Merge nuts.davemloft.net:/disk1/BK/nf-work-2.6

into nuts.davemloft.net:/disk1/BK/nf-2.6

include/linux/netfilter.h

@@ -178,7 +178,7 @@ extern inline struct ip6t_target *
 ip6t_find_target_lock(const char *name, int *error, struct semaphore *mutex);
 extern inline struct arpt_target *
 arpt_find_target_lock(const char *name, int *error, struct semaphore *mutex);
-extern void (*ip_ct_attach)(struct sk_buff *, struct nf_ct_info *);
+extern void (*ip_ct_attach)(struct sk_buff *, struct sk_buff *);
 
 #ifdef CONFIG_NETFILTER_DEBUG
 extern void nf_dump_skb(int pf, struct sk_buff *skb);

include/linux/netfilter_ipv4/ip_conntrack.h

@@ -172,9 +172,6 @@ struct ip_conntrack
            plus 1 for any connection(s) we are `master' for */
 	struct nf_conntrack ct_general;
 
-	/* These are my tuples; original and reply */
-	struct ip_conntrack_tuple_hash tuplehash[IP_CT_DIR_MAX];
-
 	/* Have we seen traffic both ways yet? (bitset) */
 	unsigned long status;
 
@@ -199,12 +196,7 @@ struct ip_conntrack
 	/* Helper, if any. */
 	struct ip_conntrack_helper *helper;
 
-	/* Our various nf_ct_info structs specify *what* relation this
-           packet has to the conntrack */
-	struct nf_ct_info infos[IP_CT_NUMBER];
-
 	/* Storage reserved for other modules: */
-
 	union ip_conntrack_proto proto;
 
 	union ip_conntrack_help help;
@@ -220,6 +212,9 @@ struct ip_conntrack
 	} nat;
 #endif /* CONFIG_IP_NF_NAT_NEEDED */
 
+	/* Traversed often, so hopefully in different cacheline to top */
+	/* These are my tuples; original and reply */
+	struct ip_conntrack_tuple_hash tuplehash[IP_CT_DIR_MAX];
 };
 
 /* get master conntrack via master expectation */
@@ -238,8 +233,12 @@ ip_conntrack_tuple_taken(const struct ip_conntrack_tuple *tuple,
 			 const struct ip_conntrack *ignored_conntrack);
 
 /* Return conntrack_info and tuple hash for given skb. */
-extern struct ip_conntrack *
-ip_conntrack_get(struct sk_buff *skb, enum ip_conntrack_info *ctinfo);
+static inline struct ip_conntrack *
+ip_conntrack_get(const struct sk_buff *skb, enum ip_conntrack_info *ctinfo)
+{
+	*ctinfo = skb->nfctinfo;
+	return (struct ip_conntrack *)skb->nfct;
+}
 
 /* decrement reference count on a conntrack */
 extern inline void ip_conntrack_put(struct ip_conntrack *ct);
@@ -306,12 +305,13 @@ struct ip_conntrack_stat
 	unsigned int insert_failed;
 	unsigned int drop;
 	unsigned int early_drop;
-	unsigned int icmp_error;
+	unsigned int error;
 	unsigned int expect_new;
 	unsigned int expect_create;
 	unsigned int expect_delete;
 };
 
+#define CONNTRACK_STAT_INC(count) (__get_cpu_var(ip_conntrack_stat).count++)
 
 /* eg. PROVIDES_CONNTRACK(ftp); */
 #define PROVIDES_CONNTRACK(name)                        \

include/linux/netfilter_ipv4/ip_conntrack_core.h

@@ -16,10 +16,6 @@ extern int ip_conntrack_init(void);
 extern void ip_conntrack_cleanup(void);
 
 struct ip_conntrack_protocol;
-extern struct ip_conntrack_protocol *ip_ct_find_proto(u_int8_t protocol);
-/* Like above, but you already have conntrack read lock. */
-extern struct ip_conntrack_protocol *__ip_ct_find_proto(u_int8_t protocol);
-extern struct list_head protocol_list;
 
 extern int
 ip_ct_get_tuple(const struct iphdr *iph,
@@ -38,14 +34,14 @@ struct ip_conntrack_tuple_hash *
 ip_conntrack_find_get(const struct ip_conntrack_tuple *tuple,
 		      const struct ip_conntrack *ignored_conntrack);
 
-extern int __ip_conntrack_confirm(struct nf_ct_info *nfct);
+extern int __ip_conntrack_confirm(struct sk_buff *skb);
 
 /* Confirm a connection: returns NF_DROP if packet must be dropped. */
 static inline int ip_conntrack_confirm(struct sk_buff *skb)
 {
 	if (skb->nfct
-	    && !is_confirmed((struct ip_conntrack *)skb->nfct->master))
-		return __ip_conntrack_confirm(skb->nfct);
+	    && !is_confirmed((struct ip_conntrack *)skb->nfct))
+		return __ip_conntrack_confirm(skb);
 	return NF_ACCEPT;
 }
 

include/linux/netfilter_ipv4/ip_conntrack_protocol.h

@@ -7,9 +7,6 @@ struct seq_file;
 
 struct ip_conntrack_protocol
 {
-	/* Next pointer. */
-	struct list_head list;
-
 	/* Protocol number. */
 	u_int8_t proto;
 
