please find the tar-ed and gzip-ped directory
radeonHwCursor.tar.gz, containing the following files:

256-08v.gif
	The file used as X root image ("xv -root 256-08v.gif -quit").

bios
	Information about the laptop obtained from the bios set-up
	menu, in particular video controller, bios-verson, and memory.

editedScreenShot.gif
	The edited screen shot mentioned above.
	The cursors are probably not drawn 100% correct, but I did
	my best.  Observed flickering zones are indicated by light
	and dark magenta pixels; cf. the legend in the picture.

hwCursor_.xsession-errors
	The error output of the X server when using hardware cursors;
	seems to contain no valuable information on the problem.

hwCursor_Xorg.0.log
	The logging output from X server configuration (copied from
	/var/log/Xorg.0.log), using a hardware cursor.

hwCursor_xorg.conf
	The X server configuration file to use a hardware cursor
	(copied from /etc/X11/xorg.conf).

swCursor_.xsession-errors
swCursor_Xorg.0.log
swCursor_xorg.conf
	Similar for software-cursors.
	In particular, hwCursor_xorg.conf and swCursor_xorg.conf
	differ only in two lines:
	'#     Option          "SWcursor"      "true"' vs.
	'      Option          "SWcursor"      "true"'.



As I am rather unexperienced in all that graphics-driver stuff, my
configuration may contain awful flaws (improvement hints are
appreciated).

Nevertheless, I guess, the observed flickering might indicate a bug in
the radeon driver, since it disappears when using a software cursor,
and the dimensions of the problem zones are powers of 2.

Thus, I hope this mail could be helpful to you.

If you need more information (maybe some other configuration or log
files), please ask.  I'll read my mail about every third day, since
I'm at home, where I've only a slow tcp connection.

Please do not publish my e-mail address on any www pages, to prevent
spam mails.



Best regards
Jochen Burghardt