Add struct net to all calls for rt_cache_flush

commit 628d4768843fc6d572596b91e29f8ce6506c276a

Author: Denis V. Lunev <[email protected]>
Date:   Thu Oct 18 12:48:38 2007 +0400

Add struct net to all calls for rt_cache_flush. Additionally, the function manipulating with timer should not actually call the garbage collector as the namespace is not known in the global rt_secret_rebuild.

Signed-off-by Denis V. Lunev <[email protected]>

include/net/route.h +1 -1

index 688c750..3372dce 100644
@@ -111,7 +111,7 @@ extern int		ip_rt_init(void);
 extern void		ip_rt_redirect(__be32 old_gw, __be32 dst, __be32 new_gw,
 				       __be32 src, struct net_device *dev);
 extern void		ip_rt_advice(struct rtable **rp, int advice);
-extern void		rt_cache_flush(int how);
+extern void		rt_cache_flush(struct net *net, int how);
 extern int		__ip_route_output_key(struct rtable **, const struct flowi *flp);
 extern int		ip_route_output_key(struct rtable **, struct flowi *flp);
 extern int		ip_route_output_flow(struct rtable **rp, struct flowi *flp, struct sock *sk, int flags);

net/ipv4/arp.c +1 -1

index 71c4d6d..7d9fb24 100644
@@ -1222,7 +1222,7 @@ static int arp_netdev_event(struct notifier_block *this, unsigned long event, vo
 	switch (event) {
 	case NETDEV_CHANGEADDR:
 		neigh_changeaddr(&arp_tbl, dev);
-		rt_cache_flush(0);
+		rt_cache_flush(dev->nd_net, 0);
 		break;
 	default:
 		break;

net/ipv4/devinet.c +5 -4

index 32f52d4..ab196a1 100644
@@ -56,6 +56,7 @@
 #include <linux/sysctl.h>
 #endif
 #include <linux/kmod.h>
+#include <linux/nsproxy.h>
 
 #include <net/arp.h>
 #include <net/ip.h>

@@ -1348,7 +1349,7 @@ void inet_forward_change(struct net *net)
 	}
 	read_unlock(&dev_base_lock);
 
-	rt_cache_flush(0);
+	rt_cache_flush(net, 0);
 }
 
 static int devinet_sysctl_forward(struct ctl_table *ctl, int write,

@@ -1364,7 +1365,7 @@ static int devinet_sysctl_forward(struct ctl_table *ctl, int write,
 		if (valp == &IPV4_DEVCONF_ALL(net, FORWARDING))
 			inet_forward_change(net);
 		else if (valp != &IPV4_DEVCONF_DFLT(net, FORWARDING))
-			rt_cache_flush(0);
+			rt_cache_flush(current->nsproxy->net_ns, 0);
 	}
 
 	return ret;

@@ -1379,7 +1380,7 @@ int ipv4_doint_and_flush(struct ctl_table *ctl, int write,
 	int ret = proc_dointvec(ctl, write, filp, buffer, lenp, ppos);
 
 	if (write && *valp != val)
-		rt_cache_flush(0);
+		rt_cache_flush(current->nsproxy->net_ns, 0);
 
 	return ret;
 }

@@ -1392,7 +1393,7 @@ int ipv4_doint_and_flush_strategy(struct ctl_table *table, int __user *name, int
 				      newval, newlen);
 
 	if (ret == 1)
-		rt_cache_flush(0);
+		rt_cache_flush(current->nsproxy->net_ns, 0);
 
 	return ret;
 }

net/ipv4/fib_frontend.c +9 -8

index ad6bfce..835a8b7 100644
@@ -113,7 +113,7 @@ static void fib_flush(struct net *net)
 	}
 
 	if (flushed)
-		rt_cache_flush(-1);
+		rt_cache_flush(net, -1);
 }
 
 /*

@@ -866,21 +866,22 @@ static void fib_disable_ip(struct net_device *dev, int force)
 	struct net *net = dev->nd_net;
 	if (fib_sync_down(net, 0, dev, force))
 		fib_flush(net);
-	rt_cache_flush(0);
+	rt_cache_flush(dev->nd_net, 0);
 	arp_ifdown(dev);
 }
 
 static int fib_inetaddr_event(struct notifier_block *this, unsigned long event, void *ptr)
 {
 	struct in_ifaddr *ifa = (struct in_ifaddr*)ptr;
+	struct net_device *dev = ifa->ifa_dev->dev;
 
 	switch (event) {
 	case NETDEV_UP:
 		fib_add_ifaddr(ifa);
 #ifdef CONFIG_IP_ROUTE_MULTIPATH
-		fib_sync_up(ifa->ifa_dev->dev);
+		fib_sync_up(dev);
 #endif
-		rt_cache_flush(-1);
+		rt_cache_flush(dev->nd_net, -1);
 		break;
 	case NETDEV_DOWN:
 		fib_del_ifaddr(ifa);

