: Revisit read-only-in-namespaces sysctls.

index aa540e6..8df751b 100644
@@ -201,8 +201,11 @@ extern void unregister_pernet_gen_device(int id, struct pernet_operations *);
 struct ctl_path;
 struct ctl_table;
 struct ctl_table_header;
+
 extern struct ctl_table_header *register_net_sysctl_table(struct net *net,
 	const struct ctl_path *path, struct ctl_table *table);
+extern struct ctl_table_header *register_net_sysctl_rotable(
+	const struct ctl_path *path, struct ctl_table *table);
 extern void unregister_net_sysctl_table(struct ctl_table_header *header);
 
 #endif /* __NET_NET_NAMESPACE_H */

index b4f0525..d8e7916 100644
@@ -40,6 +40,27 @@ static struct ctl_table_root net_sysctl_root = {
 	.lookup = net_ctl_header_lookup,
 };
 
+static LIST_HEAD(net_sysctl_ro_tables);
+static struct list_head *net_ctl_ro_header_lookup(struct ctl_table_root *root,
+		struct nsproxy *namespaces)
+{
+	return &net_sysctl_ro_tables;
+}
+
+static int net_ctl_ro_header_perms(struct ctl_table_root *root,
+		struct nsproxy *namespaces, struct ctl_table *table)
+{
+	if (namespaces->net_ns == &init_net)
+		return table->mode;
+	else
+		return table->mode & ~0222;
+}
+
+static struct ctl_table_root net_sysctl_ro_root = {
+	.lookup = net_ctl_ro_header_lookup,
+	.permissions = net_ctl_ro_header_perms,
+};
+
 static int sysctl_net_init(struct net *net)
 {
 	INIT_LIST_HEAD(&net->sysctl_table_headers);

@@ -64,6 +85,7 @@ static __init int sysctl_init(void)
 	if (ret)
 		goto out;
 	register_sysctl_root(&net_sysctl_root);
+	register_sysctl_root(&net_sysctl_ro_root);
 out:
 	return ret;
 }

@@ -80,6 +102,14 @@ struct ctl_table_header *register_net_sysctl_table(struct net *net,
 }
 EXPORT_SYMBOL_GPL(register_net_sysctl_table);
 
+struct ctl_table_header *register_net_sysctl_rotable(const
+		struct ctl_path *path, struct ctl_table *table)
+{
+	return __register_sysctl_paths(&net_sysctl_ro_root,
+			&init_nsproxy, path, table);
+}
+EXPORT_SYMBOL_GPL(register_net_sysctl_rotable);
+
 void unregister_net_sysctl_table(struct ctl_table_header *header)
 {
 	unregister_sysctl_table(header);

index 5fc8010..a570e2a 100644
@@ -125,14 +125,6 @@ static struct ctl_table net_core_table[] = {
 #endif /* CONFIG_XFRM */
 #endif /* CONFIG_NET */
 	{
-		.ctl_name	= NET_CORE_SOMAXCONN,
-		.procname	= "somaxconn",
-		.data		= &init_net.core.sysctl_somaxconn,
-		.maxlen		= sizeof(int),
-		.mode		= 0644,
-		.proc_handler	= &proc_dointvec
-	},
-	{
 		.ctl_name	= NET_CORE_BUDGET,
 		.procname	= "netdev_budget",
 		.data		= &netdev_budget,

@@ -151,6 +143,18 @@ static struct ctl_table net_core_table[] = {
 	{ .ctl_name = 0 }
 };
 
+static struct ctl_table netns_core_table[] = {
+	{
+		.ctl_name	= NET_CORE_SOMAXCONN,
+		.procname	= "somaxconn",
+		.data		= &init_net.core.sysctl_somaxconn,
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.proc_handler	= &proc_dointvec
+	},
+	{ .ctl_name = 0 }
+};
+
 static __net_initdata struct ctl_path net_core_path[] = {
 	{ .procname = "net", .ctl_name = CTL_NET, },
 	{ .procname = "core", .ctl_name = NET_CORE, },

