[PATCH] net: remove rcu locking from fib_rules_event()

[PATCH] net: remove rcu locking from fib_rules_event()

[Eric Dumazet]

We hold RTNL at this point and dont use RCU variants of list traversals,
we dont need rcu_read_lock()/rcu_read_unlock()

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
---
 net/core/fib_rules.c |   10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

diff --git a/net/core/fib_rules.c b/net/core/fib_rules.c
index 9a24377..2ff3489 100644
--- a/net/core/fib_rules.c
+++ b/net/core/fib_rules.c
@@ -108,7 +108,7 @@ fib_rules_register(struct fib_rules_ops *tmpl, struct net *net)
 	struct fib_rules_ops *ops;
 	int err;
 
-	ops = kmemdup(tmpl, sizeof (*ops), GFP_KERNEL);
+	ops = kmemdup(tmpl, sizeof(*ops), GFP_KERNEL);
 	if (ops == NULL)
 		return ERR_PTR(-ENOMEM);
 
@@ -123,7 +123,6 @@ fib_rules_register(struct fib_rules_ops *tmpl, struct net *net)
 
 	return ops;
 }
-
 EXPORT_SYMBOL_GPL(fib_rules_register);
 
 void fib_rules_cleanup_ops(struct fib_rules_ops *ops)
@@ -157,7 +156,6 @@ void fib_rules_unregister(struct fib_rules_ops *ops)
 
 	call_rcu(&ops->rcu, fib_rules_put_rcu);
 }
-
 EXPORT_SYMBOL_GPL(fib_rules_unregister);
 
 static int fib_rule_match(struct fib_rule *rule, struct fib_rules_ops *ops,
@@ -220,7 +218,6 @@ out:
 
 	return err;
 }
-
 EXPORT_SYMBOL_GPL(fib_rules_lookup);
 
 static int validate_rulemsg(struct fib_rule_hdr *frh, struct nlattr **tb,
@@ -613,7 +610,7 @@ static int fib_nl_dumprule(struct sk_buff *skb, struct netlink_callback *cb)
 			break;
 
 		cb->args[1] = 0;
-	skip:
+skip:
 		idx++;
 	}
 	rcu_read_unlock();
@@ -685,7 +682,6 @@ static int fib_rules_event(struct notifier_block *this, unsigned long event,
 	struct fib_rules_ops *ops;
 
 	ASSERT_RTNL();
-	rcu_read_lock();
 
 	switch (event) {
 	case NETDEV_REGISTER:
@@ -699,8 +695,6 @@ static int fib_rules_event(struct notifier_block *this, unsigned long event,
 		break;
 	}
 
-	rcu_read_unlock();
-
 	return NOTIFY_DONE;
 }
 

Re: [PATCH] net: remove rcu locking from fib_rules_event()

[David Miller]

From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Wed, 10 Mar 2010 07:03:38 +0100

> We hold RTNL at this point and dont use RCU variants of list traversals,
> we dont need rcu_read_lock()/rcu_read_unlock()
> 
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>

I'll apply this, but to my net-next-2.6 tree once I open that
up.

Re: [PATCH] net: remove rcu locking from fib_rules_event()

[David Miller]

From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Wed, 10 Mar 2010 07:03:38 +0100

> We hold RTNL at this point and dont use RCU variants of list traversals,
> we dont need rcu_read_lock()/rcu_read_unlock()
> 
> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>

Applied.

Re: [PATCH] net: remove rcu locking from fib_rules_event()

[Eric Dumazet]

Le mardi 16 mars 2010 à 21:28 -0700, David Miller a écrit :
> From: Eric Dumazet <eric.dumazet@gmail.com>
> Date: Wed, 10 Mar 2010 07:03:38 +0100
> 
> > We hold RTNL at this point and dont use RCU variants of list traversals,
> > we dont need rcu_read_lock()/rcu_read_unlock()
> > 
> > Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
> 
> Applied.

Thanks David.

I believe the synchronize_rcu() we have in this file is useless too,
I'll submit a patch.