@@ -888,9 +889,9 @@ static int fib_inetaddr_event(struct notifier_block *this, unsigned long event,
 			/* Last address was deleted from this interface.
 			   Disable IP.
 			 */
-			fib_disable_ip(ifa->ifa_dev->dev, 1);
+			fib_disable_ip(dev, 1);
 		} else {
-			rt_cache_flush(-1);
+			rt_cache_flush(dev->nd_net, -1);
 		}
 		break;
 	}

@@ -921,14 +922,14 @@ static int fib_netdev_event(struct notifier_block *this, unsigned long event, vo
 #ifdef CONFIG_IP_ROUTE_MULTIPATH
 		fib_sync_up(dev);
 #endif
-		rt_cache_flush(-1);
+		rt_cache_flush(dev->nd_net, -1);
 		break;
 	case NETDEV_DOWN:
 		fib_disable_ip(dev, 0);
 		break;
 	case NETDEV_CHANGEMTU:
 	case NETDEV_CHANGE:
-		rt_cache_flush(0);
+		rt_cache_flush(dev->nd_net, 0);
 		break;
 	}
 	return NOTIFY_DONE;

net/ipv4/fib_hash.c +3 -3

index 5d28d44..c32bd0a 100644
@@ -455,7 +455,7 @@ static int fn_hash_insert(struct fib_table *tb, struct fib_config *cfg)
 
 			fib_release_info(fi_drop);
 			if (state & FA_S_ACCESSED)
-				rt_cache_flush(-1);
+				rt_cache_flush(cfg->fc_nlinfo.net, -1);
 			rtmsg_fib(RTM_NEWROUTE, key, fa, cfg->fc_dst_len, tb->tb_id,
 				  &cfg->fc_nlinfo, NLM_F_REPLACE);
 			return 0;

@@ -522,7 +522,7 @@ static int fn_hash_insert(struct fib_table *tb, struct fib_config *cfg)
 
 	if (new_f)
 		fz->fz_nent++;
-	rt_cache_flush(-1);
+	rt_cache_flush(cfg->fc_nlinfo.net, -1);
 
 	rtmsg_fib(RTM_NEWROUTE, key, new_fa, cfg->fc_dst_len, tb->tb_id,
 		  &cfg->fc_nlinfo, 0);

@@ -604,7 +604,7 @@ static int fn_hash_delete(struct fib_table *tb, struct fib_config *cfg)
 		write_unlock_bh(&fib_hash_lock);
 
 		if (fa->fa_state & FA_S_ACCESSED)
-			rt_cache_flush(-1);
+			rt_cache_flush(cfg->fc_nlinfo.net, -1);
 		fn_free_alias(fa);
 		if (kill_fn) {
 			fn_free_node(f);

net/ipv4/fib_rules.c +1 -1

index c93b278..63df6f9 100644
@@ -281,7 +281,7 @@ static size_t fib4_rule_nlmsg_payload(struct fib_rule *rule)
 
 static void fib4_rule_flush_cache(struct net *net, struct fib_rules_ops *ops)
 {
-	rt_cache_flush(-1);
+	rt_cache_flush(net, -1);
 }
 
 static struct fib4_rule_table fib4_rule_table = {

net/ipv4/fib_trie.c +3 -3

index 1415ad4..8a24f21 100644
@@ -1230,7 +1230,7 @@ static int fn_trie_insert(struct fib_table *tb, struct fib_config *cfg)
 
 			fib_release_info(fi_drop);
 			if (state & FA_S_ACCESSED)
-				rt_cache_flush(-1);
+				rt_cache_flush(cfg->fc_nlinfo.net, -1);
 			rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen,
 				tb->tb_id, &cfg->fc_nlinfo, NLM_F_REPLACE);
 

@@ -1283,7 +1283,7 @@ static int fn_trie_insert(struct fib_table *tb, struct fib_config *cfg)
 	list_add_tail_rcu(&new_fa->fa_list,
 			  (fa ? &fa->fa_list : fa_head));
 