@@ -159,23 +163,17 @@ static __net_initdata struct ctl_path net_core_path[] = {
 
 static __net_init int sysctl_core_net_init(struct net *net)
 {
-	struct ctl_table *tbl, *tmp;
+	struct ctl_table *tbl;
 
 	net->core.sysctl_somaxconn = SOMAXCONN;
 
-	tbl = net_core_table;
+	tbl = netns_core_table;
 	if (net != &init_net) {
-		tbl = kmemdup(tbl, sizeof(net_core_table), GFP_KERNEL);
+		tbl = kmemdup(tbl, sizeof(netns_core_table), GFP_KERNEL);
 		if (tbl == NULL)
 			goto err_dup;
 
-		for (tmp = tbl; tmp->procname; tmp++) {
-			if (tmp->data >= (void *)&init_net &&
-					tmp->data < (void *)(&init_net + 1))
-				tmp->data += (char *)net - (char *)&init_net;
-			else
-				tmp->mode &= ~0222;
-		}
+		tbl[0].data = &net->core.sysctl_somaxconn;
 	}
 
 	net->core.sysctl_hdr = register_net_sysctl_table(net,

@@ -186,7 +184,7 @@ static __net_init int sysctl_core_net_init(struct net *net)
 	return 0;
 
 err_reg:
-	if (tbl != net_core_table)
+	if (tbl != netns_core_table)
 		kfree(tbl);
 err_dup:
 	return -ENOMEM;

@@ -198,7 +196,7 @@ static __net_exit void sysctl_core_net_exit(struct net *net)
 
 	tbl = net->core.sysctl_hdr->ctl_table_arg;
 	unregister_net_sysctl_table(net->core.sysctl_hdr);
-	BUG_ON(tbl == net_core_table);
+	BUG_ON(tbl == netns_core_table);
 	kfree(tbl);
 }
 

@@ -209,6 +207,7 @@ static __net_initdata struct pernet_operations sysctl_core_ops = {
 
 static __init int sysctl_core_init(void)
 {
+	register_net_sysctl_rotable(net_core_path, net_core_table);
 	return register_pernet_subsys(&sysctl_core_ops);
 }
 

index cd6ce6a..7f102ee 100644
@@ -598,7 +598,7 @@ int ip_defrag(struct sk_buff *skb, u32 user)
 #ifdef CONFIG_SYSCTL
 static int zero;
 
-static struct ctl_table ip4_frags_ctl_table[] = {
+static struct ctl_table ip4_frags_ns_ctl_table[] = {
 	{
 		.ctl_name	= NET_IPV4_IPFRAG_HIGH_THRESH,
 		.procname	= "ipfrag_high_thresh",

@@ -644,14 +644,14 @@ static struct ctl_table ip4_frags_ctl_table[] = {
 	{ }
 };
 
-static int ip4_frags_ctl_register(struct net *net)
+static int ip4_frags_ns_ctl_register(struct net *net)
 {
 	struct ctl_table *table;
 	struct ctl_table_header *hdr;
 
-	table = ip4_frags_ctl_table;
+	table = ip4_frags_ns_ctl_table;
 	if (net != &init_net) {
-		table = kmemdup(table, sizeof(ip4_frags_ctl_table), GFP_KERNEL);
+		table = kmemdup(table, sizeof(ip4_frags_ns_ctl_table), GFP_KERNEL);
 		if (table == NULL)
 			goto err_alloc;
 

@@ -676,7 +676,7 @@ err_alloc:
 	return -ENOMEM;
 }
 
-static void ip4_frags_ctl_unregister(struct net *net)
+static void ip4_frags_ns_ctl_unregister(struct net *net)
 {
 	struct ctl_table *table;
 

