Patchwork [2/6] netprio_cgroup: don't allocate prio table when a device is registered

login
register
mail settings
Submitter Li Zefan
Date Feb. 1, 2012, 6:55 a.m.
Message ID <4F28E1E3.80507@cn.fujitsu.com>
Download mbox | patch
Permalink /patch/138908/
State Changes Requested
Delegated to: David Miller
Headers show

Comments

Li Zefan - Feb. 1, 2012, 6:55 a.m.
So we delay the allocation till the priority is set through cgroup,
and this makes skb_update_priority() faster when it's not set.

This also eliminates an off-by-one bug similar with the one fixed
in the previous patch.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
---
 net/core/netprio_cgroup.c |    6 ------
 1 files changed, 0 insertions(+), 6 deletions(-)

Patch

diff --git a/net/core/netprio_cgroup.c b/net/core/netprio_cgroup.c
index a296cbb..2edfa6b 100644
--- a/net/core/netprio_cgroup.c
+++ b/net/core/netprio_cgroup.c
@@ -270,7 +270,6 @@  static int netprio_device_event(struct notifier_block *unused,
 {
 	struct net_device *dev = ptr;
 	struct netprio_map *old;
-	u32 max_len = atomic_read(&max_prioidx);
 
 	/*
 	 * Note this is called with rtnl_lock held so we have update side
@@ -278,11 +277,6 @@  static int netprio_device_event(struct notifier_block *unused,
 	 */
 
 	switch (event) {
-
-	case NETDEV_REGISTER:
-		if (max_len)
-			extend_netdev_table(dev, max_len);
-		break;
 	case NETDEV_UNREGISTER:
 		old = rtnl_dereference(dev->priomap);
 		RCU_INIT_POINTER(dev->priomap, NULL);