Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/kaber/nf-next-2.6

13 files changed, 159 insertions, 56 deletions

diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h
index a8eb43cf0c7e..708009be88b6 100644
--- a/include/net/net_namespace.h
+++ b/include/net/net_namespace.h
@@ -16,6 +16,9 @@
 #include <net/netns/ipv6.h>
 #include <net/netns/dccp.h>
 #include <net/netns/x_tables.h>
+#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
+#include <net/netns/conntrack.h>
+#endif
 
 struct proc_dir_entry;
 struct net_device;
@@ -67,6 +70,9 @@ struct net {
 #endif
 #ifdef CONFIG_NETFILTER
 	struct netns_xt		xt;
+#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
+	struct netns_ct		ct;
+#endif
 #endif
 	struct net_generic	*gen;
 };
diff --git a/include/net/netfilter/ipv4/nf_defrag_ipv4.h b/include/net/netfilter/ipv4/nf_defrag_ipv4.h
new file mode 100644
index 000000000000..6b00ea38546b
--- /dev/null
+++ b/include/net/netfilter/ipv4/nf_defrag_ipv4.h
@@ -0,0 +1,6 @@
+#ifndef _NF_DEFRAG_IPV4_H
+#define _NF_DEFRAG_IPV4_H
+
+extern void nf_defrag_ipv4_enable(void);
+
+#endif /* _NF_DEFRAG_IPV4_H */
diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h
index 0741ad592da0..b76a8685b5b5 100644
--- a/include/net/netfilter/nf_conntrack.h
+++ b/include/net/netfilter/nf_conntrack.h
@@ -123,7 +123,9 @@ struct nf_conn
 
 	/* Extensions */
 	struct nf_ct_ext *ext;
-
+#ifdef CONFIG_NET_NS
+	struct net *ct_net;
+#endif
 	struct rcu_head rcu;
 };
 
@@ -147,6 +149,17 @@ static inline u_int8_t nf_ct_protonum(const struct nf_conn *ct)
 /* get master conntrack via master expectation */
 #define master_ct(conntr) (conntr->master)
 
+extern struct net init_net;
+
+static inline struct net *nf_ct_net(const struct nf_conn *ct)
+{
+#ifdef CONFIG_NET_NS
+	return ct->ct_net;
+#else
+	return &init_net;
+#endif
+}
+
 /* Alter reply tuple (maybe alter helper). */
 extern void
 nf_conntrack_alter_reply(struct nf_conn *ct,
@@ -182,11 +195,11 @@ extern void nf_ct_free_hashtable(struct hlist_head *hash, int vmalloced,
 				 unsigned int size);
 
 extern struct nf_conntrack_tuple_hash *
-__nf_conntrack_find(const struct nf_conntrack_tuple *tuple);
+__nf_conntrack_find(struct net *net, const struct nf_conntrack_tuple *tuple);
 
 extern void nf_conntrack_hash_insert(struct nf_conn *ct);
 
-extern void nf_conntrack_flush(void);
+extern void nf_conntrack_flush(struct net *net);
 
 extern bool nf_ct_get_tuplepr(const struct sk_buff *skb,
 			      unsigned int nhoff, u_int16_t l3num,
@@ -248,10 +261,11 @@ extern struct nf_conn nf_conntrack_untracked;
 
 /* Iterate over all conntracks: if iter returns true, it's deleted. */
 extern void
-nf_ct_iterate_cleanup(int (*iter)(struct nf_conn *i, void *data), void *data);
+nf_ct_iterate_cleanup(struct net *net, int (*iter)(struct nf_conn *i, void *data), void *data);
 extern void nf_conntrack_free(struct nf_conn *ct);
 extern struct nf_conn *
-nf_conntrack_alloc(const struct nf_conntrack_tuple *orig,
+nf_conntrack_alloc(struct net *net,
+		   const struct nf_conntrack_tuple *orig,
 		   const struct nf_conntrack_tuple *repl,
 		   gfp_t gfp);
 