@@ -685,12 +685,12 @@ static void ip4_frags_ctl_unregister(struct net *net)
 	kfree(table);
 }
 #else
-static inline int ip4_frags_ctl_register(struct net *net)
+static inline int ip4_frags_ns_ctl_register(struct net *net)
 {
 	return 0;
 }
 
-static inline void ip4_frags_ctl_unregister(struct net *net)
+static inline void ip4_frags_ns_ctl_unregister(struct net *net)
 {
 }
 #endif

@@ -714,12 +714,12 @@ static int ipv4_frags_init_net(struct net *net)
 
 	inet_frags_init_net(&net->ipv4.frags);
 
-	return ip4_frags_ctl_register(net);
+	return ip4_frags_ns_ctl_register(net);
 }
 
 static void ipv4_frags_exit_net(struct net *net)
 {
-	ip4_frags_ctl_unregister(net);
+	ip4_frags_ns_ctl_unregister(net);
 	inet_frags_exit_net(&net->ipv4.frags, &ip4_frags);
 }
 

index 798cabc..7e008de 100644
@@ -632,7 +632,7 @@ static struct inet6_protocol frag_protocol =
 };
 
 #ifdef CONFIG_SYSCTL
-static struct ctl_table ip6_frags_ctl_table[] = {
+static struct ctl_table ip6_frags_ns_ctl_table[] = {
 	{
 		.ctl_name	= NET_IPV6_IP6FRAG_HIGH_THRESH,
 		.procname	= "ip6frag_high_thresh",

@@ -670,14 +670,14 @@ static struct ctl_table ip6_frags_ctl_table[] = {
 	{ }
 };
 
-static int ip6_frags_sysctl_register(struct net *net)
+static int ip6_frags_ns_sysctl_register(struct net *net)
 {
 	struct ctl_table *table;
 	struct ctl_table_header *hdr;
 
-	table = ip6_frags_ctl_table;
+	table = ip6_frags_ns_ctl_table;
 	if (net != &init_net) {
-		table = kmemdup(table, sizeof(ip6_frags_ctl_table), GFP_KERNEL);
+		table = kmemdup(table, sizeof(ip6_frags_ns_ctl_table), GFP_KERNEL);
 		if (table == NULL)
 			goto err_alloc;
 

@@ -701,7 +701,7 @@ err_alloc:
 	return -ENOMEM;
 }
 
-static void ip6_frags_sysctl_unregister(struct net *net)
+static void ip6_frags_ns_sysctl_unregister(struct net *net)
 {
 	struct ctl_table *table;
 

@@ -710,12 +710,12 @@ static void ip6_frags_sysctl_unregister(struct net *net)
 	kfree(table);
 }
 #else
-static inline int ip6_frags_sysctl_register(struct net *net)
+static inline int ip6_frags_ns_sysctl_register(struct net *net)
 {
 	return 0;
 }
 
-static inline void ip6_frags_sysctl_unregister(struct net *net)
+static inline void ip6_frags_ns_sysctl_unregister(struct net *net)
 {
 }
 #endif

@@ -728,12 +728,12 @@ static int ipv6_frags_init_net(struct net *net)
 
 	inet_frags_init_net(&net->ipv6.frags);
 
-	return ip6_frags_sysctl_register(net);
+	return ip6_frags_ns_sysctl_register(net);
 }
 
 static void ipv6_frags_exit_net(struct net *net)
 {
-	ip6_frags_sysctl_unregister(net);
+	ip6_frags_ns_sysctl_unregister(net);
 	inet_frags_exit_net(&net->ipv6.frags, &ip6_frags);
 }
 

index 7e008de..130d6f6 100644
@@ -750,7 +750,9 @@ int __init ipv6_frag_init(void)
 	if (ret)
 		goto out;
 
-	register_pernet_subsys(&ip6_frags_ops);
+	ret = register_pernet_subsys(&ip6_frags_ops);
+	if (ret)
+		goto err_pernet;
 
 	ip6_frags.hashfn = ip6_hashfn;
 	ip6_frags.constructor = ip6_frag_init;

