The 12/17/2021 18:12, Vladimir Oltean wrote:
>
> On Fri, Dec 17, 2021 at 04:53:53PM +0100, Horatiu Vultur wrote:
> > Extend lan966x driver with fdb support by implementing the switchdev
> > calls SWITCHDEV_FDB_ADD_TO_DEVICE and SWITCHDEV_FDB_DEL_TO_DEVICE.
> >
> > Signed-off-by: Horatiu Vultur <horatiu.vultur@xxxxxxxxxxxxx>
> > ---
>
> Looks pretty good. Just one question, since I can't figure this out by
> looking at the code. Is the CPU port in the unknown unicast flood mask
> currently?
It is not. Because under a bridge can be only lan966x ports so the
HW will do already the flooding of the frames.
>
> > .../net/ethernet/microchip/lan966x/Makefile | 2 +-
> > .../ethernet/microchip/lan966x/lan966x_fdb.c | 244 ++++++++++++++++++
> > .../ethernet/microchip/lan966x/lan966x_main.c | 5 +
> > .../ethernet/microchip/lan966x/lan966x_main.h | 14 +
> > .../microchip/lan966x/lan966x_switchdev.c | 21 ++
> > .../ethernet/microchip/lan966x/lan966x_vlan.c | 15 +-
> > 6 files changed, 298 insertions(+), 3 deletions(-)
> > create mode 100644 drivers/net/ethernet/microchip/lan966x/lan966x_fdb.c
> (...)
> > +static void lan966x_fdb_add_entry(struct lan966x *lan966x,
> > + struct switchdev_notifier_fdb_info *fdb_info)
> > +{
> > + struct lan966x_fdb_entry *fdb_entry;
> > +
> > + fdb_entry = lan966x_fdb_find_entry(lan966x, fdb_info);
> > + if (fdb_entry) {
> > + fdb_entry->references++;
> > + return;
> > + }
> > +
> > + fdb_entry = kzalloc(sizeof(*fdb_entry), GFP_KERNEL);
> > + if (!fdb_entry)
> > + return;
> > +
> > + memcpy(fdb_entry->mac, fdb_info->addr, ETH_ALEN);
>
> ether_addr_copy
>
> > + fdb_entry->vid = fdb_info->vid;
> > + fdb_entry->references = 1;
> > + list_add_tail(&fdb_entry->list, &lan966x->fdb_entries);
> > +}
--
/Horatiu
