Merge branch 'net-core-fix-a-lockdep-splat-in-the-dev_addr_list'

Taehee Yoo says:

====================
net: core: fix a lockdep splat in the dev_addr_list.

This patchset is to avoid lockdep splat.

When a stacked interface graph is changed, netif_addr_lock() is called
recursively and it internally calls spin_lock_nested().
The parameter of spin_lock_nested() is 'dev->lower_level',
this is called subclass.
The problem of 'dev->lower_level' is that while 'dev->lower_level' is
being used as a subclass of spin_lock_nested(), its value can be changed.
So, spin_lock_nested() would be called recursively with the same
subclass value, the lockdep understands a deadlock.
In order to avoid this, a new variable is needed and it is going to be
used as a parameter of spin_lock_nested().
The first and second patch is a preparation patch for the third patch.
In the third patch, the problem will be fixed.

The first patch is to add __netdev_upper_dev_unlink().
An existed netdev_upper_dev_unlink() is renamed to
__netdev_upper_dev_unlink(). and netdev_upper_dev_unlink()
is added as an wrapper of this function.

The second patch is to add the netdev_nested_priv structure.
netdev_walk_all_{ upper | lower }_dev() pass both private functions
and "data" pointer to handle their own things.
At this point, the data pointer type is void *.
In order to make it easier to expand common variables and functions,
this new netdev_nested_priv structure is added.

The third patch is to add a new variable 'nested_level'
into the net_device structure.
This variable will be used as a parameter of spin_lock_nested() of
dev->addr_list_lock.
Due to this variable, it can avoid lockdep splat.
====================

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

4 files changed, 25 insertions, 12 deletions

diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c
index ffad73bb40ff..5a76611e684a 100644
--- a/drivers/infiniband/core/cache.c
+++ b/drivers/infiniband/core/cache.c
@@ -1320,9 +1320,10 @@ struct net_device *rdma_read_gid_attr_ndev_rcu(const struct ib_gid_attr *attr)
 }
 EXPORT_SYMBOL(rdma_read_gid_attr_ndev_rcu);
 
-static int get_lower_dev_vlan(struct net_device *lower_dev, void *data)
+static int get_lower_dev_vlan(struct net_device *lower_dev,
+			      struct netdev_nested_priv *priv)
 {
-	u16 *vlan_id = data;
+	u16 *vlan_id = (u16 *)priv->data;
 
 	if (is_vlan_dev(lower_dev))
 		*vlan_id = vlan_dev_vlan_id(lower_dev);
@@ -1348,6 +1349,9 @@ static int get_lower_dev_vlan(struct net_device *lower_dev, void *data)
 int rdma_read_gid_l2_fields(const struct ib_gid_attr *attr,
 			    u16 *vlan_id, u8 *smac)
 {
+	struct netdev_nested_priv priv = {
+		.data = (void *)vlan_id,
+	};
 	struct net_device *ndev;
 
 	rcu_read_lock();
@@ -1368,7 +1372,7 @@ int rdma_read_gid_l2_fields(const struct ib_gid_attr *attr,
 			 * the lower vlan device for this gid entry.
 			 */
 			netdev_walk_all_lower_dev_rcu(attr->ndev,
-					get_lower_dev_vlan, vlan_id);
+					get_lower_dev_vlan, &priv);
 		}
 	}
 	rcu_read_unlock();
diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index 7f0e91e92968..5888311b2119 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -2865,9 +2865,10 @@ struct iboe_prio_tc_map {
 	bool found;
 };
 
-static int get_lower_vlan_dev_tc(struct net_device *dev, void *data)
+static int get_lower_vlan_dev_tc(struct net_device *dev,
+				 struct netdev_nested_priv *priv)
 {
-	struct iboe_prio_tc_map *map = data;
+	struct iboe_prio_tc_map *map = (struct iboe_prio_tc_map *)priv->data;
 
 	if (is_vlan_dev(dev))
 		map->output_tc = get_vlan_ndev_tc(dev, map->input_prio);
@@ -2886,16 +2887,18 @@ static int iboe_tos_to_sl(struct net_device *ndev, int tos)
 {
 	struct iboe_prio_tc_map prio_tc_map = {};
 	int prio = rt_tos2priority(tos);
+	struct netdev_nested_priv priv;
 
