On Wed, Mar 15, 2023 at 02:58:23AM +0100, Andrew Lunn wrote: > On Wed, Mar 15, 2023 at 02:50:00AM +0200, Vladimir Oltean wrote: > > On Tue, Mar 14, 2023 at 11:15:11AM +0100, Christian Marangi wrote: > > > Document support for LEDs node in dsa port. > > > Switch may support different LEDs that can be configured for different > > > operation like blinking on traffic event or port link. > > > > > > Also add some Documentation to describe the difference of these nodes > > > compared to PHY LEDs, since dsa-port LEDs are controllable by the switch > > > regs and the possible intergated PHY doesn't have control on them. > > > > > > Signed-off-by: Christian Marangi <ansuelsmth@xxxxxxxxx> > > > --- > > > .../devicetree/bindings/net/dsa/dsa-port.yaml | 21 +++++++++++++++++++ > > > 1 file changed, 21 insertions(+) > > > > Of all schemas, why did you choose dsa-port.yaml? Why not either something > > hardware specific (qca8k.yaml) or more generic (ethernet-controller.yaml)? > > The binding should be generic. So qca8k.yaml is way to specific. The > Marvell switch should re-use it at some point. > > Looking at the hierarchy, ethernet-controller.yaml would work since > dsa-port includes ethernet-switch-port, which includes > ethernet-controller. > > These are MAC LEDs, and there is no reason why a standalone MAC in a > NIC could not implement such LEDs. So yes, > ethernet-controller.yaml. > > Is there actually anything above ethernet-controller.yaml? This is > about a netdev really, so a wifi MAC, or even a CAN MAC could also use > the binding.... > Yes ideally when we manage to do all the things, ath10k would benefits from this since it does have leds that blink on tx/rx traffic and are specially controlled... Don't think there is something above ethernet-controller tho... -- Ansuel
