Re: CVS commit: src/sys/dev/wscons

[Joerg Sonnenberger <joerg%britannica.bec.de@localhost>]

On Tue, Jan 15, 2008 at 07:18:49PM +0100, Matthias Drochner wrote:
> [I/O devices]
> joerg%britannica.bec.de@localhost said:
> > Not from the display driver path. The X server still keeps permissions
> > for them. It just has to stop direct access to VGA registers./ 
> 
> I'm not talking about VGA registers. Let's take a mouse -- the
> X server closes the mouse device when it switches away from
> its screen, and it opens it when it switches back.
> It might well do some probing and setup on open. If the serial
> port or USB is not ready at that point this can fail.

Well, if it is using the infrastructure provided, that doesn't matter
either. Mouse and keyboard are part of wsdisplay as well.

> A terminal switch in the middle of the suspend/resume is
> of limited value. If the system dies before/after it, one loses.

The current breakage I see is either much latter or a rejected suspend.
The rejected suspend is not critical. Nothing forces you to not switch
earlier, but the kernel can and should do something sane.

This is strongly related to being able to use DDB even when X11 is
running.

Joerg