From: =?iso-8859-1?Q?Niklas_H=F6glund?= <niklas@canit.se>
  To  : ggi-develop@eskimo.com
  Date: Thu, 19 Aug 1999 14:29:09 +0200

Re: VC Switching

On Thu, Aug 19, 1999 at 12:16:00AM -0700, Michael Gersten wrote:
> >But, you have good arguments in favor of a sane vc-switching system.
> >They seem to reinforce the point that the graphics card needs
> >to be viewed as a resource server, serving up frames, sprites, 
> >textures, and whatnot, and not seen as a single resource; except for 
> >the actual focus (of which there is only one per monitor attached 
> >to the card, so those are single resources.)
> 
> Alas, the card is not just a collection of frames, sprites, etc.
> 
> Andy pointed out that it is also a single set of registers,
> whose programming is order specific, chip specific, and command-
> non-atomic.
> 
> That kills making it a sharable resource as you suggest, unless
> everyone is using the same graphics mode.

Not nesseccarily. If all accels are performed by the kernel (as is the
case when using ping-pong buffers), the kernel knows what state the
kernel is in. Wouldn't this allow the kernel to force a VT-switch
(possibly discarding the framebuffer contents (but not the card
state), if the application doesn't save it within 0.5 seconds.

-- 
						Niklas