@@ -763,6 +765,10 @@ int __init ipv6_frag_init(void)
 	inet_frags_init(&ip6_frags);
 out:
 	return ret;
+
+err_pernet:
+	inet6_del_protocol(&frag_protocol, IPPROTO_FRAGMENT);
+	goto out;
 }
 
 void ipv6_frag_exit(void)

index 7f102ee..be1cb89 100644
@@ -624,6 +624,10 @@ static struct ctl_table ip4_frags_ns_ctl_table[] = {
 		.proc_handler	= &proc_dointvec_jiffies,
 		.strategy	= &sysctl_jiffies
 	},
+	{ }
+};
+
+static struct ctl_table ip4_frags_ctl_table[] = {
 	{
 		.ctl_name	= NET_IPV4_IPFRAG_SECRET_INTERVAL,
 		.procname	= "ipfrag_secret_interval",

@@ -658,8 +662,6 @@ static int ip4_frags_ns_ctl_register(struct net *net)
 		table[0].data = &net->ipv4.frags.high_thresh;
 		table[1].data = &net->ipv4.frags.low_thresh;
 		table[2].data = &net->ipv4.frags.timeout;
-		table[3].mode &= ~0222;
-		table[4].mode &= ~0222;
 	}
 
 	hdr = register_net_sysctl_table(net, net_ipv4_ctl_path, table);

@@ -684,6 +686,11 @@ static void ip4_frags_ns_ctl_unregister(struct net *net)
 	unregister_net_sysctl_table(net->ipv4.frags_hdr);
 	kfree(table);
 }
+
+static void ip4_frags_ctl_register(void)
+{
+	register_net_sysctl_rotable(net_ipv4_ctl_path, ip4_frags_ctl_table);
+}
 #else
 static inline int ip4_frags_ns_ctl_register(struct net *net)
 {

@@ -693,6 +700,10 @@ static inline int ip4_frags_ns_ctl_register(struct net *net)
 static inline void ip4_frags_ns_ctl_unregister(struct net *net)
 {
 }
+
+static inline void ip4_frags_ctl_register(void)
+{
+}
 #endif
 
 static int ipv4_frags_init_net(struct net *net)

@@ -730,6 +741,7 @@ static struct pernet_operations ip4_frags_ops = {
 
 void __init ipfrag_init(void)
 {
+	ip4_frags_ctl_register();
 	register_pernet_subsys(&ip4_frags_ops);
 	ip4_frags.hashfn = ip4_hashfn;
 	ip4_frags.constructor = ip4_frag_init;

index 130d6f6..9391a69 100644
@@ -658,6 +658,10 @@ static struct ctl_table ip6_frags_ns_ctl_table[] = {
 		.proc_handler	= &proc_dointvec_jiffies,
 		.strategy	= &sysctl_jiffies,
 	},
+	{ }
+};
+
+static struct ctl_table ip6_frags_ctl_table[] = {
 	{
 		.ctl_name	= NET_IPV6_IP6FRAG_SECRET_INTERVAL,
 		.procname	= "ip6frag_secret_interval",

@@ -684,7 +688,6 @@ static int ip6_frags_ns_sysctl_register(struct net *net)
 		table[0].data = &net->ipv6.frags.high_thresh;
 		table[1].data = &net->ipv6.frags.low_thresh;
 		table[2].data = &net->ipv6.frags.timeout;
-		table[3].mode &= ~0222;
 	}
 
 	hdr = register_net_sysctl_table(net, net_ipv6_ctl_path, table);

