On Tue, Jan 15, 2008 at 05:36:32PM +0100, Matthias Drochner wrote:
> Actually I'm not sure it is right at all to deal with
> this in the wsdisplay suspend/resume functions: at that
> point, the X server might try to access I/O devices
> which are already/still suspended.

The only device the X server has native access to is the PCI device and
wsdisplay is a child of that. A setup where no text mode VT is present
is IMO broken -- note that this is different from running no tty on it
or so.

Joerg