Huh?  PCI power management is a *standard* defining how power management
capabilities are detected and utilized, when devices will decode, etc.
There is nothing 'magical' about it.  Similar specifications exist for other
bus standards. Why not define a standard set of device power management
interfaces?

-----Original Message-----
From: Chris Torek [mailto:torek@BSDI.COM]
Sent: Monday, February 28, 2000 2:13 AM
To: joda@pdc.kth.se; tech-kern@netbsd.org
Cc: thorpej@netbsd.org
Subject: Re: interface power management


>Doesn't [power management] belong in struct device and/or cfattach?

I would say no, or at least "not yet": power management is by no
means universal yet, and tends to have funny "special consideration"
needs, such as "device XYZ must not be powered up within 5 seconds
of device QRS".  Until some sort of sufficiently general pattern
emerges (in 10 or 15 years? :-) ), I would hold off on attempting
to "genericize" it that far.

Chris