@@ -273,16 +287,14 @@ static inline int nf_ct_is_untracked(const struct sk_buff *skb)
 
 extern int nf_conntrack_set_hashsize(const char *val, struct kernel_param *kp);
 extern unsigned int nf_conntrack_htable_size;
-extern int nf_conntrack_checksum;
-extern atomic_t nf_conntrack_count;
 extern int nf_conntrack_max;
 
-DECLARE_PER_CPU(struct ip_conntrack_stat, nf_conntrack_stat);
-#define NF_CT_STAT_INC(count) (__get_cpu_var(nf_conntrack_stat).count++)
-#define NF_CT_STAT_INC_ATOMIC(count)			\
+#define NF_CT_STAT_INC(net, count)	\
+	(per_cpu_ptr((net)->ct.stat, raw_smp_processor_id())->count++)
+#define NF_CT_STAT_INC_ATOMIC(net, count)		\
 do {							\
 	local_bh_disable();				\
-	__get_cpu_var(nf_conntrack_stat).count++;	\
+	per_cpu_ptr((net)->ct.stat, raw_smp_processor_id())->count++;	\
 	local_bh_enable();				\
 } while (0)
 
diff --git a/include/net/netfilter/nf_conntrack_acct.h b/include/net/netfilter/nf_conntrack_acct.h
index 5d5ae55d54c4..03e218f0be43 100644
--- a/include/net/netfilter/nf_conntrack_acct.h
+++ b/include/net/netfilter/nf_conntrack_acct.h
@@ -8,6 +8,7 @@
 
 #ifndef _NF_CONNTRACK_ACCT_H
 #define _NF_CONNTRACK_ACCT_H
+#include <net/net_namespace.h>
 #include <linux/netfilter/nf_conntrack_common.h>
 #include <linux/netfilter/nf_conntrack_tuple_common.h>
 #include <net/netfilter/nf_conntrack.h>
@@ -18,8 +19,6 @@ struct nf_conn_counter {
 	u_int64_t bytes;
 };
 
-extern int nf_ct_acct;
-
 static inline
 struct nf_conn_counter *nf_conn_acct_find(const struct nf_conn *ct)
 {
@@ -29,9 +28,10 @@ struct nf_conn_counter *nf_conn_acct_find(const struct nf_conn *ct)
 static inline
 struct nf_conn_counter *nf_ct_acct_ext_add(struct nf_conn *ct, gfp_t gfp)
 {
+	struct net *net = nf_ct_net(ct);
 	struct nf_conn_counter *acct;
 
-	if (!nf_ct_acct)
+	if (!net->ct.sysctl_acct)
 		return NULL;
 
 	acct = nf_ct_ext_add(ct, NF_CT_EXT_ACCT, gfp);
@@ -45,7 +45,7 @@ struct nf_conn_counter *nf_ct_acct_ext_add(struct nf_conn *ct, gfp_t gfp)
 extern unsigned int
 seq_print_acct(struct seq_file *s, const struct nf_conn *ct, int dir);
 
-extern int nf_conntrack_acct_init(void);
-extern void nf_conntrack_acct_fini(void);
+extern int nf_conntrack_acct_init(struct net *net);
+extern void nf_conntrack_acct_fini(struct net *net);
 
 #endif /* _NF_CONNTRACK_ACCT_H */
diff --git a/include/net/netfilter/nf_conntrack_core.h b/include/net/netfilter/nf_conntrack_core.h
index a81771210934..e78afe7f28e3 100644
--- a/include/net/netfilter/nf_conntrack_core.h
+++ b/include/net/netfilter/nf_conntrack_core.h
@@ -20,12 +20,13 @@
 /* This header is used to share core functionality between the
    standalone connection tracking module, and the compatibility layer's use
    of connection tracking. */
-extern unsigned int nf_conntrack_in(int pf,
+extern unsigned int nf_conntrack_in(struct net *net,
+				    u_int8_t pf,
 				    unsigned int hooknum,
 				    struct sk_buff *skb);
 
