Re: [PATCH RFC net-next v2 15/18] net: pse-pd: Add support for getting and setting port priority

[Oleksij Rempel]

On Fri, Nov 01, 2024 at 09:31:43AM +0100, Oleksij Rempel wrote:
> On Thu, Oct 31, 2024 at 12:11:04PM +0100, Kory Maincent wrote:
> > On Thu, 31 Oct 2024 07:54:08 +0100
> > Oleksij Rempel <o.rempel@xxxxxxxxxxxxxx> wrote:
> > 
> > > > diff --git a/include/uapi/linux/ethtool.h b/include/uapi/linux/ethtool.h
> > > > index a1ad257b1ec1..22664b1ea4a2 100644
> > > > --- a/include/uapi/linux/ethtool.h
> > > > +++ b/include/uapi/linux/ethtool.h
> > > > @@ -1002,11 +1002,35 @@ enum ethtool_c33_pse_pw_d_status {
> > > >   * enum ethtool_c33_pse_events - event list of the C33 PSE controller.
> > > >   * @ETHTOOL_C33_PSE_EVENT_OVER_CURRENT: PSE output current is too high.
> > > >   * @ETHTOOL_C33_PSE_EVENT_OVER_TEMP: PSE in over temperature state.
> > > > + * @ETHTOOL_C33_PSE_EVENT_CONNECTED: PD detected on the PSE.
> > > > + * @ETHTOOL_C33_PSE_EVENT_DISCONNECTED: PD has been disconnected on the
> > > > PSE.
> > > > + * @ETHTOOL_C33_PSE_EVENT_PORT_PRIO_STATIC_ERROR: PSE faced an error in
> > > > static
> > > > + *	port priority management mode.
> > > >   */
> > > >  
> > > >  enum ethtool_c33_pse_events {
> > > > -	ETHTOOL_C33_PSE_EVENT_OVER_CURRENT =	1 << 0,
> > > > -	ETHTOOL_C33_PSE_EVENT_OVER_TEMP =	1 << 1,
> > > > +	ETHTOOL_C33_PSE_EVENT_OVER_CURRENT =		1 << 0,
> > > > +	ETHTOOL_C33_PSE_EVENT_OVER_TEMP =		1 << 1,
> > > > +	ETHTOOL_C33_PSE_EVENT_CONNECTED =		1 << 2,
> > > > +	ETHTOOL_C33_PSE_EVENT_DISCONNECTED =		1 << 3,
> > > > +	ETHTOOL_C33_PSE_EVENT_PORT_PRIO_STATIC_ERROR =	1 << 4,
> > > > +};  
> > > 
> > > Same here, priority concept is not part of the spec, so the C33 prefix
> > > should be removed.
> > 
> > Ack. So we assume PoDL could have the same interruption events.
> > 
> > > > +/**
> > > > + * enum pse_port_prio_modes - PSE port priority modes.
> > > > + * @ETHTOOL_PSE_PORT_PRIO_DISABLED: Port priority disabled.
> > > > + * @ETHTOOL_PSE_PORT_PRIO_STATIC: PSE static port priority. Port priority
> > > > + *	based on the power requested during PD classification. This mode
> > > > + *	is managed by the PSE core.
> > > > + * @ETHTOOL_PSE_PORT_PRIO_DYNAMIC: PSE dynamic port priority. Port priority
> > > > + *	based on the current consumption per ports compared to the total
> > > > + *	power budget. This mode is managed by the PSE controller.
> > > > + */  
> 
> After thinking about it more overnight, I wanted to revisit the idea of having
> a priority strategy per port. Right now, if one port is set to static or
> dynamic mode, all disabled ports seem to have to follow it somehow too. This
> makes it feel like we should have a strategy for the whole power domain, not
> just for each port.
> 
> I'm having trouble imagining how a per-port priority strategy would work in
> this setup.
> 
> Another point that came to mind is that we might have two different components
> here, and we need to keep these two parts separate in follow-up discussions:
> 
> - **Budget Evaluation Strategy**: The static approach seems straightforward—if
> a class requests more than available, appropriate actions are taken. However,
> the dynamic approach has more complexity, such as determining the threshold,
> how long violations can be tolerated, and whether a safety margin should be
> maintained before exceeding maximum load.
> 
> - **Disconnection Policy**: Once a budget violation is detected, this decides
> how to react, like which ports should be disconnected and in what order.
> 
> Would it make more sense to have a unified strategy for power domains, where we
> apply the same budget evaluation mode (static or dynamic) and disconnection
> policy to all ports in that domain? This could make the configuration simpler
> and the power management more predictable.

Except of user reports, do we have documented confirmation about dynamic
Budget Evaluation Strategy in PD692x0 firmware?

Do this configuration bits are what I called Budget Evaluation Strategy?
Version 3.55:
Bits [3..0]—BT port PM mode
0x0: The port power that is used for power management purposes is
     dynamic (Iport x Vmain).
0x1: The port power that is used for power management purposes is port
     TPPL_BT.
0x2: The port power that is used for power management purposes is
     dynamic for non LLDP/CDP/Autoclass ports and TPPL_BT for
     LLDP/CDP/Autoclass ports.

-- 
Pengutronix e.K.                           |                             |
Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |