[ltp] Question on APM vs ACPI

[Nathan Kurz]

On Tue, Dec 07, 2004 at 10:14:31PM -0600, Satish Balay wrote:
> On Tue, 7 Dec 2004, Nathan Kurz wrote:
> 
> > I'm trying to understand this myself.  It seems that throttling is
> > only possible under ACPI, but that frequency scaling is independent of
> > APM/ACPI.  It also seems that scaling has a much larger effect than
> > throttling on power consumption.  My impression is that SpeedStep is
> > the combination of these two.  Is this accurate?
> 
> What is the difference between 'throttling' and 'freq scaling'?
> 
> 'speedstep' is a implemented as kernel module - and is primarily
> controled by a userland daemon (for example cpuspeed). And I don't
> think it has anything to do with APM or ACPI (although - it can
> repsond to some events triggered by APM/ACPI daemons)

Thus you are saying that SpeedStep is synonymous with frequency
scaling?  I agree this is they way it seems to work on Linux, but I'm
not sure if this is how the term is used in general.

> > I also haven't found anyone who has fully working APM suspend
> > support.
> 
> What is your definition for a 'fully working APM suspend'?

I'd call fully working suspend the ability to suspend to ram (S3) and
to disk (S4) both from the command line and based on hardware events
(low battery, lid close).  I'm in the process of configuring a 2.6.9
kernel on a Thinkpad X30, and haven't yet found a way to get both of
these to work consistently.  Occasionally they work, but more often I
hang on resume forcing me to recover with a hold-the-power-key reboot.

I've found many people with earlier kernels having success, but
haven't seen reports of success from people using this kernel.  I now
have ACPI working pretty much the way I would like, but it still feels
messy and complex.  I'd be happy for hints on replacing it with APM.

--nate