-extern int nf_conntrack_init(void);
-extern void nf_conntrack_cleanup(void);
+extern int nf_conntrack_init(struct net *net);
+extern void nf_conntrack_cleanup(struct net *net);
 
 extern int nf_conntrack_proto_init(void);
 extern void nf_conntrack_proto_fini(void);
@@ -48,7 +49,7 @@ nf_ct_invert_tuple(struct nf_conntrack_tuple *inverse,
 
 /* Find a connection corresponding to a tuple. */
 extern struct nf_conntrack_tuple_hash *
-nf_conntrack_find_get(const struct nf_conntrack_tuple *tuple);
+nf_conntrack_find_get(struct net *net, const struct nf_conntrack_tuple *tuple);
 
 extern int __nf_conntrack_confirm(struct sk_buff *skb);
 
@@ -71,8 +72,6 @@ print_tuple(struct seq_file *s, const struct nf_conntrack_tuple *tuple,
             const struct nf_conntrack_l3proto *l3proto,
             const struct nf_conntrack_l4proto *proto);
 
-extern struct hlist_head *nf_conntrack_hash;
 extern spinlock_t nf_conntrack_lock ;
-extern struct hlist_head unconfirmed;
 
 #endif /* _NF_CONNTRACK_CORE_H */
diff --git a/include/net/netfilter/nf_conntrack_ecache.h b/include/net/netfilter/nf_conntrack_ecache.h
index f0b9078235c9..35f814c1e2ca 100644
--- a/include/net/netfilter/nf_conntrack_ecache.h
+++ b/include/net/netfilter/nf_conntrack_ecache.h
@@ -8,6 +8,7 @@
 
 #include <linux/notifier.h>
 #include <linux/interrupt.h>
+#include <net/net_namespace.h>
 #include <net/netfilter/nf_conntrack_expect.h>
 
 #ifdef CONFIG_NF_CONNTRACK_EVENTS
@@ -15,9 +16,6 @@ struct nf_conntrack_ecache {
 	struct nf_conn *ct;
 	unsigned int events;
 };
-DECLARE_PER_CPU(struct nf_conntrack_ecache, nf_conntrack_ecache);
-
-#define CONNTRACK_ECACHE(x)	(__get_cpu_var(nf_conntrack_ecache).x)
 
 extern struct atomic_notifier_head nf_conntrack_chain;
 extern int nf_conntrack_register_notifier(struct notifier_block *nb);
@@ -25,17 +23,16 @@ extern int nf_conntrack_unregister_notifier(struct notifier_block *nb);
 
 extern void nf_ct_deliver_cached_events(const struct nf_conn *ct);
 extern void __nf_ct_event_cache_init(struct nf_conn *ct);
-extern void nf_ct_event_cache_flush(void);
+extern void nf_ct_event_cache_flush(struct net *net);
 
 static inline void
-nf_conntrack_event_cache(enum ip_conntrack_events event,
-			 const struct sk_buff *skb)
+nf_conntrack_event_cache(enum ip_conntrack_events event, struct nf_conn *ct)
 {
-	struct nf_conn *ct = (struct nf_conn *)skb->nfct;
+	struct net *net = nf_ct_net(ct);
 	struct nf_conntrack_ecache *ecache;
 
 	local_bh_disable();
-	ecache = &__get_cpu_var(nf_conntrack_ecache);
+	ecache = per_cpu_ptr(net->ct.ecache, raw_smp_processor_id());
 	if (ct != ecache->ct)
 		__nf_ct_event_cache_init(ct);
 	ecache->events |= event;
@@ -60,6 +57,9 @@ nf_ct_expect_event(enum ip_conntrack_expect_events event,
 	atomic_notifier_call_chain(&nf_ct_expect_chain, event, exp);
 }
 
