[lm-sensors] [PATCH v3] hwmon driver for ACPI 4.0 power meters

Darrick J. Wong djwong at us.ibm.com
Wed Sep 9 20:20:51 CEST 2009


On Wed, Sep 09, 2009 at 10:06:38AM +0200, Pavel Machek wrote:

> > +Description
> > +-----------
> > +
> > +This driver implements sensor reading support for the power meters exposed in
> > +the ACPI 4.0 spec (Chapter 10.4).  These devices have a simple set of
> > +features--a power meter that returns average power use over a configurable
> > +interval, an optional capping mechanism, and a couple of trip points.
> 
> So where do we get the average power?

/sys/class/hwmon/hwmonX/device/powerX_average.  Most of the hwmon driver
documentation assumes the reader has read the contents of
Documentation/hwmon/sysfs-interface, but I suppose it wouldn't be too
difficult to call that out explicitly.

(Actually I think you're really supposed to use libsensors and not access sysfs
directly.)

> > +Special Features
> > +----------------
> > +
> > +The power[1-*]_is_battery knob indicates if the power supply is a battery.
> > +Both power[1-*]_average_{min,max} must be set before the trip
> > points will work.
> 
> How do trip points work? It is not explained here.

Will add that too.

> > +config ACPI_POWER_METER
> > +	tristate "ACPI 4.0 power meter"
> > +	depends on HWMON
> > +	default m
> 
> default M is "evil".

I'll change it to Y, but I wonder why you think default M is evil...?

> > +	  devices.  Say Y (or M) if you have an Intel or AMD computer with
> > +	  a power meter.
> 
> WTF? Neither Intel nor AMD makes _computers_.

I'll change that to "a computer with ACPI firmware."

> > +	  To compile this driver as a module, choose M here:
> > +	  the module will be called power-meter.
> 
> .ko.

The other config options in drivers/acpi/Kconfig omit the .ko; I was trying to
be consistent with them.

> > +#define POWER_CAP_NAME		"power1_cap"
> 
> I did not see power cap explained.

I'll add that, thanks for the continued review!

--D




More information about the lm-sensors mailing list