Merge branch 'net-allow-per-net-notifier-to-follow-netdev-into-namespace'

Jiri Pirko says:

====================
net: allow per-net notifier to follow netdev into namespace

Currently we have per-net notifier, which allows to get only
notifications relevant to particular network namespace. That is enough
for drivers that have netdevs local in a particular namespace (cannot
move elsewhere).

However if netdev can change namespace, per-net notifier cannot be used.
Introduce dev_net variant that is basically per-net notifier with an
extension that re-registers the per-net notifier upon netdev namespace
change. Basically the per-net notifier follows the netdev into
namespace.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/mellanox/mlx5/core/en/fs.h

@@ -21,6 +21,7 @@ struct mlx5e_tc_table {
 	DECLARE_HASHTABLE(hairpin_tbl, 8);
 
 	struct notifier_block     netdevice_nb;
+	struct netdev_net_notifier	netdevice_nn;
 };
 
 struct mlx5e_flow_table {

drivers/net/ethernet/mellanox/mlx5/core/en_main.c

@@ -5144,6 +5144,7 @@ static void mlx5e_nic_enable(struct mlx5e_priv *priv)
 
 static void mlx5e_nic_disable(struct mlx5e_priv *priv)
 {
+	struct net_device *netdev = priv->netdev;
 	struct mlx5_core_dev *mdev = priv->mdev;
 
 #ifdef CONFIG_MLX5_CORE_EN_DCB
@@ -5164,7 +5165,7 @@ static void mlx5e_nic_disable(struct mlx5e_priv *priv)
 		mlx5e_monitor_counter_cleanup(priv);
 
 	mlx5e_disable_async_events(priv);
-	mlx5_lag_remove(mdev);
+	mlx5_lag_remove(mdev, netdev);
 }
 
 int mlx5e_update_nic_rx(struct mlx5e_priv *priv)

drivers/net/ethernet/mellanox/mlx5/core/en_rep.c

@@ -1731,7 +1731,9 @@ static int mlx5e_init_uplink_rep_tx(struct mlx5e_rep_priv *rpriv)
 	/* init indirect block notifications */
 	INIT_LIST_HEAD(&uplink_priv->tc_indr_block_priv_list);
 	uplink_priv->netdevice_nb.notifier_call = mlx5e_nic_rep_netdevice_event;
-	err = register_netdevice_notifier(&uplink_priv->netdevice_nb);
+	err = register_netdevice_notifier_dev_net(rpriv->netdev,
+						  &uplink_priv->netdevice_nb,
+						  &uplink_priv->netdevice_nn);
 	if (err) {
 		mlx5_core_err(priv->mdev, "Failed to register netdev notifier\n");
 		goto tc_esw_cleanup;
@@ -1770,8 +1772,12 @@ static int mlx5e_init_rep_tx(struct mlx5e_priv *priv)
 
 static void mlx5e_cleanup_uplink_rep_tx(struct mlx5e_rep_priv *rpriv)
 {
+	struct mlx5_rep_uplink_priv *uplink_priv = &rpriv->uplink_priv;
+
 	/* clean indirect TC block notifications */
-	unregister_netdevice_notifier(&rpriv->uplink_priv.netdevice_nb);
+	unregister_netdevice_notifier_dev_net(rpriv->netdev,
+					      &uplink_priv->netdevice_nb,
+					      &uplink_priv->netdevice_nn);
 	mlx5e_rep_indr_clean_block_privs(rpriv);
 
 	/* delete shared tc flow table */
@@ -1855,6 +1861,7 @@ static void mlx5e_uplink_rep_enable(struct mlx5e_priv *priv)
 
 static void mlx5e_uplink_rep_disable(struct mlx5e_priv *priv)
 {
+	struct net_device *netdev = priv->netdev;
 	struct mlx5_core_dev *mdev = priv->mdev;
 	struct mlx5e_rep_priv *rpriv = priv->ppriv;
 
