bridge: switch order of rx_handler reg and upper dev link (0f49579a) · Commits · 戴 / test

The thing is that netdev_master_upper_dev_link calls call_netdevice_notifiers(NETDEV_CHANGEUPPER, dev). That generates rtnl link message and during that, rtnl_link_ops->fill_slave_info is called. But with current ordering, rx_handler and IFF_BRIDGE_PORT are not set yet so there would have to be check for that in fill_slave_info callback. Resolve this by reordering to similar what bonding and team does to avoid the check. Also add removal of IFF_BRIDGE_PORT flag into error path. Signed-off-by:

Jiri Pirko <jiri@resnulli.us> Signed-off-by:

David S. Miller <davem@davemloft.net>

net/bridge/br_if.c

+8 −7

Original line number	Diff line number	Diff line
		@@ -252,12 +252,12 @@ static void del_nbp(struct net_bridge_port *p)
		br_fdb_delete_by_port(br, p, 1);
		nbp_update_port_count(br);

		netdev_upper_dev_unlink(dev, br->dev);

		dev->priv_flags &= ~IFF_BRIDGE_PORT;

		netdev_rx_handler_unregister(dev);

		netdev_upper_dev_unlink(dev, br->dev);

		br_multicast_del_port(p);

		kobject_uevent(&p->kobj, KOBJ_REMOVE);
		@@ -476,16 +476,16 @@ int br_add_if(struct net_bridge br, struct net_device dev)
		if (err)
		goto err3;

		err = netdev_master_upper_dev_link(dev, br->dev);
		err = netdev_rx_handler_register(dev, br_handle_frame, p);
		if (err)
		goto err4;

		err = netdev_rx_handler_register(dev, br_handle_frame, p);
		dev->priv_flags \|= IFF_BRIDGE_PORT;

		err = netdev_master_upper_dev_link(dev, br->dev);
		if (err)
		goto err5;

		dev->priv_flags \|= IFF_BRIDGE_PORT;

		dev_disable_lro(dev);

		list_add_rcu(&p->list, &br->port_list);
		@@ -520,7 +520,8 @@ int br_add_if(struct net_bridge br, struct net_device dev)
		return 0;

		err5:
		netdev_upper_dev_unlink(dev, br->dev);
		dev->priv_flags &= ~IFF_BRIDGE_PORT;
		netdev_rx_handler_unregister(dev);
		err4:
		br_netpoll_disable(p);
		err3:

Admin message