1 |
On Monday 14 September 2009 13:59:45 William Kenworthy wrote: |
2 |
> I ran into the same problem though with different versions of the |
3 |
> software a couple of days ago. The only fix that worked was to -hal |
4 |
> xorg-server, and recreate the xorg.conf file that I had previously |
5 |
> deleted, making sure that EDID and DDC were not being used. |
6 |
> |
7 |
> Section "Device" |
8 |
> Identifier "AtiRadeon" |
9 |
> Driver "radeon" |
10 |
> VendorName "ATI" |
11 |
> Option "DPMS" "On" |
12 |
> Option "EnablePageFlip" "1" |
13 |
> Option "RenderAccel" "1" |
14 |
> Option "AGPMode" "4" |
15 |
> Option "IgnoreEDID" "1" |
16 |
> Option "NoDDC" "1" |
17 |
> EndSection |
18 |
> |
19 |
> Not sure all the settings are optimal, but I have a display thats at |
20 |
> least at a usable resolution ... |
21 |
> |
22 |
> It might seem all and good that xorg automaticly chooses the best |
23 |
> resolution - but it clearly doesnt. This is on a system running as |
24 |
> 1600x1200 for years on the same hardware, with xorg suddenly deciding it |
25 |
> can only do 1280x1024 (and even then, it first defaulted to 1024x768). |
26 |
> Whatever happened to the idea that in Linux (and esp gentoo-linux) its |
27 |
> the "user" thats in control :) |
28 |
> |
29 |
> It certainly seems someone - seemingly xorg - dropped the ball |
30 |
> recently :( |
31 |
|
32 |
I notice that panels are almost always detected correctly - they have a single |
33 |
native resolution defined by the number of elements in the display. |
34 |
|
35 |
CRTs are another story - my spare machine can do better 1600x1200 with a CRT |
36 |
and autodetect logic often sets it lower. Anecdotal evidence via my eyeballs |
37 |
tells me it's because it first figures "optimum" frequencies to run at, then |
38 |
picks the best resolution for that frequency. Personally I don't see the need |
39 |
to run that monitor at 85Hz, lower frequencies are just fine for me. |
40 |
|
41 |
Pre-hal Xorg will do what you tell it to. |
42 |
Post-hal Xorg apparently does what a toss of the coin tells it to. |
43 |
|
44 |
-- |
45 |
alan dot mckinnon at gmail dot com |