Subject: Re: APM and suspend mode
To: Jonas Oberg <jonas@gnu.org>
From: David Laight <david@l8s.co.uk>
List: port-hpcarm
Date: 09/16/2002 09:56:05
On Mon, Sep 16, 2002 at 08:43:15AM +0200, Jonas Oberg wrote:
> manu@netbsd.org (Emmanuel Dreyfus) writes:
> 
>     This manual?
>     http://developer.intel.com/design/strong/manuals/278240.htm
> 
> Yes. Page 101 and forward, if I remember correctly.
>     
>     How? If it's in ROM, I don't see how we'll be able to manage that.
>     That's bad...
> 
> The CPU executes the standard Windows ROM. That code looks at certain
> registers to find out why it booted and then perform the appropriate
> thing. It would be most helpful if someone could dump the ROM contents
> and disassemble the boot sequence so we can see what it actually does.


Getting into (and out of) sleep correctly is somewhat tiresome.
(I did it for a SA1100/SA1101 board).  As well as the obvious
things all the IO lines have to be set the the correct state
(typically 0v - even if they are active low) in order to avoid
back-powering parts of the system.

Since all the IO is powered down it all needs re-initialising,
also you need to ensure that no device driver code is active
(even if actually sleeping) when you go to sleep.

Powering down the pcmcia slot is a useful power save.
Some of the network cards tuen the system into a nice hand warmer!

	David

-- 
David Laight: david@l8s.co.uk