[Devel] [PATCH RH7] net: skip genenerating uevents for network namespaces that are exiting

[Kirill Tkhai]

From: Andrey Vagin <avagin at openvz.org>

ms commit 002d8a1a6c11

No one can see these events, because a network namespace can not be
destroyed, if it has sockets.

Unlike other devices, uevent-s for network devices are generated
only inside their network namespaces. They are filtered in
kobj_bcast_filter()

My experiments shows that net namespaces are destroyed more 30% faster
with this optimization.

Here is a perf output for destroying network namespaces without this
patch.

-   94.76%     0.02%  kworker/u48:1  [kernel.kallsyms]     [k] cleanup_net
   - 94.74% cleanup_net
      - 94.64% ops_exit_list.isra.4
         - 41.61% default_device_exit_batch
            - 41.47% unregister_netdevice_many
               - rollback_registered_many
                  - 40.36% netdev_unregister_kobject
                     - 14.55% device_del
                        + 13.71% kobject_uevent
                     - 13.04% netdev_queue_update_kobjects
                        + 12.96% kobject_put
                     - 12.72% net_rx_queue_update_kobjects
                          kobject_put
                        - kobject_release
                           + 12.69% kobject_uevent
                  + 0.80% call_netdevice_notifiers_info
         + 19.57% nfsd_exit_net
         + 11.15% tcp_net_metrics_exit
         + 8.25% rpcsec_gss_exit_net

It's very critical to optimize the exit path for network namespaces,
because they are destroyed under net_mutex and many namespaces can be
destroyed for one iteration.

v2: use dev_set_uevent_suppress()

Cc: Cong Wang <xiyou.wangcong at gmail.com>
Cc: "David S. Miller" <davem at davemloft.net>
Cc: Eric W. Biederman <ebiederm at xmission.com>
Signed-off-by: Andrei Vagin <avagin at openvz.org>
Signed-off-by: David S. Miller <davem at davemloft.net>

[ktkhai: Added missed exit_list initialization]

Signed-off-by: Kirill Tkhai <ktkhai at virtuozzo.com>
---
 net/core/net-sysfs.c     |   16 ++++++++++++++--
 net/core/net_namespace.c |    2 ++
 2 files changed, 16 insertions(+), 2 deletions(-)

diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
index 37fbab2bdd40..2580de8ebfc8 100644
--- a/net/core/net-sysfs.c
+++ b/net/core/net-sysfs.c
@@ -917,8 +917,14 @@ net_rx_queue_update_kobjects(struct net_device *net, int old_num, int new_num)
 		}
 	}
 
-	while (--i >= new_num)
-		kobject_put(&net->_rx[i].kobj);
+	while (--i >= new_num) {
+		struct kobject *kobj = &net->_rx[i].kobj;
+
+		if (!list_empty(&dev_net(net)->exit_list))
+			kobj->uevent_suppress = 1;
+
+		kobject_put(kobj);
+	}
 
 	return error;
 #else
@@ -1331,6 +1337,9 @@ netdev_queue_update_kobjects(struct net_device *net, int old_num, int new_num)
 	while (--i >= new_num) {
 		struct netdev_queue *queue = net->_tx + i;
 
+		if (!list_empty(&dev_net(net)->exit_list))
+			queue->kobj.uevent_suppress = 1;
+
 #ifdef CONFIG_BQL
 		sysfs_remove_group(&queue->kobj, &dql_group);
 #endif
@@ -1489,6 +1498,9 @@ void netdev_unregister_kobject(struct net_device * net)
 {
 	struct device *dev = &(net->dev);
 
+	if (!list_empty(&dev_net(net)->exit_list))
+		dev_set_uevent_suppress(dev, 1);
+
 	kobject_get(&dev->kobj);
 
 	remove_queue_kobjects(net);
diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c
index 8076eb163ffc..314a84e3f0ff 100644
--- a/net/core/net_namespace.c
+++ b/net/core/net_namespace.c
@@ -42,6 +42,7 @@ EXPORT_SYMBOL_GPL(net_namespace_list);
 
 struct net init_net = {
 	.dev_base_head = LIST_HEAD_INIT(init_net.dev_base_head),
+	.exit_list = LIST_HEAD_INIT(init_net.exit_list),
 #ifdef CONFIG_VE
 	.owner_ve = &ve0,
 #ifdef CONFIG_VE_IPTABLES
@@ -321,6 +322,7 @@ static __net_init int setup_net(struct net *net, struct user_namespace *user_ns)
 	net->user_ns = user_ns;
 	idr_init(&net->netns_ids);
 	spin_lock_init(&net->nsid_lock);
+	INIT_LIST_HEAD(&net->exit_list);
 
 	list_for_each_entry(ops, &pernet_list, list) {
 		error = ops_init(ops, net);