@@ -1863,7 +1870,7 @@ static void mlx5e_uplink_rep_disable(struct mlx5e_priv *priv)
 #endif
 	mlx5_notifier_unregister(mdev, &priv->events_nb);
 	cancel_work_sync(&rpriv->uplink_priv.reoffload_flows_work);
-	mlx5_lag_remove(mdev);
+	mlx5_lag_remove(mdev, netdev);
 }
 
 static MLX5E_DEFINE_STATS_GRP(sw_rep, 0);

drivers/net/ethernet/mellanox/mlx5/core/en_rep.h

@@ -73,6 +73,7 @@ struct mlx5_rep_uplink_priv {
 	 */
 	struct list_head	    tc_indr_block_priv_list;
 	struct notifier_block	    netdevice_nb;
+	struct netdev_net_notifier  netdevice_nn;
 
 	struct mlx5_tun_entropy tun_entropy;
 

drivers/net/ethernet/mellanox/mlx5/core/en_tc.c

@@ -4251,7 +4251,10 @@ int mlx5e_tc_nic_init(struct mlx5e_priv *priv)
 		return err;
 
 	tc->netdevice_nb.notifier_call = mlx5e_tc_netdev_event;
-	if (register_netdevice_notifier(&tc->netdevice_nb)) {
+	err = register_netdevice_notifier_dev_net(priv->netdev,
+						  &tc->netdevice_nb,
+						  &tc->netdevice_nn);
+	if (err) {
 		tc->netdevice_nb.notifier_call = NULL;
 		mlx5_core_warn(priv->mdev, "Failed to register netdev notifier\n");
 	}
@@ -4273,7 +4276,9 @@ void mlx5e_tc_nic_cleanup(struct mlx5e_priv *priv)
 	struct mlx5e_tc_table *tc = &priv->fs.tc;
 
 	if (tc->netdevice_nb.notifier_call)
-		unregister_netdevice_notifier(&tc->netdevice_nb);
+		unregister_netdevice_notifier_dev_net(priv->netdev,
+						      &tc->netdevice_nb,
+						      &tc->netdevice_nn);
 
 	mutex_destroy(&tc->mod_hdr.lock);
 	mutex_destroy(&tc->hairpin_tbl_lock);

drivers/net/ethernet/mellanox/mlx5/core/lag.c

@@ -586,7 +586,8 @@ void mlx5_lag_add(struct mlx5_core_dev *dev, struct net_device *netdev)
 
 	if (!ldev->nb.notifier_call) {
 		ldev->nb.notifier_call = mlx5_lag_netdev_event;
-		if (register_netdevice_notifier(&ldev->nb)) {
+		if (register_netdevice_notifier_dev_net(netdev, &ldev->nb,
+							&ldev->nn)) {
 			ldev->nb.notifier_call = NULL;
 			mlx5_core_err(dev, "Failed to register LAG netdev notifier\n");
 		}
@@ -599,7 +600,7 @@ void mlx5_lag_add(struct mlx5_core_dev *dev, struct net_device *netdev)
 }
 
 /* Must be called with intf_mutex held */
-void mlx5_lag_remove(struct mlx5_core_dev *dev)
+void mlx5_lag_remove(struct mlx5_core_dev *dev, struct net_device *netdev)
 {
 	struct mlx5_lag *ldev;
 	int i;
@@ -619,7 +620,8 @@ void mlx5_lag_remove(struct mlx5_core_dev *dev)
 
 	if (i == MLX5_MAX_PORTS) {
 		if (ldev->nb.notifier_call)
-			unregister_netdevice_notifier(&ldev->nb);
+			unregister_netdevice_notifier_dev_net(netdev, &ldev->nb,
+							      &ldev->nn);
 		mlx5_lag_mp_cleanup(ldev);
 		cancel_delayed_work_sync(&ldev->bond_work);
 		mlx5_lag_dev_free(ldev);

drivers/net/ethernet/mellanox/mlx5/core/lag.h

@@ -44,6 +44,7 @@ struct mlx5_lag {
 	struct workqueue_struct   *wq;
 	struct delayed_work       bond_work;
 	struct notifier_block     nb;
+	struct netdev_net_notifier	nn;
 	struct lag_mp             lag_mp;
 };
 

drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h

@@ -157,7 +157,7 @@ int mlx5_query_qcam_reg(struct mlx5_core_dev *mdev, u32 *qcam,
 			u8 feature_group, u8 access_reg_group);
 
 void mlx5_lag_add(struct mlx5_core_dev *dev, struct net_device *netdev);
-void mlx5_lag_remove(struct mlx5_core_dev *dev);
+void mlx5_lag_remove(struct mlx5_core_dev *dev, struct net_device *netdev);
 
 int mlx5_irq_table_init(struct mlx5_core_dev *dev);
 void mlx5_irq_table_cleanup(struct mlx5_core_dev *dev);

include/linux/netdevice.h

@@ -939,6 +939,11 @@ struct netdev_name_node {
 int netdev_name_node_alt_create(struct net_device *dev, const char *name);
 int netdev_name_node_alt_destroy(struct net_device *dev, const char *name);
 