@@ -58,14 +55,23 @@ struct ip_conntrack_protocol
 	struct module *me;
 };
 
+#define MAX_IP_CT_PROTO 256
+extern struct ip_conntrack_protocol *ip_ct_protos[MAX_IP_CT_PROTO];
+
 /* Protocol registration. */
 extern int ip_conntrack_protocol_register(struct ip_conntrack_protocol *proto);
 extern void ip_conntrack_protocol_unregister(struct ip_conntrack_protocol *proto);
 
+static inline struct ip_conntrack_protocol *ip_ct_find_proto(u_int8_t protocol)
+{
+	return ip_ct_protos[protocol];
+}
+
 /* Existing built-in protocols */
 extern struct ip_conntrack_protocol ip_conntrack_protocol_tcp;
 extern struct ip_conntrack_protocol ip_conntrack_protocol_udp;
 extern struct ip_conntrack_protocol ip_conntrack_protocol_icmp;
+extern struct ip_conntrack_protocol ip_conntrack_generic_protocol;
 extern int ip_conntrack_protocol_tcp_init(void);
 
 /* Log invalid packets */

include/linux/netfilter_ipv4/ip_nat.h

@@ -80,27 +80,18 @@ struct ip_nat_info_manip
 /* Protects NAT hash tables, and NAT-private part of conntracks. */
 DECLARE_RWLOCK_EXTERN(ip_nat_lock);
 
-/* Hashes for by-source and IP/protocol. */
-struct ip_nat_hash
-{
-	struct list_head list;
-
-	/* conntrack we're embedded in: NULL if not in hash. */
-	struct ip_conntrack *conntrack;
-};
-
 /* The structure embedded in the conntrack structure. */
 struct ip_nat_info
 {
 	/* Set to zero when conntrack created: bitmask of maniptypes */
-	int initialized;
+	u_int16_t initialized;
 
-	unsigned int num_manips;
+	u_int16_t num_manips;
 
 	/* Manipulations to be done on this conntrack. */
 	struct ip_nat_info_manip manips[IP_NAT_MAX_MANIPS];
 
-	struct ip_nat_hash bysource, byipsproto;
+	struct list_head bysource, byipsproto;
 
 	/* Helper (NULL if none). */
 	struct ip_nat_helper *helper;

include/linux/netfilter_ipv4/ip_nat_core.h

@@ -14,8 +14,6 @@ extern unsigned int do_bindings(struct ip_conntrack *ct,
 				unsigned int hooknum,
 				struct sk_buff **pskb);
 
-extern struct list_head protos;
-
 extern int icmp_reply_translation(struct sk_buff **pskb,
 				  struct ip_conntrack *conntrack,
 				  unsigned int hooknum,
@@ -26,8 +24,4 @@ extern void replace_in_hashes(struct ip_conntrack *conntrack,
 extern void place_in_hashes(struct ip_conntrack *conntrack,
 			    struct ip_nat_info *info);
 
-/* Built-in protocols. */
-extern struct ip_nat_protocol ip_nat_protocol_tcp;
-extern struct ip_nat_protocol ip_nat_protocol_udp;
-extern struct ip_nat_protocol ip_nat_protocol_icmp;
 #endif /* _IP_NAT_CORE_H */

include/linux/netfilter_ipv4/ip_nat_protocol.h

@@ -9,8 +9,6 @@ struct ip_nat_range;
 
 struct ip_nat_protocol
 {
-	struct list_head list;
-
 	/* Protocol name */
 	const char *name;
 
@@ -47,10 +45,24 @@ struct ip_nat_protocol
 				    const struct ip_nat_range *range);
 };
 
+#define MAX_IP_NAT_PROTO 256
+extern struct ip_nat_protocol *ip_nat_protos[MAX_IP_NAT_PROTO];
+
 /* Protocol registration. */
 extern int ip_nat_protocol_register(struct ip_nat_protocol *proto);
 extern void ip_nat_protocol_unregister(struct ip_nat_protocol *proto);
 
