improve cx_lowest logic

[Vitaly Magerya]

Andriy Gapon <avg at freebsd.org> wrote:
> I would like to propose the following change for review and testing:
> http://people.freebsd.org/~avg/acpi_cpu_cx_lowest.diff

I just tested this on a laptop that jumps between reporting only
C1, C1 and C2, and C1 through C3; in all cases your patch does
the right thing, so this removes the need for kludges I had to
do earlier; thank you, Andriy.

Will you also change how /etc/rc.d/power_profile selects
lowest_value? It currently uses the maximum from dev.cpu.0.cx_supported,
but with your patch it'll be the right thing to simply use Cmax.