suspend/hibernate on desktops

[Dave Jones]

On Mon, Jan 16, 2006 at 11:57:19PM +0000, Andy Burns wrote:
 > Jeremy Katz wrote:
 > 
 > >I think it's worth filing bugs as the framework is definitely viable for
 > >both.  
 > 
 > I already filed BZ 177869  and 177871
 > 
 > >Desktop suspend/resume definitely isn't the focus, but if we can
 > >get it working more, it's a nice side effect ;-)  One thing that's
 > >likely to be more problematic with desktops is BIOS bugs[1] just because
 > >the laptop vendors at least consciously consider suspend/resume.
 > 
 > And there I was thinking lack of batteries, docking stations and 
 > lid-switches might make it simpler ;-)

The biggest problem we have with suspend/resume is making the video
come back on when we resume.  On some laptops, the BIOS
re-POST's the chip itself, so nothing needs to be done.

Due to the modular nature of most desktops, it's unlikely the BIOS
has any clue whatsoever what to do in this case, so we're entirely
reliant upon horrors like re-running the video bios by hand ourselves[*].
Desktop machines with onboard VGA _may_ have BIOSes that do the
relevant magic too, but I haven't seen a significant sample
to say one way or the other.

		Dave

[*] Sadly we have to do this on a lot of laptops too, so if nothing else,
     it increases coverage testing :-)