+static inline struct ip_nat_protocol *ip_nat_find_proto(u_int8_t protocol)
+{
+	return ip_nat_protos[protocol];
+}
+
+/* Built-in protocols. */
+extern struct ip_nat_protocol ip_nat_protocol_tcp;
+extern struct ip_nat_protocol ip_nat_protocol_udp;
+extern struct ip_nat_protocol ip_nat_protocol_icmp;
+extern struct ip_nat_protocol ip_nat_unknown_protocol;
+
 extern int init_protocols(void) __init;
 extern void cleanup_protocols(void);
 extern struct ip_nat_protocol *find_nat_proto(u_int16_t protonum);

include/linux/netfilter_ipv4/ipt_comment.h

+#ifndef _IPT_COMMENT_H
+#define _IPT_COMMENT_H
+
+#define IPT_MAX_COMMENT_LEN 256
+
+struct ipt_comment_info {
+	unsigned char comment[IPT_MAX_COMMENT_LEN];
+};
+
+#endif /* _IPT_COMMENT_H */

include/linux/skbuff.h

@@ -97,10 +97,6 @@ struct nf_conntrack {
 	void (*destroy)(struct nf_conntrack *);
 };
 
-struct nf_ct_info {
-	struct nf_conntrack *master;
-};
-
 #ifdef CONFIG_BRIDGE_NETFILTER
 struct nf_bridge_info {
 	atomic_t use;
@@ -186,6 +182,7 @@ struct skb_shared_info {
  *	@nfmark: Can be used for communication between hooks
  *	@nfcache: Cache info
  *	@nfct: Associated connection, if any
+ *	@nfctinfo: Relationship of this skb to the connection
  *	@nf_debug: Netfilter debugging
  *	@nf_bridge: Saved data about a bridged frame - see br_netfilter.c
  *      @private: Data which is private to the HIPPI implementation
@@ -253,7 +250,8 @@ struct sk_buff {
 #ifdef CONFIG_NETFILTER
         unsigned long		nfmark;
 	__u32			nfcache;
-	struct nf_ct_info	*nfct;
+	struct nf_conntrack	*nfct;
+	__u32			nfctinfo;
 #ifdef CONFIG_NETFILTER_DEBUG
         unsigned int		nf_debug;
 #endif
@@ -1141,15 +1139,15 @@ extern int skb_iter_next(const struct sk_buff *skb, struct skb_iter *i);
 extern void skb_iter_abort(const struct sk_buff *skb, struct skb_iter *i);
 
 #ifdef CONFIG_NETFILTER
-static inline void nf_conntrack_put(struct nf_ct_info *nfct)
+static inline void nf_conntrack_put(struct nf_conntrack *nfct)
 {
-	if (nfct && atomic_dec_and_test(&nfct->master->use))
-		nfct->master->destroy(nfct->master);
+	if (nfct && atomic_dec_and_test(&nfct->use))
+		nfct->destroy(nfct);
 }
-static inline void nf_conntrack_get(struct nf_ct_info *nfct)
+static inline void nf_conntrack_get(struct nf_conntrack *nfct)
 {
 	if (nfct)
-		atomic_inc(&nfct->master->use);
+		atomic_inc(&nfct->use);
 }
 static inline void nf_reset(struct sk_buff *skb)
 {

include/linux/sysctl.h

@@ -424,6 +424,7 @@ enum
  	NET_IPV4_NF_CONNTRACK_SCTP_TIMEOUT_SHUTDOWN_SENT=24,
  	NET_IPV4_NF_CONNTRACK_SCTP_TIMEOUT_SHUTDOWN_RECD=25,
  	NET_IPV4_NF_CONNTRACK_SCTP_TIMEOUT_SHUTDOWN_ACK_SENT=26,
+	NET_IPV4_NF_CONNTRACK_COUNT=27,
 };
  
 /* /proc/sys/net/ipv6 */

net/core/netfilter.c

@@ -806,7 +806,7 @@ EXPORT_SYMBOL(nf_log_packet);
    tracking in use: without this, connection may not be in hash table,
    and hence manufactured ICMP or RST packets will not be associated
    with it. */
-void (*ip_ct_attach)(struct sk_buff *, struct nf_ct_info *);
+void (*ip_ct_attach)(struct sk_buff *, struct sk_buff *);
 
 void __init netfilter_init(void)
 {

net/core/skbuff.c

@@ -311,6 +311,7 @@ struct sk_buff *skb_clone(struct sk_buff *skb, int gfp_mask)
 	C(nfcache);
 	C(nfct);
 	nf_conntrack_get(skb->nfct);
+	C(nfctinfo);
 #ifdef CONFIG_NETFILTER_DEBUG
 	C(nf_debug);
 #endif
@@ -377,6 +378,7 @@ static void copy_skb_header(struct sk_buff *new, const struct sk_buff *old)
 	new->nfcache	= old->nfcache;
 	new->nfct	= old->nfct;
 	nf_conntrack_get(old->nfct);
+	new->nfctinfo	= old->nfctinfo;
 #ifdef CONFIG_NETFILTER_DEBUG
 	new->nf_debug	= old->nf_debug;
 #endif

net/ipv4/ip_output.c

@@ -422,6 +422,7 @@ static void ip_copy_metadata(struct sk_buff *to, struct sk_buff *from)
 	nf_conntrack_put(to->nfct);
 	to->nfct = from->nfct;
 	nf_conntrack_get(to->nfct);
+	to->nfctinfo = from->nfctinfo;
 #ifdef CONFIG_BRIDGE_NETFILTER
 	nf_bridge_put(to->nf_bridge);
 	to->nf_bridge = from->nf_bridge;

net/ipv4/netfilter/Kconfig

@@ -332,6 +332,16 @@ config IP_NF_MATCH_SCTP
 	  If you want to compile it as a module, say M here and read
 	  Documentation/modules.txt.  If unsure, say `N'.
 
