Re: [RFC] hwmon: (tps23861) PSE-PD support, extend hwmon driver or move to pse-pd?

[Kory Maincent <kory.maincent@bootlin.com>]

On 11/2/26 18:33, Andrew Lunn wrote:
> On Mon, Nov 02, 2026 at 05:57:06PM +0100, Kory Maincent wrote:
>>
>>
>> On 11/2/26 17:36, Andrew Lunn wrote:
>>>> I am not using at all the hwmon framework in that development.
>>>> But as the binding tps23861 is already reserved for hwmon I use a new
>>>> "tps23861-poe" compatible name binding.
>>>> To me, I agree to remove the hwmon driver but this will break backward
>>>> compatibility. I am not a hwmon maintainer so it's not my decision to take.
>>>
>>> Is the hwmon code applicable to the tps238XX range of devices?
>>>
>>> Can that hwmon code by merged into the PoE driver? Without API
>>> changes, apart from dropping the API abuse case? The compatible can
>>> then be transplanted into the PoE driver?
>>
>> Should we really want to add the hwmon framework in the driver only to keep
>> the binding name?
> 
> That is why i asked if the code is applicable. If it is using hwmon to
> show the temperature of the PoE controller, it probably is
> applicable.
> 
> If it is just repeating the same information which is available in
> ethtool, then it is not so useful. If it is not useful, the hwmon
> driver should be marked as deprecated, and a pointer added to the real
> PoE driver.

Oh ok. The information reported by the hwmon driver that are not in the 
PSE driver are:
- current and voltage while PSE is reporting the power
- the temperature
- the port resistance

Not sure if they are useful. For the temperature the PSE raise a 
notification in case of over temperature.
Anyway if we want to report these information I think we will report 
them through ethtool not in sysfs. We can later add a new ops to report 
the temperature if people want it.

Regards,
-- 
Köry Maincent, Bootlin
Embedded Linux and kernel engineering
https://bootlin.com