Intel 2.0.0 driver not seeing laptop screen resolutions properly

[Keith Packard]

On Tue, 2007-05-08 at 21:21 -0700, Eric Anholt wrote:

> So, we lose out on starting X after a resume from S3 has happened, on
> that hardware (which, as I understand, is not uncommon)?  That seems
> like a recipe for even more pain than the current situation.

I don't know which is more common -- hardware that unmaps the BIOS on S3
resume or hardware which has bogus parameters in the video registers.
And, at this point, we're at 'failing to ever work' being worse than
'failing to work after S3'.

In any case, we should certainly add config options that can control
which method (video reg vs BIOS) we use for the initial mode selection.

Can we get the driver to pull both the current timing mode *and* the
BIOS mode and then use some logic to select between them? It seems like
we could tell which was more sensible most of the time...

One thing seems clear -- we should update Arjan's BIOS stuff to complain
if the LVDS doesn't offer EDID. That's the only reliable source of
information which doesn't disappear after S3 and which isn't subject to
random BIOS bugs in how the text mode is configured.

-- 
keith.packard at intel.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://lists.x.org/archives/xorg/attachments/20070508/fde3606c/attachment.pgp>