+config IP_NF_MATCH_COMMENT
+	tristate  'comment match support'
+	depends on IP_NF_IPTABLES
+	help
+	  This option adds a `comment' dummy-match, which allows you to put
+	  comments in your iptables ruleset.
+
+	  If you want to compile it as a module, say M here and read
+	  Documentation/modules.txt.  If unsure, say `N'.
+
 # `filter', generic and specific targets
 config IP_NF_FILTER
 	tristate "Packet filtering"

net/ipv4/netfilter/Makefile

@@ -50,28 +50,23 @@ obj-$(CONFIG_IP_NF_MATCH_SCTP) += ipt_sctp.o
 obj-$(CONFIG_IP_NF_MATCH_MARK) += ipt_mark.o
 obj-$(CONFIG_IP_NF_MATCH_MAC) += ipt_mac.o
 obj-$(CONFIG_IP_NF_MATCH_IPRANGE) += ipt_iprange.o
-
 obj-$(CONFIG_IP_NF_MATCH_PKTTYPE) += ipt_pkttype.o
 obj-$(CONFIG_IP_NF_MATCH_MULTIPORT) += ipt_multiport.o
 obj-$(CONFIG_IP_NF_MATCH_OWNER) += ipt_owner.o
 obj-$(CONFIG_IP_NF_MATCH_TOS) += ipt_tos.o
-
 obj-$(CONFIG_IP_NF_MATCH_RECENT) += ipt_recent.o
-
 obj-$(CONFIG_IP_NF_MATCH_ECN) += ipt_ecn.o
 obj-$(CONFIG_IP_NF_MATCH_DSCP) += ipt_dscp.o
 obj-$(CONFIG_IP_NF_MATCH_AH_ESP) += ipt_ah.o ipt_esp.o
-
 obj-$(CONFIG_IP_NF_MATCH_LENGTH) += ipt_length.o
-
 obj-$(CONFIG_IP_NF_MATCH_TTL) += ipt_ttl.o
 obj-$(CONFIG_IP_NF_MATCH_STATE) += ipt_state.o
 obj-$(CONFIG_IP_NF_MATCH_CONNTRACK) += ipt_conntrack.o
 obj-$(CONFIG_IP_NF_MATCH_TCPMSS) += ipt_tcpmss.o
 obj-$(CONFIG_IP_NF_MATCH_REALM) += ipt_realm.o
 obj-$(CONFIG_IP_NF_MATCH_ADDRTYPE) += ipt_addrtype.o
-
 obj-$(CONFIG_IP_NF_MATCH_PHYSDEV) += ipt_physdev.o
+obj-$(CONFIG_IP_NF_MATCH_COMMENT) += ipt_comment.o
 
 # targets
 obj-$(CONFIG_IP_NF_TARGET_REJECT) += ipt_REJECT.o

net/ipv4/netfilter/ip_conntrack_proto_icmp.c

@@ -195,7 +195,8 @@ icmp_error_message(struct sk_buff *skb,
 	}
 
 	/* Update skb to refer to this connection */
-	skb->nfct = &h->ctrack->infos[*ctinfo];
+	skb->nfct = &h->ctrack->ct_general;
+	skb->nfctinfo = *ctinfo;
 	return -NF_ACCEPT;
 }
 

net/ipv4/netfilter/ip_conntrack_proto_sctp.c

@@ -430,16 +430,16 @@ static int sctp_new(struct ip_conntrack *conntrack,
 	DEBUGP("\n");
 
 	if (skb_copy_bits(skb, skb->nh.iph->ihl * 4, &sctph, sizeof(sctph)) != 0)
-		return -1;
+		return 0;
 
 	if (do_basic_checks(conntrack, skb, map) != 0)
-		return -1;
+		return 0;
 
 	/* If an OOTB packet has any of these chunks discard (Sec 8.4) */
 	if ((test_bit (SCTP_CID_ABORT, (void *)map))
 		|| (test_bit (SCTP_CID_SHUTDOWN_COMPLETE, (void *)map))
 		|| (test_bit (SCTP_CID_COOKIE_ACK, (void *)map))) {
-		return -1;
+		return 0;
 	}
 
 	newconntrack = SCTP_CONNTRACK_MAX;
@@ -461,7 +461,7 @@ static int sctp_new(struct ip_conntrack *conntrack,
 
 				if (skb_copy_bits(skb, offset + sizeof (sctp_chunkhdr_t), 
 					&inithdr, sizeof(inithdr)) != 0) {
-						return -1;
+						return 0;
 				}
 
