[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug

>xserver -xorg-video-ati 1:6.6.3-2 is very old, lots of changes were made
>since then, but no new stable version has been released. So there are 2
>major release candidates that you could try:

>* 1:6.6.193-1 is the natural successor of 1:6.6.3-2 with several fixes,
>it is available from http://people.debian.org/~bgoglin/ati-for-sid/

>* Then the driver got a large rework, especially in the area regarding
>monitor/mode detection, with randr-1.2 support. So 1:6.7.192-1 is
>available for unstable from

>Please try with any of these versions.

The new 6.7.192 version of the driver has a couple of quirks:

- Even though I specify a DisplaySize in my xorg.conf, it uses it's own... So 
if I want the DPI to be actual 'dpi' I have to: 

xrandr --fbmm 328x246

(Ye olde ruler told me the screen dimensions) This is important under KDE for 
example, because it doesn't force 96 or 120 dpi. Quite probably because that 
isn't 'literal' DPI. :p

- The new driver auto-detects and sets a resolution of 1280 x 768 at a nausea 
inducing 60 HZ. I prefer my CRTs 1152x864 @ 75 Hz. Indeed, 1280 x 768 isn't 
even listed as a mode for that reason. OK so the 'new magic' detects it. I 
discovered that:

Option		"PreferredMode"		"1152x864"

in the monitor section will clear that up...

But then I'm blessed with this in the logs:
(II) RADEON(0): Output VGA-0 using monitor section Generic Monitor
(**) RADEON(0): Option "PreferredMode" "1152x864"

(II) RADEON(0): RandR 1.2 enabled, ignore the following RandR disabled 
(WW) RADEON(0): Option "PreferredMode" is not used

The above is snipped from the same log file... The warning is misleading-- it 
sets the resolution in PreferredMode because if I remove the line back to 
1280 I go.

With the above caveats on the 6.7.x driver, the 6.6.19x driver has worked 
comparatively well so far.


PS- For what it's worth, I've had these problems with the 6.7.192 driver vs. 
the new xorg in experimental, and with a source rebuild of the driver deb vs. 
the newest non-experimental xorg version. So it's definitely isolated to the 
ati driver.

Reply to: