On 07/30/2012 01:08 PM, Alex Deucher wrote:
On Mon, Jul 30, 2012 at 3:43 PM, Adam Jackson <a...@redhat.com> wrote:
If more than one output is connected, and there's only one ACPI lid
device, proxy the state of the lid into the LVDS / eDP connectivity.
This way we don't suffer from broken lid-status from the firmware in
a way that would result in nothing lighting up.
This is terribly Linux-specific and I don't think we should merge it
as-is, I'm mostly trying to generate some discussion. There's some
weird corner cases that come out of this since we're still doing the
"poke one thing at a time" model, and we don't explicitly turn _off_
outputs during initial setup, so X and kernel's view of the world
gets out of sync.
I think ideally we'd do this in the DRM, but I was unable to come up
with a clean way to do it there. Maybe it just needs to be ugly.
Seems like something X shouldn't care about. Too many ways to do the
wrong thing. Maybe just report always report connected for LVDS/eDP
and let the desktop environment figure out what to light up based on
lid state, saved preferences, etc.
Would reporting lid state through a standardized RandR 1.x property be a
reasonable compromise? That would keep the /proc/acpi / /sys / whatever mucking
about confined to the drivers, or at the very least, the X server.
-- Aaron
_______________________________________________
xorg-devel@lists.x.org: X.Org development
Archives: http://lists.x.org/archives/xorg-devel
Info: http://lists.x.org/mailman/listinfo/xorg-devel
