Bug#564174: xserver-xorg-video-intel: xrandr shows outputs that do not exist on Lenovo

[Julien Cristau <jcristau@debian.org>]

On Tue, Sep 13, 2011 at 10:53:29 -0700, Josh Triplett wrote:

> On Tue, Sep 13, 2011 at 11:11:59AM +0200, Julien Cristau wrote:
> > On Tue, Sep 13, 2011 at 11:03:46 +0200, maximilian attems wrote:
> > 
> > > > On my laptop, a Lenovo X200 with GM45 chipset, xrandr shows many outputs
> > > > that do not exist; more if using KMS.
> > > 
> > > Does this still occur with newer linux images 3.0 and
> > > xserver-xorg-video-intel 2.16 ?
> > > 
> > I expect that it does, and I think this should be treated as a non-bug.
> 
> Older versions of the intel driver contained a substantial table of
> output quirks for various systems, specifically to address this problem.
> This quirk handling regressed in the transition to moving more of the
> logic in the kernel.  I don't know where the bug needs fixing, whether
> in the kernel or in the KMS-enabled bits of the Intel driver, but either
> way I'd really like to see a fix for this regression.
> 
I don't think that's true.  There's a bunch of quirks for "doesn't have
LVDS" and similar, because there's no way to detect connection status
for those connectors and you don't want to light up something that's not
there.  When connection status can be (mostly) reliably detected,
exposing more connectors than are actually available on the specific
machine does not hurt anything, and I don't remember ever seeing quirks
to change that (it's a largely intractable problem, with pretty much no
benefit).

Cheers,
Julien