On Wed, May 17, 2023 at 05:49:13PM +0530, Krishna Kurapati PSSNV wrote: > On 5/17/2023 5:14 PM, Johan Hovold wrote: > > On Wed, May 17, 2023 at 04:40:11PM +0530, Krishna Kurapati PSSNV wrote: > >> On 5/16/2023 4:29 PM, Johan Hovold wrote: > >>> On Sun, May 14, 2023 at 11:19:09AM +0530, Krishna Kurapati wrote: > > > >>>> + interrupts: > >>>> + maxItems: 7 > >>>> + interrupt-names: > >>>> + items: > >>>> + - const: dp_hs_phy_irq > >>>> + - const: dm_hs_phy_irq > >>>> + - const: ss_phy_irq > >>> > >>> I assume that these are only for the first port, and that you need to > >>> define these interrupts also for ports 2-4. > > > >> I wanted to add them when wakeup-source is enabled but since you > >> mentioned that these must be added now and driver support can be added > >> later, I will make a patch separately for this in v9. > > > >> Can I use the following notation for the new interrupts ? > >> > >> dp_hs_port2_irq > >> dm_hs_port2_irq > >> dp_hs_port3_irq > >> dm_hs_port3_irq > >> dp_hs_port4_irq > >> dm_hs_port4_irq > >> > >> > >> That way the interrupt names for first port will be same as ones for > >> single port. > > > > For consistency, I'd say: use the same scheme also for port1. Perhaps > > "port" is unnecessary too. > > > > And since these are getting new names, you can drop the redundant "_irq" > > suffix as you did for the power-event lines. > The reason I wanted to mark it as dp_hs_portX_irq is to keep code > changes to driver simple. The existing code to read current IRQ's can > stay as it. Only need to add changes for reading IRQ's of new ports. I understand why you want to do it this way, but again, the devicetree binding is supposed to be hardware description that is independent from any particular implementation. This is also why I said that it may be preferable/easier to just implement wakeup for MP from the start. > > For example: > > > > pwr_event_1 > > dp_hs_phy_1 > > dm_hs_phy_1 > > ss_phy_1 > > ... Johan