+extern int nf_conntrack_ecache_init(struct net *net);
+extern void nf_conntrack_ecache_fini(struct net *net);
+
 #else /* CONFIG_NF_CONNTRACK_EVENTS */
 
 static inline void nf_conntrack_event_cache(enum ip_conntrack_events event,
@@ -69,7 +69,15 @@ static inline void nf_conntrack_event(enum ip_conntrack_events event,
 static inline void nf_ct_deliver_cached_events(const struct nf_conn *ct) {}
 static inline void nf_ct_expect_event(enum ip_conntrack_expect_events event,
 				      struct nf_conntrack_expect *exp) {}
-static inline void nf_ct_event_cache_flush(void) {}
+static inline void nf_ct_event_cache_flush(struct net *net) {}
+
+static inline int nf_conntrack_ecache_init(struct net *net)
+{
+	return 0;
+
+static inline void nf_conntrack_ecache_fini(struct net *net)
+{
+}
 #endif /* CONFIG_NF_CONNTRACK_EVENTS */
 
 #endif /*_NF_CONNTRACK_ECACHE_H*/
diff --git a/include/net/netfilter/nf_conntrack_expect.h b/include/net/netfilter/nf_conntrack_expect.h
index dfdf4b459475..37a7fc1164b0 100644
--- a/include/net/netfilter/nf_conntrack_expect.h
+++ b/include/net/netfilter/nf_conntrack_expect.h
@@ -6,7 +6,6 @@
 #define _NF_CONNTRACK_EXPECT_H
 #include <net/netfilter/nf_conntrack.h>
 
-extern struct hlist_head *nf_ct_expect_hash;
 extern unsigned int nf_ct_expect_hsize;
 extern unsigned int nf_ct_expect_max;
 
@@ -56,6 +55,15 @@ struct nf_conntrack_expect
 	struct rcu_head rcu;
 };
 
+static inline struct net *nf_ct_exp_net(struct nf_conntrack_expect *exp)
+{
+#ifdef CONFIG_NET_NS
+	return exp->master->ct_net;	/* by definition */
+#else
+	return &init_net;
+#endif
+}
+
 struct nf_conntrack_expect_policy
 {
 	unsigned int	max_expected;
@@ -67,17 +75,17 @@ struct nf_conntrack_expect_policy
 #define NF_CT_EXPECT_PERMANENT	0x1
 #define NF_CT_EXPECT_INACTIVE	0x2
 
-int nf_conntrack_expect_init(void);
-void nf_conntrack_expect_fini(void);
+int nf_conntrack_expect_init(struct net *net);
+void nf_conntrack_expect_fini(struct net *net);
 
 struct nf_conntrack_expect *
-__nf_ct_expect_find(const struct nf_conntrack_tuple *tuple);
+__nf_ct_expect_find(struct net *net, const struct nf_conntrack_tuple *tuple);
 
 struct nf_conntrack_expect *
-nf_ct_expect_find_get(const struct nf_conntrack_tuple *tuple);
+nf_ct_expect_find_get(struct net *net, const struct nf_conntrack_tuple *tuple);
 
 struct nf_conntrack_expect *
-nf_ct_find_expectation(const struct nf_conntrack_tuple *tuple);
+nf_ct_find_expectation(struct net *net, const struct nf_conntrack_tuple *tuple);
 
 void nf_ct_unlink_expect(struct nf_conntrack_expect *exp);
 void nf_ct_remove_expectations(struct nf_conn *ct);
@@ -86,7 +94,7 @@ void nf_ct_unexpect_related(struct nf_conntrack_expect *exp);
 /* Allocate space for an expectation: this is mandatory before calling
    nf_ct_expect_related.  You will have to call put afterwards. */
 struct nf_conntrack_expect *nf_ct_expect_alloc(struct nf_conn *me);
-void nf_ct_expect_init(struct nf_conntrack_expect *, unsigned int, int,
+void nf_ct_expect_init(struct nf_conntrack_expect *, unsigned int, u_int8_t,
 		       const union nf_inet_addr *,
 		       const union nf_inet_addr *,
 		       u_int8_t, const __be16 *, const __be16 *);
diff --git a/include/net/netfilter/nf_conntrack_l4proto.h b/include/net/netfilter/nf_conntrack_l4proto.h
index 723df9d1cc35..7f2f43c77284 100644
--- a/include/net/netfilter/nf_conntrack_l4proto.h
+++ b/include/net/netfilter/nf_conntrack_l4proto.h
@@ -39,7 +39,7 @@ struct nf_conntrack_l4proto
 		      const struct sk_buff *skb,
 		      unsigned int dataoff,
 		      enum ip_conntrack_info ctinfo,
-		      int pf,
+		      u_int8_t pf,
 		      unsigned int hooknum);
 
