Re: [Unichrome] xorg and CLE266 : won't activate panel and weird colors

[Luc V. <li...@sk...>]

On Thu, Apr 27, 2006 at 11:03:11PM -0400, Alexis Bilodeau wrote:
> Hi,
> 
> I have two problems with the unichrome driver and xorg 6.9.0 :
> 
> First of all, the driver won't activate my LCD flat panel (with the "no
> panel found" error). If I modify via_mode.c in order to bypass the bios
> settings verification and I force it to activate the panel -dirty hack, I
> know- (won't work with 0.2.5, only with the driver bundled with xorg 6.9.0),
> the panel turns on ok.

Provide me with a log with --logverbose 8

This could be breakage in the work i've been doing with respect to 
output devices and modesetting.

Yes, the output selection is seriously different from what's included in 
X. It's progress, but with no hardware and no reports of breakage, i 
can't fix that what invariably gets gotten broken along the way.

> 
> However, when it does, the colors are all messed up. I saw a couple of posts
> on this list reporting this, but with no solutions.
> 
-ENOHW, i've done things VIA has been unable to do, but i did have the 
hardware to extensively test with. But feel free to remind me of those 
posts.

> I'm willing to help and I have enough programming knowledge to try a couple
> of solutions, but I'm not familiar with the video drivers stuff. So if
> anyone can direct me a little bit, it would be greatly appreciated.
> 
> The colors problem is hard to describe. The screen is - barely- usable, but
> it is. The colors aren't quite inverted, they're, well, mixed up. In 8-bit,
> they seem to be correct, but, 8-bit color, you know.
> 
> I had exactly the same problem with the via driver shipped with Xfree86
> 4.5.0 (I guess that it's about the same driver).

Yes, unlike almost every other part of the driver, the panel code has 
only been shifted around, never fixed or cleaned up. -ENOHW. But 
shifting around is enough to break things.

Provide a log first, we'll take it from there.

Luc Verhaegen.