Re: A simple cpufreq(9)

[Jukka Ruohonen <jruohonen%iki.fi@localhost>]

On Fri, Sep 23, 2011 at 08:42:16PM +0200, Joerg Sonnenberger wrote:
> Consider making the "unit" of scaling an additional attribute of the
> list and provide userland with:

Another strong case I forgot to mention is that avoiding lists or explicit
states (as available in recent x86) rules out hardware implementations that
may operate with a single boolean.

There are already ichlpcib(4) and piixpcib(4) that provide no state
information. Presumably the same approach is used in some parts of the
embedded world too. Using a percentage would cover also these cases.

- Jukka.