 				DEBUGP("Setting vtag %x for new conn\n", 
@@ -471,7 +471,7 @@ static int sctp_new(struct ip_conntrack *conntrack,
 								inithdr.init_tag;
 			} else {
 				/* Sec 8.5.1 (A) */
-				return -1;
+				return 0;
 			}
 		}
 		/* If it is a shutdown ack OOTB packet, we expect a return
@@ -496,7 +496,6 @@ static int sctp_exp_matches_pkt(struct ip_conntrack_expect *exp,
 }
 
 struct ip_conntrack_protocol ip_conntrack_protocol_sctp = { 
-	.list 		 = { NULL, NULL }, 
 	.proto 		 = IPPROTO_SCTP, 
 	.name 		 = "sctp",
 	.pkt_to_tuple 	 = sctp_pkt_to_tuple, 

net/ipv4/netfilter/ip_conntrack_standalone.c

@@ -112,7 +112,7 @@ static int ct_seq_real_show(const struct ip_conntrack_tuple_hash *hash,
 	if (DIRECTION(hash))
 		return 0;
 
-	proto = __ip_ct_find_proto(conntrack->tuplehash[IP_CT_DIR_ORIGINAL]
+	proto = ip_ct_find_proto(conntrack->tuplehash[IP_CT_DIR_ORIGINAL]
 			       .tuple.dst.protonum);
 	IP_NF_ASSERT(proto);
 
@@ -242,7 +242,7 @@ static int exp_seq_show(struct seq_file *s, void *v)
 		   expect->tuple.dst.protonum);
 
 	print_tuple(s, &expect->tuple,
-		    __ip_ct_find_proto(expect->tuple.dst.protonum));
+		    ip_ct_find_proto(expect->tuple.dst.protonum));
 	return seq_putc(s, '\n');
 }
 
@@ -317,7 +317,7 @@ static int ct_cpu_seq_show(struct seq_file *seq, void *v)
 		   st->insert_failed,
 		   st->drop,
 		   st->early_drop,
-		   st->icmp_error,
+		   st->error,
 
 		   st->expect_new,
 		   st->expect_create,
@@ -515,6 +515,14 @@ static ctl_table ip_ct_sysctl_table[] = {
 		.mode		= 0644,
 		.proc_handler	= &proc_dointvec,
 	},
+	{
+		.ctl_name	= NET_IPV4_NF_CONNTRACK_COUNT,
+		.procname	= "ip_conntrack_count",
+		.data		= &ip_conntrack_count,
+		.maxlen		= sizeof(int),
+		.mode		= 0444,
+		.proc_handler	= &proc_dointvec,
+	},
 	{
 		.ctl_name	= NET_IPV4_NF_CONNTRACK_BUCKETS,
 		.procname	= "ip_conntrack_buckets",
@@ -816,19 +824,13 @@ static int init_or_cleanup(int init)
 int ip_conntrack_protocol_register(struct ip_conntrack_protocol *proto)
 {
 	int ret = 0;
-	struct list_head *i;
 
 	WRITE_LOCK(&ip_conntrack_lock);
-	list_for_each(i, &protocol_list) {
-		if (((struct ip_conntrack_protocol *)i)->proto
-		    == proto->proto) {
+	if (ip_ct_protos[proto->proto] != &ip_conntrack_generic_protocol) {
 		ret = -EBUSY;
 		goto out;
 	}
-	}
-
-	list_prepend(&protocol_list, proto);
-
+	ip_ct_protos[proto->proto] = proto;
  out:
 	WRITE_UNLOCK(&ip_conntrack_lock);
 	return ret;
@@ -837,10 +839,7 @@ int ip_conntrack_protocol_register(struct ip_conntrack_protocol *proto)
 void ip_conntrack_protocol_unregister(struct ip_conntrack_protocol *proto)
 {
 	WRITE_LOCK(&ip_conntrack_lock);
-
-	/* ip_ct_find_proto() returns proto_generic in case there is no protocol 
-	 * helper. So this should be enough - HW */
-	LIST_DELETE(&protocol_list, proto);
+	ip_ct_protos[proto->proto] = &ip_conntrack_generic_protocol;
 	WRITE_UNLOCK(&ip_conntrack_lock);
 	
