[Openchrome-users] OpenChrome on an OQO 02

[Xavier Bachelot]

Karim Yaghmour wrote:
> I've been trying to get decent display on an OQO 02 (CX700) for
> a few days now. Here are my basic setup possibilities:
> 1- Office: OQO 02 on docking station w/ DVI-connected 1280x1024 LCD
> 2- Mobile: OQO 02's internal LCD display - 800x480
> 3- Projector: OQO O2 driving VGA (could be 1024x768 or 800x600)
> 
> Basically, I don't really care about 3D or anything of that kind,
> I just want to be able to have decent 2D (browser, OpenOffice,
> cli, emacs, vmware, etc.)
> 
> I've tried a few different approaches:
> - basic vesa driver: I get #2 working pretty easily, but DVI is
>    basically a 1280x1024 rendering of an 800x480 image ... :( Yuck.
> - VIA's drivers: I get #1 working, but it's mutually exclusive w/
>    #2 -- plus I really get the impression that VIA has no idea of
>    how to interface with the FLOSS community. It's a shame, they
>    have the knowledge and there are people who'd love to point them
>    in the right direction ... not to mention that using the source
>    code of a company that has no idea how to package OSS package
>    for distribution scares the !#$# out of me.
> - openchrome (Ubuntu 7.10 source w/ trivial patch from OQO:
>    http://www.oqo.com/unsupported/linux/openchrome-vt3157.patch --
>    instructions here: http://www.oqo.com/unsupported/linux/):
>    same result as basic vesa driver.
> 
> I'm not new to funky hardware, but I would like a few pointers.
> Basically, from what I gather, I should be using xrandr 1.2 drivers.
> Looking into the openchrome branches, there seems to be a randr
> branch with a couple of new files (when diff'ed against trunk.)
> Though, from the comments in the sources I gather randr 1.2 support
> isn't there yet.
> 
> So here are a few questions:
> 1- Can I *safely* use the randr branch on my hardware? I don't mind
>     testing new stuff and reporting back, I just don't want any hardware
>     to break.

Yup, the branch is safe to use as long as you know this is a work in
progress, that's why it is not in trunk.

> 2- I noticed this code snippet in via_driver.c:
> 	case VIA_CX700:
>              /* FIXME:
>               * Only for testing, this code will be removed
>               * when randr 1.2 support is added
>               */
>              pVia->pBIOSInfo->FirstCRTC->IsActive = TRUE ;
>              pVia->UseLegacyModeSwitch = FALSE ;
>     can I replace it with something like this:
> 	case VIA_CX700:
>              /* FIXME:
>               * Only for testing, this code will be removed
>               * when randr 1.2 support is added
>               */
>              pVia->pBIOSInfo->Panel->IsActive = TRUE ;
>              pVia->pBIOSInfo->Lvds->IsActive = TRUE ;
>              pVia->UseLegacyModeSwitch = FALSE ;

Yup, this should be OK.

You probably also have to remove line 1538 from via_driver.c (randr rev. 
501). Gabriel, can you confirm this is all that is needed to enable your 
new panel code on CX700 ?

>     Nothing precludes having 2 versions of the driver with some
>     kind of option menu or boot option (one driver that fits
>     use cases #1 and #2 and one that fits #2 and #3) ... or some
>     dynamically modifiable symlink ... until randr 1.2 support is
>     there ...

The randr branch should fit #2 and #3, there's no support for DVI in any
of the branches yet. The only way to have DVI currently is to rely on
the VGA bios by setting VBEModes to true and this will prevent the
driver from driving the other outputs properly.

> 3- When is randr 1.2 support expected? No pressure, just thought
>     I'd ask in case someone has an idea. Related question: how
>     much work is randr 1.2 support?
> 
Gabriel is working on this, he'll give you more accurate answers.

> Thanks for your help,
> 
> Karim Yaghmour
> Founder and CEO
> Kryptiva inc.
> 
Regards,
Xavier