Skip to contentSkip to conversation

netns: Simplify the network namespace list locking rules.

PATCH Applied Eric W. Biederman 3 participants 3 messages Sep 27, 2007 1 message not in the archive

Subsystems (2)
lore
Share & tools
lore
#1 lore raw

PATCH 1/1 netns: Simplify the network namespace list locking rules.

In reply to a message not in the archive

Denis V. Lunev <[email protected]> noticed that the locking rules for the network namespace list are over complicated and broken.

In particular the current register_netdev_notifier currently does not take any lock making the for_each_net iteration racy with network namespace creation and destruction. Oops.

The fact that we need to use for_each_net in rtnl_unlock() when the rtnetlink support becomes per network namespace makes designing the proper locking tricky. In addition we need to be able to call rtnl_lock() and rtnl_unlock() when we have the net_mutex held.

After thinking about it and looking at the alternatives carefully it looks like the simplest and most maintainable solution is to remove net_list_mutex altogether, and to use the rtnl_mutex instead.

---
 include/net/net_namespace.h |    3 ---
 net/core/net_namespace.c    |   23 ++++++-----------------
 2 files changed, 6 insertions(+), 20 deletions(-)
include/net/net_namespace.h +0 -3
index 13b0e3b..934c840 100644
@@ -96,9 +96,6 @@ static inline void release_net(struct net *net)
 #endif
 }
 
-extern void net_lock(void);
-extern void net_unlock(void);
-
 #define for_each_net(VAR)				\
 	list_for_each_entry(VAR, &net_namespace_list, list)
 
net/core/net_namespace.c +6 -17
index e478e35..0e0ca6d 100644
@@ -15,7 +15,6 @@ static LIST_HEAD(pernet_list);
 static struct list_head *first_device = &pernet_list;
 static DEFINE_MUTEX(net_mutex);
 
-static DEFINE_MUTEX(net_list_mutex);
 LIST_HEAD(net_namespace_list);
 
 static struct kmem_cache *net_cachep;
@@ -23,16 +22,6 @@ static struct kmem_cache *net_cachep;
 struct net init_net;
 EXPORT_SYMBOL_GPL(init_net);
 
-void net_lock(void)
-{
-	mutex_lock(&net_list_mutex);
-}
-
-void net_unlock(void)
-{
-	mutex_unlock(&net_list_mutex);
-}
-
 static struct net *net_alloc(void)
 {
 	return kmem_cache_alloc(net_cachep, GFP_KERNEL);
@@ -62,9 +51,9 @@ static void cleanup_net(struct work_struct *work)
 	mutex_lock(&net_mutex);
 
 	/* Don't let anyone else find us. */
-	net_lock();
+	rtnl_lock();
 	list_del(&net->list);
-	net_unlock();
+	rtnl_unlock();
 
 	/* Run all of the network namespace exit methods */
 	list_for_each_entry_reverse(ops, &pernet_list, list) {
@@ -151,9 +140,9 @@ struct net *copy_net_ns(unsigned long flags, struct net *old_net)
 	if (err)
 		goto out_unlock;
 
-	net_lock();
+	rtnl_lock();
 	list_add_tail(&new_net->list, &net_namespace_list);
-	net_unlock();
+	rtnl_unlock();
 
 
 out_unlock:
@@ -178,9 +167,9 @@ static int __init net_ns_init(void)
 	mutex_lock(&net_mutex);
 	err = setup_net(&init_net);
 
-	net_lock();
+	rtnl_lock();
 	list_add_tail(&init_net.list, &net_namespace_list);
-	net_unlock();
+	rtnl_unlock();
 
 	mutex_unlock(&net_mutex);
 	if (err)
Signature · 1 line
1.5.3.rc6.17.g1911
Acked-by on v1 1/1
#2 lore raw
Quoted 12 lines · Eric W. Biederman

Denis V. Lunev <[email protected]> noticed that the locking rules for the network namespace list are over complicated and broken.

In particular the current register_netdev_notifier currently does not take any lock making the for_each_net iteration racy with network namespace creation and destruction. Oops.

The fact that we need to use for_each_net in rtnl_unlock() when the rtnetlink support becomes per network namespace makes designing the proper locking tricky. In addition we need to be able to call rtnl_lock() and rtnl_unlock() when we have the net_mutex held.

Quoted 5 lines

After thinking about it and looking at the alternatives carefully it looks like the simplest and most maintainable solution is to remove net_list_mutex altogether, and to use the rtnl_mutex instead.

maintainer Applied
#3 lore raw
In reply to Eric W. Biederman

From: [email protected] (Eric W. Biederman)
Date: Wed, 26 Sep 2007 21:54:47 -0600

Quoted 13 lines

Denis V. Lunev <[email protected]> noticed that the locking rules for the network namespace list are over complicated and broken.

In particular the current register_netdev_notifier currently does not take any lock making the for_each_net iteration racy with network namespace creation and destruction. Oops.

The fact that we need to use for_each_net in rtnl_unlock() when the rtnetlink support becomes per network namespace makes designing the proper locking tricky. In addition we need to be able to call rtnl_lock() and rtnl_unlock() when we have the net_mutex held.

Quoted 5 lines

After thinking about it and looking at the alternatives carefully it looks like the simplest and most maintainable solution is to remove net_list_mutex altogether, and to use the rtnl_mutex instead.

Applied, thanks Eric.