 	/* Somebody could be still looking at the proto in bh. */
@@ -880,8 +879,8 @@ EXPORT_SYMBOL(ip_conntrack_helper_register);
 EXPORT_SYMBOL(ip_conntrack_helper_unregister);
 EXPORT_SYMBOL(ip_ct_selective_cleanup);
 EXPORT_SYMBOL(ip_ct_refresh_acct);
+EXPORT_SYMBOL(ip_ct_protos);
 EXPORT_SYMBOL(ip_ct_find_proto);
-EXPORT_SYMBOL(__ip_ct_find_proto);
 EXPORT_SYMBOL(ip_ct_find_helper);
 EXPORT_SYMBOL(ip_conntrack_expect_alloc);
 EXPORT_SYMBOL(ip_conntrack_expect_related);

net/ipv4/netfilter/ip_fw_compat_masq.c

@@ -146,7 +146,7 @@ check_for_demasq(struct sk_buff **pskb)
 	case IPPROTO_ICMP:
 		/* ICMP errors. */
 		protocol->error(*pskb, &ctinfo, NF_IP_PRE_ROUTING);
-		ct = (struct ip_conntrack *)(*pskb)->nfct->master;
+		ct = (struct ip_conntrack *)(*pskb)->nfct;
 		if (ct) {
 			/* We only do SNAT in the compatibility layer.
 			   So we can manipulate ICMP errors from
@@ -187,7 +187,7 @@ check_for_demasq(struct sk_buff **pskb)
 				      NULL, NULL, NULL);
 
 		/* Put back the reference gained from find_get */
-		nf_conntrack_put(&h->ctrack->infos[0]);
+		nf_conntrack_put(&h->ctrack->ct_general);
 		if (ret == NF_ACCEPT) {
 			struct ip_conntrack *ct;
 			ct = ip_conntrack_get(*pskb, &ctinfo);
@@ -206,7 +206,7 @@ check_for_demasq(struct sk_buff **pskb)
 	} else {
 		if (h)
 			/* Put back the reference gained from find_get */
-			nf_conntrack_put(&h->ctrack->infos[0]);
+			nf_conntrack_put(&h->ctrack->ct_general);
 		ret = NF_ACCEPT;
 	}
 

net/ipv4/netfilter/ip_nat_core.c

@@ -48,9 +48,8 @@ static unsigned int ip_nat_htable_size;
 
 static struct list_head *bysource;
 static struct list_head *byipsproto;
-LIST_HEAD(protos);
+struct ip_nat_protocol *ip_nat_protos[MAX_IP_NAT_PROTO];
 
-extern struct ip_nat_protocol unknown_nat_protocol;
 
 /* We keep extra hashes for each conntrack, for fast searching. */
 static inline size_t
@@ -77,9 +76,6 @@ static void ip_nat_cleanup_conntrack(struct ip_conntrack *conn)
 	if (!info->initialized)
 		return;
 
-	IP_NF_ASSERT(info->bysource.conntrack);
-	IP_NF_ASSERT(info->byipsproto.conntrack);
-
 	hs = hash_by_src(&conn->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src,
 	                 conn->tuplehash[IP_CT_DIR_ORIGINAL]
 	                 .tuple.dst.protonum);
@@ -90,8 +86,8 @@ static void ip_nat_cleanup_conntrack(struct ip_conntrack *conn)
 	                      .tuple.dst.protonum);
 
 	WRITE_LOCK(&ip_nat_lock);
-	LIST_DELETE(&bysource[hs], &info->bysource);
-	LIST_DELETE(&byipsproto[hp], &info->byipsproto);
+	list_del(&info->bysource);
+	list_del(&info->byipsproto);
 	WRITE_UNLOCK(&ip_nat_lock);
 }
 
@@ -106,23 +102,6 @@ ip_nat_cheat_check(u_int32_t oldvalinv, u_int32_t newval, u_int16_t oldcheck)
 				      oldcheck^0xFFFF));
 }
 
-static inline int cmp_proto(const struct ip_nat_protocol *i, int proto)
-{
-	return i->protonum == proto;
-}
-
-struct ip_nat_protocol *
-find_nat_proto(u_int16_t protonum)
-{
-	struct ip_nat_protocol *i;
-
-	MUST_BE_READ_LOCKED(&ip_nat_lock);
-	i = LIST_FIND(&protos, cmp_proto, struct ip_nat_protocol *, protonum);
-	if (!i)
-		i = &unknown_nat_protocol;
-	return i;
-}
-
 /* Is this tuple already taken? (not by us) */
 int
 ip_nat_used_tuple(const struct ip_conntrack_tuple *tuple,
@@ -145,7 +124,7 @@ in_range(const struct ip_conntrack_tuple *tuple,
 	 const struct ip_conntrack_manip *manip,
 	 const struct ip_nat_multi_range *mr)
 {
-	struct ip_nat_protocol *proto = find_nat_proto(tuple->dst.protonum);
+	struct ip_nat_protocol *proto = ip_nat_find_proto(tuple->dst.protonum);
 	unsigned int i;
 	struct ip_conntrack_tuple newtuple = { *manip, tuple->dst };
 
