> I did get apm to function, but it wasn't the way i wanted. I set an
> inactivity timeout in the bios and after that period of time, the cpu
> did indeed downshift. *However*, when the cpu became non-inactive,
> the bios did not speed the cpu up. Apparently, the bios was simply
> monitoring things like key presses to determine inactivity. I would
> like to set it up so that the kernel decides when it is inactive and
> when it is not. That way, when left unattended, the cpu could slow
> down and speed up according to cpu demand (since there is a great deal
> of unattended activity) and not simply user interaction at the
> console.
>
I had the same problem. I enabled APM then started the Byte Unix Benchmarks.
The result was that my cpu "went to sleep" and didn't finished the benchmark
until I pressed a key hours later. The resulting numbers were what you would
expect if linux ran on a Sinclair.
-Mike