[Tfug] APM mechanisms

johngalt johngalt1 at uswest.net
Thu Dec 27 20:51:46 MST 2007


----- Original Message ----- 
From: "Bexley Hall" <bexley401>
To: <tfug at tfug.org>
Sent: Wednesday, December 26, 2007 4:52 PM
Subject: [Tfug] APM mechanisms


> Hi,
>
> I have a Compaq Presario 3020 (?) that I am
> currently using as my DNS/TFTP/SMTP/whatever
> server.  It's a Pentium 166 (?) class machine.
> Nothing spectacular.  Except, it has two slots
> which lets me cram a dual NIC and a SCSI HBA
> into it (the dual NIC lets it act as my
> firewall and the HBA lets me attach external
> storage as well as tape backup, etc. as needed).

A proper identification of the laptop might be helpful to better define the 
problem.

If it is a Pentium 166 era box, its original OS would probably be win95, 
which wasn't terribly clever about anything. So, power monitoring for that 
type of box may not offer too much to work with.

However, a brief search on Hpaq's web site shows two types of presario 3020 
laptops; one of which has a 1.6 Ghz processor and the other has a 3Ghz 
processor For both of these, their native hardware power monitoring support 
is probably very good.

Regardless, if it gets to be more trouble than it's worth, couldn't you just 
switch the display from LCD to external VGA connector by some keyboard 
Fn+keypress manoeuvre?

FWIW

>
> But, the *biggest* win is an attached LCD display
> which saves me the hassle of having to make room
> for a monitor *or* run it headless and have to
> drag out a monitor when/if it dies, etc.
>
> The box runs NetBSD 3.1 (quite nicely).  *But*,
> the damn LCD backlight is *always* on!  :<
>
> So, with all this as background information,
> my question is:
>
> How does APM work on these boxes?
>
> I would assume it is implemented in the BIOS
> (since the mechanics of it would obviously vary
> from machine to machine).
>
> But, how/when/why does the kernel talk to that
> part of the BIOS.  I.e., how does it decide the
> machine should go into standby?  I mean, the
> machine is never truly *idle*.  E.g., cron
> ticks off every few minutes, there are frequent
> DNS requests made of it, etc.
>
> And, is the APM API fat enough that it includes
> hooks for various "devices" so the caller
> (i.e. the kernel) can specify *what* to "wind
> down"?  (e.g., "put disk into standby mode",
> "put display into standby mode", "put ... ").
> Or, does the BIOS author decide what "standby
> mode" entails and the kernel can just say
> "invoke THE standby mode" or "invoke THE
> sleep mode", etc.?
>
> <shrug>  I want to get a feel for how it
> *should* be working before I decide if it is
> "broken" (misconfigured) or not.
>
> N.B. It may be easier for me to just put a switch
> in series with the CCFL's than to deal with
> sorting out stuff in the BIOS, etc.
>
> (Note that I haven't found anything in the
> machine's "setup" that would pertain to this)
>
> Thx,
> --don
>
>
> 
> ____________________________________________________________________________________
> Never miss a thing.  Make Yahoo your home page.
> http://www.yahoo.com/r/hs
>
> _______________________________________________
> Tucson Free Unix Group - tfug at tfug.org
> Subscription Options:
> http://www.tfug.org/mailman/listinfo/tfug_tfug.org
> 





More information about the tfug mailing list