@@ -171,20 +150,18 @@ in_range(const struct ip_conntrack_tuple *tuple,
 }
 
 static inline int
-src_cmp(const struct ip_nat_hash *i,
+src_cmp(const struct ip_conntrack *ct,
 	const struct ip_conntrack_tuple *tuple,
 	const struct ip_nat_multi_range *mr)
 {
-	return (i->conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum
+	return (ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum
 		== tuple->dst.protonum
-		&& i->conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip
+		&& ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip
 		== tuple->src.ip
-		&& i->conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.u.all
+		&& ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.u.all
 		== tuple->src.u.all
 		&& in_range(tuple,
-			    &i->conntrack->tuplehash[IP_CT_DIR_ORIGINAL]
-			    .tuple.src,
-			    mr));
+			    &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src, mr));
 }
 
 /* Only called for SRC manip */
@@ -193,13 +170,12 @@ find_appropriate_src(const struct ip_conntrack_tuple *tuple,
 		     const struct ip_nat_multi_range *mr)
 {
 	unsigned int h = hash_by_src(&tuple->src, tuple->dst.protonum);
-	struct ip_nat_hash *i;
+	struct ip_conntrack *ct;
 
 	MUST_BE_READ_LOCKED(&ip_nat_lock);
-	i = LIST_FIND(&bysource[h], src_cmp, struct ip_nat_hash *, tuple, mr);
-	if (i)
-		return &i->conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src;
-	else
+	list_for_each_entry(ct, &bysource[h], nat.info.bysource)
+		if (src_cmp(ct, tuple, mr))
+			return &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src;
 	return NULL;
 }
 
@@ -226,19 +202,17 @@ do_extra_mangle(u_int32_t var_ip, u_int32_t *other_ipp)
 #endif
 
 /* Simple way to iterate through all. */
-static inline int fake_cmp(const struct ip_nat_hash *i,
+static inline int fake_cmp(const struct ip_conntrack *ct,
 			   u_int32_t src, u_int32_t dst, u_int16_t protonum,
-			   unsigned int *score,
-			   const struct ip_conntrack *conntrack)
+			   unsigned int *score, const struct ip_conntrack *ct2)
 {
 	/* Compare backwards: we're dealing with OUTGOING tuples, and
            inside the conntrack is the REPLY tuple.  Don't count this
            conntrack. */
-	if (i->conntrack != conntrack
-	    && i->conntrack->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip == dst
-	    && i->conntrack->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip == src
-	    && (i->conntrack->tuplehash[IP_CT_DIR_REPLY].tuple.dst.protonum
-		== protonum))
+	if (ct != ct2
+	    && ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip == dst
+	    && ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip == src
+	    && (ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.protonum == protonum))
 		(*score)++;
 	return 0;
 }
@@ -247,13 +221,14 @@ static inline unsigned int
 count_maps(u_int32_t src, u_int32_t dst, u_int16_t protonum,
 	   const struct ip_conntrack *conntrack)
 {
+	struct ip_conntrack *ct;
 	unsigned int score = 0;
 	unsigned int h;
 
 	MUST_BE_READ_LOCKED(&ip_nat_lock);
 	h = hash_by_ipsproto(src, dst, protonum);
-	LIST_FIND(&byipsproto[h], fake_cmp, struct ip_nat_hash *,
-	          src, dst, protonum, &score, conntrack);
+	list_for_each_entry(ct, &byipsproto[h], nat.info.byipsproto)
+		fake_cmp(ct, src, dst, protonum, &score, conntrack);
 
 	return score;
 }
@@ -401,7 +376,7 @@ get_unique_tuple(struct ip_conntrack_tuple *tuple,
 		 unsigned int hooknum)
 {
 	struct ip_nat_protocol *proto
-		= find_nat_proto(orig_tuple->dst.protonum);
+		= ip_nat_find_proto(orig_tuple->dst.protonum);
 	struct ip_nat_range *rptr;
 	unsigned int i;
 	int ret;
@@ -640,12 +615,10 @@ ip_nat_setup_info(struct ip_conntrack *conntrack,
 	/* It's done. */
 	info->initialized |= (1 << HOOK2MANIP(hooknum));
 
-	if (in_hashes) {
-		IP_NF_ASSERT(info->bysource.conntrack);
+	if (in_hashes)
 		replace_in_hashes(conntrack, info);
-	} else {
+	else
 		place_in_hashes(conntrack, info);
-	}
 
 	return NF_ACCEPT;
 }
@@ -669,14 +642,9 @@ void replace_in_hashes(struct ip_conntrack *conntrack,
 				   conntrack->tuplehash[IP_CT_DIR_REPLY]
 				   .tuple.dst.protonum);
 
-	IP_NF_ASSERT(info->bysource.conntrack == conntrack);
 	MUST_BE_WRITE_LOCKED(&ip_nat_lock);
-
-	list_del(&info->bysource.list);
-	list_del(&info->byipsproto.list);
-
-	list_prepend(&bysource[srchash], &info->bysource);
-	list_prepend(&byipsproto[ipsprotohash], &info->byipsproto);
+	list_move(&info->bysource, &bysource[srchash]);
+	list_move(&info->byipsproto, &byipsproto[ipsprotohash]);
 }
 
 void place_in_hashes(struct ip_conntrack *conntrack,
@@ -697,14 +665,9 @@ void place_in_hashes(struct ip_conntrack *conntrack,
 				   conntrack->tuplehash[IP_CT_DIR_REPLY]
 				   .tuple.dst.protonum);
 
