[ltp] DVI problem again

[Alex Deucher]

On 11/16/05, Martin H=F6fling <martin.hoefling@gmx.de> wrote:
> Am Mittwoch, 16. November 2005 15:43 schrieb Alex Deucher:
>
> > right now the digital outputs in the radeon driver (LVDS and TMDS) do
> > not allow user modelines to prevent users from damaging their panels,
> > however, in some cases, it's advantageous to override the driver's
> > mode as per the bug I mentioned above.
>
> All calculation from various tools gave me a dot clock greater than 165. =
 In
> the analog case, the mode is set properly:
>
> (II) RADEON(0): Supported additional Video Mode:
> (II) RADEON(0): clock: 162.0 MHz   Image Size:  408 x 306 mm
> (II) RADEON(0): h_active: 1600  h_sync: 1664  h_sync_end 1856 h_blank_end=
 2160
> h_border: 0
> (II) RADEON(0): v_active: 1200  v_sync: 1201  v_sync_end 1204 v_blanking:=
 1250
> v_border: 0
> (II) RADEON(0): Ranges: V min: 56  V max: 75 Hz, H min: 30  H max: 81 kHz=
,
> PixClock max 170 MHz
>
> I used a xls sheet in gnumeric to calculate the modelines with reduced
> blanking times and with this modeline, it should be possible to use the t=
ft
> with 60 Hz and a dot clock < 165MHz.
> Is the dot-clock limited in the same way on the analog side? If yes, then
> your information corresponds to the problem, that the analog output is se=
t
> independent compared to the TMDS.

the analog output is driven by one of the DACs (most radeons have 2).=20
the driver allows the DACs to be prgrammed with user defined
modelines.  The TMDS and LVDS controllers do not. However, you can
comment out a line or two in the source (see one of the patches linked
from the bug I mentioned earlier) and the digital outputs will use
user defined modelines.

Alex

>
> Best wishes
>
>         Martin
> --
> Glaubw=FCrdigkeit ist ein soziales Konstrukt, das wesentlich aus
> Anerkennung erw=E4chst. (Andreas Peeck in de.admin.news.groups)
>
>
>