[Openchrome-devel] Fix modeline list generation for tv outputs

[Harry de Boer]

Some good news: I got the the basic 720x576 PAL mode working again.

It turns out the list of modelines in DisplayModePtr now has a ->next 
element that is used to go to the next item on the list. These next 
items were all NULL so there was only one modeline recognised by the server.

I created a helper function that creates DisplayModePtrs with next 
elements from a DisplayModeRec[] much like the xorg edid code does. As 
this affects all tv encoders I applied it throughout the codebase.

With this fix it was possible to select the 720x756 mode again which 
works for me (as it did in the past). Other modes are still a mess 
(including 720x756Over which also used to work). I'll look into the 
other modes next.

Patches so far are attached.

Cheers,
Harry

On 26-11-12 15:30, Harry de Boer wrote:
> Hi,
>
> I have made some progress:
>
> - I updated the output sensing patch to support the VT1625S. The power 
> register has the same layout as the status register. A 0 is always for 
> the DACs that are not present, so writing all 1-bits and reading back 
> the register results in a mask of which DACs are actually present.
>
> - The number of tv registers on the VT1625 is 0x82, not 0x6C. This 
> should not matter for anything currently in place.
>
> - The tv screen is lighting up now. Apparently 
> xf86OutputPtr->possible_crtcs should have a value of 1 or higher for 
> it to be a valid configuration.
>
> Don't get excited yet, the output is a complete mess. It seems the 
> driver only reports one modeline, 640x480, which I believe has never 
> worked for me anyway. Only 720x576 worked. I guess that is because the 
> some of the registers are not set correctly and 720x576 works because 
> it is the default set by the BIOS (for PAL).
>
> But first I'll try to get the driver to report all supported 
> modelines. Attached are the patches of what I have so far.
>
> On 23-11-12 00:40, Willem van Asperen wrote:
>> I've applied Harry's patch and it now correctly senses the TV-Out, so 
>> good work Harry!
>
> I kind of expected it to work for you since we seem to have exactly 
> the same board :-)
>
> Cheers,
> Harry
>
>
> _______________________________________________
> Openchrome-devel mailing list
> Openchrome-devel at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/openchrome-devel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/openchrome-devel/attachments/20121126/19bcc5ba/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-Fix-VT1625-output-sensing.patch
Type: text/x-patch
Size: 2586 bytes
Desc: not available
URL: <http://lists.freedesktop.org/archives/openchrome-devel/attachments/20121126/19bcc5ba/attachment-0004.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0002-VT1625-register-count-is-0x82.patch
Type: text/x-patch
Size: 766 bytes
Desc: not available
URL: <http://lists.freedesktop.org/archives/openchrome-devel/attachments/20121126/19bcc5ba/attachment-0005.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0003-There-is-a-possible-crtc-for-TV-1.patch
Type: text/x-patch
Size: 1401 bytes
Desc: not available
URL: <http://lists.freedesktop.org/archives/openchrome-devel/attachments/20121126/19bcc5ba/attachment-0006.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0004-Create-valid-modeline-lists-for-tv-outputs.patch
Type: text/x-patch
Size: 11253 bytes
Desc: not available
URL: <http://lists.freedesktop.org/archives/openchrome-devel/attachments/20121126/19bcc5ba/attachment-0007.bin>