 	/* If VLAN device, get it directly from the VLAN netdev */
 	if (is_vlan_dev(ndev))
 		return get_vlan_ndev_tc(ndev, prio);
 
 	prio_tc_map.input_prio = prio;
+	priv.data = (void *)&prio_tc_map;
 	rcu_read_lock();
 	netdev_walk_all_lower_dev_rcu(ndev,
 				      get_lower_vlan_dev_tc,
-				      &prio_tc_map);
+				      &priv);
 	rcu_read_unlock();
 	/* If map is found from lower device, use it; Otherwise
 	 * continue with the current netdevice to get priority to tc map.
diff --git a/drivers/infiniband/core/roce_gid_mgmt.c b/drivers/infiniband/core/roce_gid_mgmt.c
index 2860def84f4d..6b8364bb032d 100644
--- a/drivers/infiniband/core/roce_gid_mgmt.c
+++ b/drivers/infiniband/core/roce_gid_mgmt.c
@@ -531,10 +531,11 @@ struct upper_list {
 	struct net_device *upper;
 };
 
-static int netdev_upper_walk(struct net_device *upper, void *data)
+static int netdev_upper_walk(struct net_device *upper,
+			     struct netdev_nested_priv *priv)
 {
 	struct upper_list *entry = kmalloc(sizeof(*entry), GFP_ATOMIC);
-	struct list_head *upper_list = data;
+	struct list_head *upper_list = (struct list_head *)priv->data;
 
 	if (!entry)
 		return 0;
@@ -553,12 +554,14 @@ static void handle_netdev_upper(struct ib_device *ib_dev, u8 port,
 						      struct net_device *ndev))
 {
 	struct net_device *ndev = cookie;
+	struct netdev_nested_priv priv;
 	struct upper_list *upper_iter;
 	struct upper_list *upper_temp;
 	LIST_HEAD(upper_list);
 
+	priv.data = &upper_list;
 	rcu_read_lock();
-	netdev_walk_all_upper_dev_rcu(ndev, netdev_upper_walk, &upper_list);
+	netdev_walk_all_upper_dev_rcu(ndev, netdev_upper_walk, &priv);
 	rcu_read_unlock();
 
 	handle_netdev(ib_dev, port, ndev);
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c
index ab75b7f745d4..f772fe8c5b66 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_main.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
@@ -342,9 +342,10 @@ struct ipoib_walk_data {
 	struct net_device *result;
 };
 
-static int ipoib_upper_walk(struct net_device *upper, void *_data)
+static int ipoib_upper_walk(struct net_device *upper,
+			    struct netdev_nested_priv *priv)
 {
-	struct ipoib_walk_data *data = _data;
+	struct ipoib_walk_data *data = (struct ipoib_walk_data *)priv->data;
 	int ret = 0;
 
 	if (ipoib_is_dev_match_addr_rcu(data->addr, upper)) {
@@ -368,10 +369,12 @@ static int ipoib_upper_walk(struct net_device *upper, void *_data)
 static struct net_device *ipoib_get_net_dev_match_addr(
 		const struct sockaddr *addr, struct net_device *dev)
 {
+	struct netdev_nested_priv priv;
 	struct ipoib_walk_data data = {
 		.addr = addr,
 	};
 
+	priv.data = (void *)&data;
 	rcu_read_lock();
 	if (ipoib_is_dev_match_addr_rcu(addr, dev)) {
 		dev_hold(dev);
@@ -379,7 +382,7 @@ static struct net_device *ipoib_get_net_dev_match_addr(
 		goto out;
 	}
 
-	netdev_walk_all_upper_dev_rcu(dev, ipoib_upper_walk, &data);
+	netdev_walk_all_upper_dev_rcu(dev, ipoib_upper_walk, &priv);
 out:
 	rcu_read_unlock();
 	return data.result;