Does FBSD support 'software suspend' ?

[Bruno Ducrot]

On Sat, Sep 10, 2005 at 05:47:15PM +0530, Pranav Peshwe wrote:
> >AFAIK "software suspend" does not equal to "suspend to RAM".
> >There are two different "suspend to disk" modes. One is "software suspend"
> >the other "BIOS suspend". The second needs help by the BIOS and a special
> >formated slice, the first doesn't.
> >FreeBSD supports "BIOS supend" if the hardware supports it,
> >you can tell by checking hw.acpi.s4bios with sysctl.
> >GNU/Linux supports software suspend for S4 as well, ATM FreeBSD doesn't.
> 
> >Fabian
> 
> Thanks for the info.
> 
> So, what i surmise is that : the 'suspend to disk' feature in BSD is
> currently  h/w dependent i.e  BSD does not have a 'software suspend'
> feature like in linux provided by suspend2 (http://www.suspend2.net).
> 
> How useful is this feature(s/w suspend) , if implemented ?

It's architecture independant.  For example, under Linux, software
suspend work under amd64, which can't use APM nor APCI S4Bios.  I've
heard it work under apple computers too, though I never tested myself.

> I am planning to
> take up its development as my final year project.I have not decided for
> final to take it up;any suggestions/remarks regarding its use and feasiblity
> would be very very valuable in helping me to take the right decision.

Well, you should also look at those I think:
- DragonflyBSD, especially their implementation of process
  checkpointing (I must admit I've not looked at this too much
  though);
- implementation of system suspend by system checkpointing under
  OpenSolaris.

In case of OpenSolaris, it seems it is based upon software suspend
under Linux, but to my eyes they have done a much cleaner
implementation IMHO.

BTW the licence of OpenSolaris do not allow direct integration to
FreeBSD IIRC, but that give some insight about a possible design
under our OS.

Cheers,

-- 
Bruno Ducrot

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.