Re: [Linux-fbdev-devel] [PATCH] nvidiafb: allow ignoring EDID info

From: Luca Tettamanti <kronos@people.it>
To: linux-fbdev-devel@lists.sourceforge.net
Cc: Andrew Morton <akpm@osdl.org>, Dave Airlie <airlied@gmail.com>,
	Giuseppe Bilotta <giuseppe.bilotta@gmail.com>,
	linux-kernel@vger.kernel.org
Subject: Re: [Linux-fbdev-devel] [PATCH] nvidiafb: allow ignoring EDID info
Date: Sun, 4 Feb 2007 21:17:13 +0100	[thread overview]
Message-ID: <20070204201713.GA12867@dreamland.darkstar.lan> (raw)
In-Reply-To: <20070130203301.GA15067@dreamland.darkstar.lan>

Il Tue, Jan 30, 2007 at 09:33:01PM +0100, Luca Tettamanti ha scritto: 
> Il Mon, Jan 29, 2007 at 03:37:22PM +0100, Giuseppe Bilotta ha scritto: 
> > On 1/29/07, Dave Airlie <airlied@gmail.com> wrote:
> > > > > > Because most users won't even be aware of the module option: they'll just
> > > > > > know that their card doesn't work right.
> > > > >
> > > > > This isn't a card problem this is a monitor problem, the card just
> > > > > passes through the edid data from the monitor... or else the
> > > > > programming of the card registers from edid is wrong..
> > > >
> > > > In which case the same problem would occur with different video cards, so
> > > > this patch should be some generic thing, available to all drivers, no?
> > >
> > > It should be in the fb layer not card specific.. as it may happen on any card...
> > 
> > Although solving the problem at the fb layer level is probably the
> > correct/best way to do it, I am not aware of people having problems
> > with broken EDIDs and non-nVidia cards. By contrast, workarounds for
> > nVidia and broken EDIDs is a very common thing to do even on Windows.
> 
> Not all drivers have support for reading EDID information, so in many
> cases it just "works". My old monitor had garbage in the EDID eeprom and
> I had to pass "ignore_edid" to radeonfb (before I reprogrammed the
> eeprom that is :P).
> 
> > Plus, providing an fb-layer-level solution is way beyond my kernel
> > programming experience :)
> 
> I'll try to code something in the weekend.

I'm somewhat stuck. fbmon has the library functions to parse and
optionally fix EDID block, but it lacks the knowledge of the source of
the data; IOW it's not possible to ignore the EDID coming from a certain
head, the flag would be global (which may be less that optimal with
multi-head setups).
So instead of adding an option I'd use the existing broken EDID db,
adding another state (say 'unfixable') which causes the data to be
discarded. Comments?

Giuseppe, can you send me the EDID block of your monitor?

Luca
-- 
Sbagliare ? umano, ma per incasinare davvero le cose serve un computer.