 	/* Called when a new connection for this protocol found;
@@ -50,9 +50,9 @@ struct nf_conntrack_l4proto
 	/* Called when a conntrack entry is destroyed */
 	void (*destroy)(struct nf_conn *ct);
 
-	int (*error)(struct sk_buff *skb, unsigned int dataoff,
+	int (*error)(struct net *net, struct sk_buff *skb, unsigned int dataoff,
 		     enum ip_conntrack_info *ctinfo,
-		     int pf, unsigned int hooknum);
+		     u_int8_t pf, unsigned int hooknum);
 
 	/* Print out the per-protocol part of the tuple. Return like seq_* */
 	int (*print_tuple)(struct seq_file *s,
@@ -117,20 +117,19 @@ extern int nf_ct_port_nlattr_to_tuple(struct nlattr *tb[],
 				      struct nf_conntrack_tuple *t);
 extern const struct nla_policy nf_ct_port_nla_policy[];
 
-/* Log invalid packets */
-extern unsigned int nf_ct_log_invalid;
-
 #ifdef CONFIG_SYSCTL
 #ifdef DEBUG_INVALID_PACKETS
-#define LOG_INVALID(proto) \
-	(nf_ct_log_invalid == (proto) || nf_ct_log_invalid == IPPROTO_RAW)
+#define LOG_INVALID(net, proto)				\
+	((net)->ct.sysctl_log_invalid == (proto) ||	\
+	 (net)->ct.sysctl_log_invalid == IPPROTO_RAW)
 #else
-#define LOG_INVALID(proto) \
-	((nf_ct_log_invalid == (proto) || nf_ct_log_invalid == IPPROTO_RAW) \
+#define LOG_INVALID(net, proto)				\
+	(((net)->ct.sysctl_log_invalid == (proto) ||	\
+	  (net)->ct.sysctl_log_invalid == IPPROTO_RAW)	\
 	 && net_ratelimit())
 #endif
 #else
-#define LOG_INVALID(proto) 0
+#define LOG_INVALID(net, proto) 0
 #endif /* CONFIG_SYSCTL */
 
 #endif /*_NF_CONNTRACK_PROTOCOL_H*/
diff --git a/include/net/netfilter/nf_log.h b/include/net/netfilter/nf_log.h
index 8c6b5ae45534..7182c06974f4 100644
--- a/include/net/netfilter/nf_log.h
+++ b/include/net/netfilter/nf_log.h
@@ -28,7 +28,7 @@ struct nf_loginfo {
 	} u;
 };
 
-typedef void nf_logfn(unsigned int pf,
+typedef void nf_logfn(u_int8_t pf,
 		      unsigned int hooknum,
 		      const struct sk_buff *skb,
 		      const struct net_device *in,
@@ -43,12 +43,12 @@ struct nf_logger {
 };
 
 /* Function to register/unregister log function. */
-int nf_log_register(int pf, const struct nf_logger *logger);
+int nf_log_register(u_int8_t pf, const struct nf_logger *logger);
 void nf_log_unregister(const struct nf_logger *logger);
-void nf_log_unregister_pf(int pf);
+void nf_log_unregister_pf(u_int8_t pf);
 
 /* Calls the registered backend logging function */
-void nf_log_packet(int pf,
+void nf_log_packet(u_int8_t pf,
 		   unsigned int hooknum,
 		   const struct sk_buff *skb,
 		   const struct net_device *in,
diff --git a/include/net/netfilter/nf_queue.h b/include/net/netfilter/nf_queue.h
index d030044e9235..252fd1010b77 100644
--- a/include/net/netfilter/nf_queue.h
+++ b/include/net/netfilter/nf_queue.h
@@ -8,7 +8,7 @@ struct nf_queue_entry {
 	unsigned int		id;
 
 	struct nf_hook_ops	*elem;
-	int			pf;
+	u_int8_t		pf;
 	unsigned int		hook;
 	struct net_device	*indev;
 	struct net_device	*outdev;
@@ -24,9 +24,9 @@ struct nf_queue_handler {
 	char			*name;
 };
 
