Re: [PATCH net-next 08/14] net: pse-pd: Split ethtool_get_status into multiple callbacks

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

On Tue, 7 Jan 2025 17:15:54 -0800
Jakub Kicinski <kuba@kernel.org> wrote:

> On Sat, 04 Jan 2025 23:27:33 +0100 Kory Maincent wrote:
> > diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h
> > index f711bfd75c4d..2bdf7e72ee50 100644
> > --- a/include/linux/ethtool.h
> > +++ b/include/linux/ethtool.h
> > @@ -1323,4 +1323,40 @@ struct ethtool_c33_pse_pw_limit_range {
> >  	u32 min;
> >  	u32 max;
> >  };
> > +
> > +/**
> > + * struct ethtool_pse_control_status - PSE control/channel status.
> > + *
> > + * @podl_admin_state: operational state of the PoDL PSE
> > + *	functions. IEEE 802.3-2018 30.15.1.1.2 aPoDLPSEAdminState
> > + * @podl_pw_status: power detection status of the PoDL PSE.
> > + *	IEEE 802.3-2018 30.15.1.1.3 aPoDLPSEPowerDetectionStatus:
> > + * @c33_admin_state: operational state of the PSE
> > + *	functions. IEEE 802.3-2022 30.9.1.1.2 aPSEAdminState
> > + * @c33_pw_status: power detection status of the PSE.
> > + *	IEEE 802.3-2022 30.9.1.1.5 aPSEPowerDetectionStatus:
> > + * @c33_pw_class: detected class of a powered PD
> > + *	IEEE 802.3-2022 30.9.1.1.8 aPSEPowerClassification
> > + * @c33_actual_pw: power currently delivered by the PSE in mW
> > + *	IEEE 802.3-2022 30.9.1.1.23 aPSEActualPower
> > + * @c33_ext_state_info: extended state information of the PSE
> > + * @c33_avail_pw_limit: available power limit of the PSE in mW
> > + *	IEEE 802.3-2022 145.2.5.4 pse_avail_pwr
> > + * @c33_pw_limit_ranges: supported power limit configuration range. The
> > driver
> > + *	is in charge of the memory allocation
> > + * @c33_pw_limit_nb_ranges: number of supported power limit configuration
> > + *	ranges
> > + */  
> 
> Is there a reason this is defined in ethtool.h?

I moved in to ethtool because the PSE drivers does not need it anymore.
I can keep it in pse.h.

> I have a weak preference towards keeping it in pse-pd/pse.h
> since touching ethtool.h rebuilds bulk of networking code.
> From that perspective it's also suboptimal that pse-pd/pse.h
> pulls in ethtool.h.

Do you prefer the other way around, ethtool.h pulls in pse.h?
Several structure are used in ethtool, PSE core and even drivers at the same
time so I don't have much choice. Or, is it preferable to add a new header?

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