[cpufreq] powernow drivers
ducrot at poupinou.org
Fri Mar 11 05:37:41 PST 2005
On Thu, Mar 10, 2005 at 12:34:01PM -0800, Nate Lawson wrote:
> Bruno Ducrot wrote:
> >The powernow stuff is nearly complete but need testings and more
> >cleanup, especially the 2 drivers shoud be merged.
> >I'm providing them right now at
> >before the merge because I maybe offline for one or two weeks in
> >a few days.
> >The powernow k6 is not finished, but mainly because I dont have the
> >hardware. It would be fine though that I know for sure that someone
> >test it before I make it public, so if you have a cyrix K6/II or III
> >with powernow! capability and you want to test please write me.
> Excellent, your work looks really nice. Just a few comments:
> Please remove my name from the copyright. You did all the work and I
> just provided suggestions.
> Both drivers should just add a child called "powernow" instead of
> k*_powernow. You can't have k7 and k8 powernow on the same system,
> right? They can announce what hardware they are via device_set_desc()
> which it appears you're already doing. Also, DRIVER_MODULE for them
> should not have "pn8" as the first arg.
I got a message like "cpu/pn already exist" or so if I do that. That's
why I have to do a pn7/pn8 distinction for the DRIVER_MODULE.
> You can remove the check for a freq >= 0xffff. I committed this to
> acpi_perf already.
> Have you tested the ACPI attachment for this? Does it work ok? I'm
> just curious because I don't have hw to test acpi-assisted probing.
This work like a charm, or else I would have send to you a complaint
> It looks like these can be merged pretty easily. Once you've done that,
> I'd like to import this driver.
I'm not sure if I will have time right now to do so. I will do this
this week-end before my vacation.
It's also why I have send this email to acpi- before that merge so
that people can atleast have those drivers to play with.
-- Which is worse: ignorance or apathy?
-- Don't know. Don't care.
More information about the freebsd-acpi