-	rt_cache_flush(-1);
+	rt_cache_flush(cfg->fc_nlinfo.net, -1);
 	rtmsg_fib(RTM_NEWROUTE, htonl(key), new_fa, plen, tb->tb_id,
 		  &cfg->fc_nlinfo, 0);
 succeeded:

@@ -1649,7 +1649,7 @@ static int fn_trie_delete(struct fib_table *tb, struct fib_config *cfg)
 		trie_leaf_remove(t, key);
 
 	if (fa->fa_state & FA_S_ACCESSED)
-		rt_cache_flush(-1);
+		rt_cache_flush(cfg->fc_nlinfo.net, -1);
 
 	fib_release_info(fa->fa_info);
 	alias_free_mem_rcu(fa);

net/ipv4/route.c +15 -7

index f9a59ff..ed1842b 100644
@@ -91,6 +91,7 @@
 #include <linux/jhash.h>
 #include <linux/rcupdate.h>
 #include <linux/times.h>
+#include <linux/nsproxy.h>
 #include <net/net_namespace.h>
 #include <net/protocol.h>
 #include <net/ip.h>

@@ -670,7 +671,7 @@ static void rt_run_flush(unsigned long dummy)
 
 static DEFINE_SPINLOCK(rt_flush_lock);
 
-void rt_cache_flush(int delay)
+static int __rt_cache_flush(int delay)
 {
 	unsigned long now = jiffies;
 	int user_mode = !in_softirq();

@@ -699,8 +700,7 @@ void rt_cache_flush(int delay)
 
 	if (delay <= 0) {
 		spin_unlock_bh(&rt_flush_lock);
-		rt_run_flush(0);
-		return;
+		return 1;
 	}
 
 	if (rt_deadline == 0)

@@ -708,13 +708,21 @@ void rt_cache_flush(int delay)
 
 	mod_timer(&rt_flush_timer, now+delay);
 	spin_unlock_bh(&rt_flush_lock);
+	return 0;
+}
+
+void rt_cache_flush(struct net *net, int delay)
+{
+	if (__rt_cache_flush(delay))
+		rt_run_flush(0);
 }
 
 static void rt_secret_rebuild(unsigned long dummy)
 {
 	unsigned long now = jiffies;
 
-	rt_cache_flush(0);
+	__rt_cache_flush(0);
+	rt_run_flush(0);
 	mod_timer(&rt_secret_timer, now + ip_rt_secret_interval);
 }
 

@@ -2716,7 +2724,7 @@ done:
 
 void ip_rt_multicast_event(struct in_device *in_dev)
 {
-	rt_cache_flush(0);
+	rt_cache_flush(in_dev->dev->nd_net, 0);
 }
 
 #ifdef CONFIG_SYSCTL

@@ -2728,7 +2736,7 @@ static int ipv4_sysctl_rtcache_flush(struct ctl_table *ctl, int write,
 {
 	if (write) {
 		proc_dointvec(ctl, write, filp, buffer, lenp, ppos);
-		rt_cache_flush(flush_delay);
+		rt_cache_flush(current->nsproxy->net_ns, flush_delay);
 		return 0;
 	}
 

@@ -2748,7 +2756,7 @@ static int ipv4_sysctl_rtcache_flush_strategy(struct ctl_table *table,
 		return -EINVAL;
 	if (get_user(delay, (int __user *)newval))
 		return -EFAULT;
-	rt_cache_flush(delay);
+	rt_cache_flush(current->nsproxy->net_ns, delay);
 	return 0;
 }
 

Quoted 11 lines · Denis V. Lunev

commit 628d4768843fc6d572596b91e29f8ce6506c276a

Author: Denis V. Lunev <[email protected]>
Date:   Thu Oct 18 12:48:38 2007 +0400

Add struct net to all calls for rt_cache_flush. Additionally, the
function
manipulating with timer should not actually call the garbage
collector as
the namespace is not known in the global rt_secret_rebuild.

Signed-off-by Denis V. Lunev <[email protected]>

I see you are using current->nsproxy->net_ns for sysctl stuff. Normally you should use ctl->extra2 where the struct net is stored. But in the case of route sysctl, it is not per namespace, so ctl->extra2 is NULL. I will look at that and make it per namespace, so you will be able to build this patch on top of it and remove current->nsproxy->net_ns.

I think that this is quite boring. It is impossible to get the namespace almost anywhere even when it is safe to use this :(

extra2 is a bad place, as it has a conventional meaning for other kernel. For example it is used for quite a lot of sysctl-s anywhere.