> >> Ports 1 and 2 should hopefully be protected by the > >> invalid_port_mask. It should not even be possible to create those > >> ports. port 0 is interesting, and possibly currently broken on > >> 6393. Please take a look at that. > > > > Why would port 0 be broken on 6393x ? > By "broken", I guess Andrew means that if we feed port 0 to > mv88e6xxx_phy_is_internal, it will return true, which is wrong since there is no > internal phy for port 0 on 6393X ? Yes, that is what i was thinking. But i did not spend the time to look at the code see if this is actually true. There might be a special case somewhere in the code. But in general, we try to avoid special cases, and add device specific ops. Andrew
