On Tue, May 03, 2005 at 12:38:46PM +1000, Benjamin Herrenschmidt wrote: > On Mon, 2005-05-02 at 10:16 +0200, Bernhard Reiter wrote: > > > Section "Device" > > Identifier "ATI Technologies, Inc. RV350 [Mobility Radeon 9600 > > M10]" > > # Driver "ati" > > Driver "radeon" > > BusID "PCI:0:16:0" > > # Option "UseFBDev" "true" > > #Option "MetaModes" "1280x854-1024x768" > > Option "MetaModes" "1280x854-1280x1024" > > EndSection > > > > Section "Monitor" > > Identifier "Standardbildschirm" > > #HorizSync 28-50 > > #VertRefresh 43-75 > > HorizSync 24-80 > > VertRefresh 50-75 > > Option "DPMS" > > EndSection > > What if you don't put any of the HorizSync/VertRefresh ? Does it manage > to pick them up from DDC or not ?
If I do not put in HorizSync/VertRefresh as below,
it picks it up correctly from DDC.
ection "Monitor"
Identifier "Standardbildschirm"
#HorizSync 24-80
#VertRefresh 50-75
Option "DPMS"
EndSection
> Also, what do you rendering problems look like ?
For reference I put a small image up on
http://intevation.de/~bernhard/ppc/powerbook5%2C6/xorg.renderingdriproblem-20050503.png
> Do they still happen if
> you force DRI not to load ? (Try removing the module
> from /lib/modules/*)
When I uncomment the "dri" module, the rendering problems are also gone.
What is the difference to using NoAccel?
> r300 DRI isn't quite ready yet for public
> consumption but X.org CVS may still try to use it ... (it will also
> cause crashes on wakeup from sleep).
I cannot judge how much changes are going on in X.org CVS,
the CVS checkout I did on 20050414 did not have that rendering problems
with UseFBDev. With the updated CVS build from 20050429,
I could use MergeFB for the first time, but had that rendering problems
with it and UseFBDev.
Now I do have a crash in the server when trying to run magicpoint.
Thus I am trying to build 6.8.2 with your patches and see if this is better.
Bernhard
pgpHP0O8RViDu.pgp
Description: PGP signature
