Re: Manual control of LVDS pixel formats and dual channel mode

From: Mike Isely <isely@isely.net>
To: Chris Wilson <chris@chris-wilson.co.uk>
Cc: intel-gfx@lists.freedesktop.org
Subject: Re: Manual control of LVDS pixel formats and dual channel mode
Date: Wed, 13 Apr 2011 02:27:18 -0500 (CDT)	[thread overview]
Message-ID: <alpine.DEB.1.10.1104130215590.5473@ivanova.isely.net> (raw)
In-Reply-To: <1302649260-11996-1-git-send-email-chris@chris-wilson.co.uk>

On Wed, 13 Apr 2011, Chris Wilson wrote:

> Mike, I wasn't keen on introducing this option to disable the LVDS fixed
> mode. I think we want to allow the user to specify the fixed mode to use
> instead, without the intermediate stop-gap solution of completely disabling
> the probe. Later we can extend that ability (if possible or still desirable
> on reflection) so that we can specify all the LVDS configuration options
> in the single string.

Chris:

Keep in mind that if the probe action is returning wrong results, that's 
worse than no results.  This isn't a question about "supplying missing 
data" so much as it a case of "overriding bad data".  Without that fixed 
mode disabling ability, it is very possible that the panel will simply 
be set up wrong with a corrupted display, with no possible means for the 
user to repair the problem.  That's what led me to make the change, and 
that same change had been living peacefully in the userspace xorg 
driver, up until the point that UMS had been removed from it.  Right now 
without the ability to disable fixed mode there simply *is no* 
workaround.

Now with that said I understand where you're driving at here and if we 
can get a means to actually specify an independent fixed mode solution 
then that clearly also solves the problem and I agree it's more elegant.  
My only worry is the fact that the more elegant solution is also more 
complex may result in there being no solution implemented, which is even 
worse than the other two choices (disable fixed mode vs specifying fixed 
mode).  I guess I just want to make sure that if the ability to disable 
fixed mode is not merged that there at least be continuous progress 
towards being able to set the fixed mode.  And I'm willing to do 
whatever I can to help code and/or test there.

> 
> But first can you check that I've correctly forward ported your other
> patches on top of Eric's refactoring.

Thanks.  Will do.  Stay tuned.

  -Mike

-- 

Mike Isely
isely @ isely (dot) net
PGP: 03 54 43 4D 75 E5 CC 92 71 16 01 E2 B5 F5 C1 E8