@@ -709,6 +712,20 @@ static void ip6_frags_ns_sysctl_unregister(struct net *net)
 	unregister_net_sysctl_table(net->ipv6.sysctl.frags_hdr);
 	kfree(table);
 }
+
+static struct ctl_table_header *ip6_ctl_header;
+
+static int ip6_frags_sysctl_register(void)
+{
+	ip6_ctl_header = register_net_sysctl_rotable(net_ipv6_ctl_path,
+			ip6_frags_ctl_table);
+	return ip6_ctl_header == NULL ? -ENOMEM : 0;
+}
+
+static void ip6_frags_sysctl_unregister(void)
+{
+	unregister_net_sysctl_table(ip6_ctl_header);
+}
 #else
 static inline int ip6_frags_ns_sysctl_register(struct net *net)
 {

@@ -718,6 +735,15 @@ static inline int ip6_frags_ns_sysctl_register(struct net *net)
 static inline void ip6_frags_ns_sysctl_unregister(struct net *net)
 {
 }
+
+static inline int ip6_frags_sysctl_register(void)
+{
+	return 0;
+}
+
+static inline void ip6_frags_sysctl_unregister(void)
+{
+}
 #endif
 
 static int ipv6_frags_init_net(struct net *net)

@@ -750,6 +776,10 @@ int __init ipv6_frag_init(void)
 	if (ret)
 		goto out;
 
+	ret = ip6_frags_sysctl_register();
+	if (ret)
+		goto err_sysctl;
+
 	ret = register_pernet_subsys(&ip6_frags_ops);
 	if (ret)
 		goto err_pernet;

@@ -767,6 +797,8 @@ out:
 	return ret;
 
 err_pernet:
+	ip6_frags_sysctl_unregister();
+err_sysctl:
 	inet6_del_protocol(&frag_protocol, IPPROTO_FRAGMENT);
 	goto out;
 }

@@ -774,6 +806,7 @@ err_pernet:
 void ipv6_frag_exit(void)
 {
 	inet_frags_fini(&ip6_frags);
+	ip6_frags_sysctl_unregister();
 	unregister_pernet_subsys(&ip6_frags_ops);
 	inet6_del_protocol(&frag_protocol, IPPROTO_FRAGMENT);
 }

index 3804dcb..5c99274 100644
@@ -37,6 +37,10 @@ static ctl_table ipv6_table_template[] = {
 		.mode		= 0644,
 		.proc_handler	= &proc_dointvec
 	},
+	{ .ctl_name = 0 }
+};
+
+static ctl_table ipv6_table[] = {
 	{
 		.ctl_name	= NET_IPV6_MLD_MAX_MSF,
 		.procname	= "mld_max_msf",

@@ -80,12 +84,6 @@ static int ipv6_sysctl_net_init(struct net *net)
 
 	ipv6_table[2].data = &net->ipv6.sysctl.bindv6only;
 
-	/* We don't want this value to be per namespace, it should be global
-	   to all namespaces, so make it read-only when we are not in the
-	   init network namespace */
-	if (net != &init_net)
-		ipv6_table[3].mode = 0444;
-
 	net->ipv6.sysctl.table = register_net_sysctl_table(net, net_ipv6_ctl_path,
 							   ipv6_table);
 	if (!net->ipv6.sysctl.table)

@@ -126,12 +124,29 @@ static struct pernet_operations ipv6_sysctl_net_ops = {
 	.exit = ipv6_sysctl_net_exit,
 };
 
+static struct ctl_table_header *ip6_header;
+
 int ipv6_sysctl_register(void)
 {
-	return register_pernet_subsys(&ipv6_sysctl_net_ops);
+	int err = -ENOMEM;;
+
+	ip6_header = register_net_sysctl_rotable(net_ipv6_ctl_path, ipv6_table);
+	if (ip6_header == NULL)
+		goto out;
+
+	err = register_pernet_subsys(&ipv6_sysctl_net_ops);
+	if (err)
+		goto err_pernet;
+out:
+	return err;
+
+err_pernet:
+	unregister_net_sysctl_table(ip6_header);
+	goto out;
 }
 
 void ipv6_sysctl_unregister(void)
 {
+	unregister_net_sysctl_table(ip6_header);
 	unregister_pernet_subsys(&ipv6_sysctl_net_ops);
 }