Patch for fpdi and some questions, maybe a bug in atom parser
Hello, I currently try to dig into the code to fix somehow the problem with the granulary output on my lcd display (Lenovo with x1400 and a LCD 1440x1050). I've got the following lines in my xorg.log (II) RADEONHD(0): LVDS SEQ Dig onto DE: 40 (II) RADEONHD(0): LVDS SEQ DE to BL: 250 (II) RADEONHD(0): LVDS Off Delay: 170 (II) RADEONHD(0): LVDS Duallink: 0x1 (II) RADEONHD(0): LVDS 24Bit: 0x0 (II) RADEONHD(0): LVDS FPDI: 0x490 (--) RADEONHD(0): Detected a 18bit dual link panel. (--) RADEONHD(0): Attaching Output LVDS to Connector PANEL 24 Bit is disabled, thats why only 18bit are used for the panel. But maybe this is a bug in the parser. Because when I use fglrx or Windows the colors seam to be more brilliant and also for example videos/pictures do not show those artefacts/granularity. An other thing is that fpdi returns 0x490 as value, which doesn't make sense, or? When I looked into the spec there is only 0/1 possible. Possible because line 1021 is '*' and not a '&' and in line 1063 would be the same problem too. So a more general question: How do you know which are the correct bits you use? Does atom bios just provide an other way to access the registers, which are described in the reference guide or is there some other magic. Thank you in advance. Martin
On Mon, Dec 10, 2007 at 04:10:29PM +0100, Martin Nowack wrote:
Hello,
I currently try to dig into the code to fix somehow the problem with the granulary output on my lcd display (Lenovo with x1400 and a LCD 1440x1050).
I've got the following lines in my xorg.log
(II) RADEONHD(0): LVDS SEQ Dig onto DE: 40 (II) RADEONHD(0): LVDS SEQ DE to BL: 250 (II) RADEONHD(0): LVDS Off Delay: 170 (II) RADEONHD(0): LVDS Duallink: 0x1 (II) RADEONHD(0): LVDS 24Bit: 0x0 (II) RADEONHD(0): LVDS FPDI: 0x490 (--) RADEONHD(0): Detected a 18bit dual link panel. (--) RADEONHD(0): Attaching Output LVDS to Connector PANEL
24 Bit is disabled, thats why only 18bit are used for the panel. But maybe this is a bug in the parser. Because when I use fglrx or Windows the colors seam to be more brilliant and also for example videos/pictures do not show those artefacts/granularity.
An other thing is that fpdi returns 0x490 as value, which doesn't make sense, or? When I looked into the spec there is only 0/1 possible. Possible because line 1021 is '*' and not a '&' and in line 1063 would be the same problem too.
So a more general question: How do you know which are the correct bits you use? Does atom bios just provide an other way to access the registers, which are described in the reference guide or is there some other magic.
Thank you in advance.
Martin
The fixed reading of FPDI was pushed now, sorry for the long delay here :) You were fully correct btw, on all three points. The reading of FPDI was wrong in all cases, amazingly :) Thanks, Luc Verhaegen. SUSE/Novell X Driver Developer. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
participants (2)
-
Luc Verhaegen -
Martin Nowack