[X/Y,SRU] net: use net->count to check whether a netns is alive or not

Message ID 1491290816-30962-1-git-send-email-tim.gardner@canonical.com
State New
Headers show

Commit Message

Tim Gardner April 4, 2017, 7:26 a.m.
From: Andrey Vagin <avagin@openvz.org>

BugLink: http://bugs.launchpad.net/bugs/1672144

The previous idea was to check whether a net namespace is in
net_exit_list or not. It doesn't work, because net->exit_list is used in
__register_pernet_operations and __unregister_pernet_operations where
all namespaces are added to a temporary list to make cleanup in a error
case, so list_empty(&net->exit_list) always returns false.

Reported-by: Mantas Mikul─Śnas <grawity@gmail.com>
Fixes: 002d8a1a6c11 ("net: skip genenerating uevents for network namespaces that are exiting")
Signed-off-by: Andrei Vagin <avagin@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
(back ported from commit 91864f5852f9996210fad400cf70fb85af091243)
Signed-off-by: Tim Gardner <tim.gardner@canonical.com>

Conflicts:
	net/core/net-sysfs.c
---

V2 - noted that this is a back ported patch.

 net/core/net-sysfs.c | 9 +++++++++
 1 file changed, 9 insertions(+)

Comments

Joseph Salisbury April 4, 2017, 9:08 a.m. | #1

Marcelo Henrique Cerri April 6, 2017, 2:43 p.m. | #2

Stefan Bader April 13, 2017, 4:10 p.m. | #3

Patch

diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
index f88a62a..056b5ca 100644
--- a/net/core/net-sysfs.c
+++ b/net/core/net-sysfs.c
@@ -931,6 +931,10 @@  net_rx_queue_update_kobjects(struct net_device *dev, int old_num, int new_num)
 	}
 
 	while (--i >= new_num) {
+		struct kobject *kobj = &dev->_rx[i].kobj;
+
+		if (!atomic_read(&dev_net(dev)->count))
+			kobj->uevent_suppress = 1;
 		if (dev->sysfs_rx_queue_group)
 			sysfs_remove_group(&dev->_rx[i].kobj,
 					   dev->sysfs_rx_queue_group);
@@ -1321,6 +1325,8 @@  netdev_queue_update_kobjects(struct net_device *dev, int old_num, int new_num)
 	while (--i >= new_num) {
 		struct netdev_queue *queue = dev->_tx + i;
 
+		if (!atomic_read(&dev_net(dev)->count))
+			queue->kobj.uevent_suppress = 1;
 #ifdef CONFIG_BQL
 		sysfs_remove_group(&queue->kobj, &dql_group);
 #endif
@@ -1506,6 +1512,9 @@  void netdev_unregister_kobject(struct net_device *ndev)
 {
 	struct device *dev = &(ndev->dev);
 
+	if (!atomic_read(&dev_net(ndev)->count))
+		dev_set_uevent_suppress(dev, 1);
+
 	kobject_get(&dev->kobj);
 
 	remove_queue_kobjects(ndev);