Skip to contentSkip to conversation

netns: foreach_netdev_safe is insufficient in default_device_exit

PATCH In review Eric W. Biederman 2 participants 2 messages Dec 28, 2008 – Dec 30, 2008

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

PATCH 1/1 netns: foreach_netdev_safe is insufficient in default_device_exit

During network namespace teardown we either move or delete all of the network devices associated with a network namespace. In the case of veth devices deleting one will also delete it's pair device. If both devices are in the same network namespace then for_each_netdev_safe is insufficient as next may point to the second veth device we have deleted.

To avoid problems I do what we do in __rtnl_kill_links and restart the scan of the device list, after we have deleted a device.

Currently dev_change_netnamespace does not appear to suffer from this problem, but wireless devices are also paired and likely should be moved between network namespaces together. So I have errored on the side of caution and restart the scan of the network devices in that case as well.

---
 net/core/dev.c |    8 +++++---
 1 files changed, 5 insertions(+), 3 deletions(-)
net/core/dev.c +5 -3
index daca72e..1f154d2 100644
@@ -5059,13 +5059,14 @@ static struct pernet_operations __net_initdata netdev_net_ops = {
 
 static void __net_exit default_device_exit(struct net *net)
 {
-	struct net_device *dev, *next;
+	struct net_device *dev;
 	/*
 	 * Push all migratable of the network devices back to the
 	 * initial network namespace
 	 */
 	rtnl_lock();
-	for_each_netdev_safe(net, dev, next) {
+restart:
+	for_each_netdev(net, dev) {
 		int err;
 		char fb_name[IFNAMSIZ];
 
@@ -5076,7 +5077,7 @@ static void __net_exit default_device_exit(struct net *net)
 		/* Delete virtual devices */
 		if (dev->rtnl_link_ops && dev->rtnl_link_ops->dellink) {
 			dev->rtnl_link_ops->dellink(dev);
-			continue;
+			goto restart;
 		}
 
 		/* Push remaing network devices to init_net */
@@ -5087,6 +5088,7 @@ static void __net_exit default_device_exit(struct net *net)
 				__func__, dev->name, err);
 			BUG();
 		}
+		goto restart;
 	}
 	rtnl_unlock();
 }
Signature · 1 line
1.5.3.rc6.17.g1911
maintainer
#1 lore raw

From: [email protected] (Eric W. Biederman)
Date: Sun, 28 Dec 2008 00:10:35 -0800

Quoted 11 lines

During network namespace teardown we either move or delete all of the network devices associated with a network namespace. In the case of veth devices deleting one will also delete it's pair device. If both devices are in the same network namespace then for_each_netdev_safe is insufficient as next may point to the second veth device we have deleted.

To avoid problems I do what we do in __rtnl_kill_links and restart the scan of the device list, after we have deleted a device.

Quoted 7 lines

Currently dev_change_netnamespace does not appear to suffer from this problem, but wireless devices are also paired and likely should be moved between network namespaces together. So I have errored on the side of caution and restart the scan of the network devices in that case as well.

Looks good, applied.

Thanks Eric.