+struct netdev_net_notifier {
+	struct list_head list;
+	struct notifier_block *nb;
+};
+
 /*
  * This structure defines the management hooks for network devices.
  * The following hooks can be defined; unless noted otherwise, they are
@@ -1793,6 +1798,10 @@ enum netdev_priv_flags {
  *
  *	@wol_enabled:	Wake-on-LAN is enabled
  *
+ *	@net_notifier_list:	List of per-net netdev notifier block
+ *				that follow this device when it is moved
+ *				to another network namespace.
+ *
  *	FIXME: cleanup struct net_device such that network protocol info
  *	moves out.
  */
@@ -2085,6 +2094,8 @@ struct net_device {
 	struct lock_class_key	addr_list_lock_key;
 	bool			proto_down;
 	unsigned		wol_enabled:1;
+
+	struct list_head	net_notifier_list;
 };
 #define to_net_dev(d) container_of(d, struct net_device, dev)
 
@@ -2529,6 +2540,12 @@ int unregister_netdevice_notifier(struct notifier_block *nb);
 int register_netdevice_notifier_net(struct net *net, struct notifier_block *nb);
 int unregister_netdevice_notifier_net(struct net *net,
 				      struct notifier_block *nb);
+int register_netdevice_notifier_dev_net(struct net_device *dev,
+					struct notifier_block *nb,
+					struct netdev_net_notifier *nn);
+int unregister_netdevice_notifier_dev_net(struct net_device *dev,
+					  struct notifier_block *nb,
+					  struct netdev_net_notifier *nn);
 
 struct netdev_notifier_info {
 	struct net_device	*dev;

net/core/dev.c

@@ -1764,7 +1764,6 @@ EXPORT_SYMBOL(register_netdevice_notifier);
 
 int unregister_netdevice_notifier(struct notifier_block *nb)
 {
-	struct net_device *dev;
 	struct net *net;
 	int err;
 
@@ -1775,16 +1774,9 @@ int unregister_netdevice_notifier(struct notifier_block *nb)
 	if (err)
 		goto unlock;
 
-	for_each_net(net) {
-		for_each_netdev(net, dev) {
-			if (dev->flags & IFF_UP) {
-				call_netdevice_notifier(nb, NETDEV_GOING_DOWN,
-							dev);
-				call_netdevice_notifier(nb, NETDEV_DOWN, dev);
-			}
-			call_netdevice_notifier(nb, NETDEV_UNREGISTER, dev);
-		}
-	}
+	for_each_net(net)
+		call_netdevice_unregister_net_notifiers(nb, net);
+
 unlock:
 	rtnl_unlock();
 	up_write(&pernet_ops_rwsem);
@@ -1792,6 +1784,42 @@ int unregister_netdevice_notifier(struct notifier_block *nb)
 }
 EXPORT_SYMBOL(unregister_netdevice_notifier);
 
+static int __register_netdevice_notifier_net(struct net *net,
+					     struct notifier_block *nb,
+					     bool ignore_call_fail)
+{
+	int err;
+
+	err = raw_notifier_chain_register(&net->netdev_chain, nb);
+	if (err)
+		return err;
+	if (dev_boot_phase)
+		return 0;
+
+	err = call_netdevice_register_net_notifiers(nb, net);
+	if (err && !ignore_call_fail)
+		goto chain_unregister;
+
+	return 0;
+
+chain_unregister:
+	raw_notifier_chain_unregister(&net->netdev_chain, nb);
+	return err;
+}
+
+static int __unregister_netdevice_notifier_net(struct net *net,
+					       struct notifier_block *nb)
+{
+	int err;
+
+	err = raw_notifier_chain_unregister(&net->netdev_chain, nb);
+	if (err)
+		return err;
+
+	call_netdevice_unregister_net_notifiers(nb, net);
+	return 0;
+}
+
 /**
  * register_netdevice_notifier_net - register a per-netns network notifier block
  * @net: network namespace
@@ -1812,23 +1840,9 @@ int register_netdevice_notifier_net(struct net *net, struct notifier_block *nb)
 	int err;
 