-extern int nf_register_queue_handler(int pf,
+extern int nf_register_queue_handler(u_int8_t pf,
 				     const struct nf_queue_handler *qh);
-extern int nf_unregister_queue_handler(int pf,
+extern int nf_unregister_queue_handler(u_int8_t pf,
 				       const struct nf_queue_handler *qh);
 extern void nf_unregister_queue_handlers(const struct nf_queue_handler *qh);
 extern void nf_reinject(struct nf_queue_entry *entry, unsigned int verdict);
diff --git a/include/net/netfilter/nf_tproxy_core.h b/include/net/netfilter/nf_tproxy_core.h
new file mode 100644
index 000000000000..208b46f4d6d2
--- /dev/null
+++ b/include/net/netfilter/nf_tproxy_core.h
@@ -0,0 +1,32 @@
+#ifndef _NF_TPROXY_CORE_H
+#define _NF_TPROXY_CORE_H
+
+#include <linux/types.h>
+#include <linux/in.h>
+#include <linux/skbuff.h>
+#include <net/sock.h>
+#include <net/inet_sock.h>
+#include <net/tcp.h>
+
+/* look up and get a reference to a matching socket */
+extern struct sock *
+nf_tproxy_get_sock_v4(struct net *net, const u8 protocol,
+		      const __be32 saddr, const __be32 daddr,
+		      const __be16 sport, const __be16 dport,
+		      const struct net_device *in, bool listening);
+
+static inline void
+nf_tproxy_put_sock(struct sock *sk)
+{
+	/* TIME_WAIT inet sockets have to be handled differently */
+	if ((sk->sk_protocol == IPPROTO_TCP) && (sk->sk_state == TCP_TIME_WAIT))
+		inet_twsk_put(inet_twsk(sk));
+	else
+		sock_put(sk);
+}
+
+/* assign a socket to the skb -- consumes sk */
+int
+nf_tproxy_assign_sock(struct sk_buff *skb, struct sock *sk);
+
+#endif
diff --git a/include/net/netns/conntrack.h b/include/net/netns/conntrack.h
new file mode 100644
index 000000000000..f4498a62881b
--- /dev/null
+++ b/include/net/netns/conntrack.h
@@ -0,0 +1,30 @@
+#ifndef __NETNS_CONNTRACK_H
+#define __NETNS_CONNTRACK_H
+
+#include <linux/list.h>
+#include <asm/atomic.h>
+
+struct ctl_table_header;
+struct nf_conntrack_ecache;
+
+struct netns_ct {
+	atomic_t		count;
+	unsigned int		expect_count;
+	struct hlist_head	*hash;
+	struct hlist_head	*expect_hash;
+	struct hlist_head	unconfirmed;
+	struct ip_conntrack_stat *stat;
+#ifdef CONFIG_NF_CONNTRACK_EVENTS
+	struct nf_conntrack_ecache *ecache;
+#endif
+	int			sysctl_acct;
+	int			sysctl_checksum;
+	unsigned int		sysctl_log_invalid; /* Log invalid packets */
+#ifdef CONFIG_SYSCTL
+	struct ctl_table_header	*sysctl_header;
+	struct ctl_table_header	*acct_sysctl_header;
+#endif
+	int			hash_vmalloc;
+	int			expect_vmalloc;
+};
+#endif
diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h
index a6ed83853dcc..ece1c926b5d1 100644
--- a/include/net/netns/ipv4.h
+++ b/include/net/netns/ipv4.h
@@ -38,6 +38,9 @@ struct netns_ipv4 {
 	struct xt_table		*iptable_raw;
 	struct xt_table		*arptable_filter;
 	struct xt_table		*iptable_security;
+	struct xt_table		*nat_table;
+	struct hlist_head	*nat_bysource;
+	int			nat_vmalloced;
 #endif
 
 	int sysctl_icmp_echo_ignore_all;