Re: [RFC 7/8] hwmon: Altera Arria10 System Resource Chip - HW Monitor

[Guenter Roeck <linux@roeck-us.net>]

On Tue, Mar 29, 2016 at 04:43:41PM -0500, Thor Thayer wrote:
> 
> 
> On 03/29/2016 03:16 PM, Guenter Roeck wrote:
> >On Tue, Mar 29, 2016 at 02:13:10PM -0500, tthayer@opensource.altera.com wrote:
> >>From: Thor Thayer <tthayer@opensource.altera.com>
> >>
> >>This patch adds the hwmon functionality to the Arria10 System
> >>Resource Chip. The hwmon encapsulates the PCIe Enable, USB Enable,
> >>and all the Power Good signals on the System Controller.
> >>
> >
> >I may be completely wrong, but a glance through the driver suggests
> >that, if anything, this should be a regulator driver, not a hwmon driver.
> >A hardware monitoring driver would be expected to report the voltages,
> >not (just) the voltage status. Am I missing something ?
> >
> >Please have a look into Documentation/hwmon/sysfs-interface for
> >acceptable hwmon attribute names and their meaning.
> >
> >Thanks,
> >Guenter
> >
> 
> Hi Guenter,
> 
> <adding voltage and current regulator framework moderators>
> 
> Yes, I see your point. In looking at the regulator drivers, I interpret
> those as being controlled by the driver whereas this chip is passively
> reporting status.
> 
> The success/fail indication seemed at first glance to fit the hwmon model. I
> thought the fan indication would be a good analog but even it reports speed
> and not success/fail.
> 
Yes, alarm and/or fault attributes are supposed to be secondary.

> After reading the referenced document, I agree that hwmon probably isn't
> appropriate. However, the regulator doesn't seem appropriate either (the
> only status appears to be tied to battery properties).
> 
Not really sure myself where this would fit if it is just status bits.

Does the chip report anything else besides the status ? One of the attributes
includes "pmbus", so one could conclude that there must be a PMBus compatible
chip somewhere.

Thanks,
Guenter