 	rtnl_lock();
-	err = raw_notifier_chain_register(&net->netdev_chain, nb);
-	if (err)
-		goto unlock;
-	if (dev_boot_phase)
-		goto unlock;
-
-	err = call_netdevice_register_net_notifiers(nb, net);
-	if (err)
-		goto chain_unregister;
-
-unlock:
+	err = __register_netdevice_notifier_net(net, nb, false);
 	rtnl_unlock();
 	return err;
-
-chain_unregister:
-	raw_notifier_chain_unregister(&netdev_chain, nb);
-	goto unlock;
 }
 EXPORT_SYMBOL(register_netdevice_notifier_net);
 
@@ -1854,18 +1868,54 @@ int unregister_netdevice_notifier_net(struct net *net,
 	int err;
 
 	rtnl_lock();
-	err = raw_notifier_chain_unregister(&net->netdev_chain, nb);
-	if (err)
-		goto unlock;
-
-	call_netdevice_unregister_net_notifiers(nb, net);
-
-unlock:
+	err = __unregister_netdevice_notifier_net(net, nb);
 	rtnl_unlock();
 	return err;
 }
 EXPORT_SYMBOL(unregister_netdevice_notifier_net);
 
+int register_netdevice_notifier_dev_net(struct net_device *dev,
+					struct notifier_block *nb,
+					struct netdev_net_notifier *nn)
+{
+	int err;
+
+	rtnl_lock();
+	err = __register_netdevice_notifier_net(dev_net(dev), nb, false);
+	if (!err) {
+		nn->nb = nb;
+		list_add(&nn->list, &dev->net_notifier_list);
+	}
+	rtnl_unlock();
+	return err;
+}
+EXPORT_SYMBOL(register_netdevice_notifier_dev_net);
+
+int unregister_netdevice_notifier_dev_net(struct net_device *dev,
+					  struct notifier_block *nb,
+					  struct netdev_net_notifier *nn)
+{
+	int err;
+
+	rtnl_lock();
+	list_del(&nn->list);
+	err = __unregister_netdevice_notifier_net(dev_net(dev), nb);
+	rtnl_unlock();
+	return err;
+}
+EXPORT_SYMBOL(unregister_netdevice_notifier_dev_net);
+
+static void move_netdevice_notifiers_dev_net(struct net_device *dev,
+					     struct net *net)
+{
+	struct netdev_net_notifier *nn;
+
+	list_for_each_entry(nn, &dev->net_notifier_list, list) {
+		__unregister_netdevice_notifier_net(dev_net(dev), nn->nb);
+		__register_netdevice_notifier_net(net, nn->nb, true);
+	}
+}
+
 /**
  *	call_netdevice_notifiers_info - call all network notifier blocks
  *	@val: value passed unmodified to notifier function
@@ -9778,6 +9828,7 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name,
 	INIT_LIST_HEAD(&dev->adj_list.lower);
 	INIT_LIST_HEAD(&dev->ptype_all);
 	INIT_LIST_HEAD(&dev->ptype_specific);
+	INIT_LIST_HEAD(&dev->net_notifier_list);
 #ifdef CONFIG_NET_SCHED
 	hash_init(dev->qdisc_hash);
 #endif
@@ -10041,6 +10092,9 @@ int dev_change_net_namespace(struct net_device *dev, struct net *net, const char
 	kobject_uevent(&dev->dev.kobj, KOBJ_REMOVE);
 	netdev_adjacent_del_links(dev);
 
+	/* Move per-net netdevice notifiers that are following the netdevice */
+	move_netdevice_notifiers_dev_net(dev, net);
+
 	/* Actually switch the network namespace */
 	dev_net_set(dev, net);
 	dev->ifindex = new_ifindex;