-	IP_NF_ASSERT(!info->bysource.conntrack);
-
 	MUST_BE_WRITE_LOCKED(&ip_nat_lock);
-	info->byipsproto.conntrack = conntrack;
-	info->bysource.conntrack = conntrack;
-
-	list_prepend(&bysource[srchash], &info->bysource);
-	list_prepend(&byipsproto[ipsprotohash], &info->byipsproto);
+	list_add(&info->bysource, &bysource[srchash]);
+	list_add(&info->byipsproto, &byipsproto[ipsprotohash]);
 }
 
 /* Returns true if succeeded. */
@@ -724,8 +687,7 @@ manip_pkt(u_int16_t proto,
 	iph = (void *)(*pskb)->data + iphdroff;
 
 	/* Manipulate protcol part. */
-	if (!find_nat_proto(proto)->manip_pkt(pskb,
-					      iphdroff + iph->ihl*4,
+	if (!ip_nat_find_proto(proto)->manip_pkt(pskb, iphdroff + iph->ihl*4,
 	                                         manip, maniptype))
 		return 0;
 
@@ -750,7 +712,7 @@ static inline int exp_for_packet(struct ip_conntrack_expect *exp,
 	int ret = 1;
 
 	MUST_BE_READ_LOCKED(&ip_conntrack_lock);
-	proto = __ip_ct_find_proto(skb->nh.iph->protocol);
+	proto = ip_ct_find_proto(skb->nh.iph->protocol);
 	if (proto->exp_matches_pkt)
 		ret = proto->exp_matches_pkt(exp, skb);
 
@@ -890,12 +852,8 @@ icmp_reply_translation(struct sk_buff **pskb,
 	}
 
 	/* Must be RELATED */
-	IP_NF_ASSERT((*pskb)->nfct
-		     - ((struct ip_conntrack *)(*pskb)->nfct->master)->infos
-		     == IP_CT_RELATED
-		     || (*pskb)->nfct
-		     - ((struct ip_conntrack *)(*pskb)->nfct->master)->infos
-		     == IP_CT_RELATED+IP_CT_IS_REPLY);
+	IP_NF_ASSERT((*pskb)->nfctinfo == IP_CT_RELATED ||
+		     (*pskb)->nfctinfo == IP_CT_RELATED+IP_CT_IS_REPLY);
 
 	/* Redirects on non-null nats must be dropped, else they'll
            start talking to each other without our translation, and be
@@ -995,9 +953,11 @@ int __init ip_nat_init(void)
 
 	/* Sew in builtin protocols. */
 	WRITE_LOCK(&ip_nat_lock);
-	list_append(&protos, &ip_nat_protocol_tcp);
-	list_append(&protos, &ip_nat_protocol_udp);
-	list_append(&protos, &ip_nat_protocol_icmp);
+	for (i = 0; i < MAX_IP_NAT_PROTO; i++)
+		ip_nat_protos[i] = &ip_nat_unknown_protocol;
+	ip_nat_protos[IPPROTO_TCP] = &ip_nat_protocol_tcp;
+	ip_nat_protos[IPPROTO_UDP] = &ip_nat_protocol_udp;
+	ip_nat_protos[IPPROTO_ICMP] = &ip_nat_protocol_icmp;
 	WRITE_UNLOCK(&ip_nat_lock);
 
 	for (i = 0; i < ip_nat_htable_size; i++) {

net/ipv4/netfilter/ip_nat_proto_icmp.c

@@ -104,7 +104,7 @@ icmp_print_range(char *buffer, const struct ip_nat_range *range)
 }
 
 struct ip_nat_protocol ip_nat_protocol_icmp
-= { { NULL, NULL }, "ICMP", IPPROTO_ICMP,
+= { "ICMP", IPPROTO_ICMP,
     icmp_manip_pkt,
     icmp_in_range,
     icmp_unique_tuple,

net/ipv4/netfilter/ip_nat_proto_tcp.c

@@ -162,7 +162,7 @@ tcp_print_range(char *buffer, const struct ip_nat_range *range)
 }
 
 struct ip_nat_protocol ip_nat_protocol_tcp
-= { { NULL, NULL }, "TCP", IPPROTO_TCP,
+= { "TCP", IPPROTO_TCP,
     tcp_manip_pkt,
     tcp_in_range,
     tcp_unique_tuple,

net/ipv4/netfilter/ip_nat_proto_udp.c

@@ -148,7 +148,7 @@ udp_print_range(char *buffer, const struct ip_nat_range *range)
 }
 
 struct ip_nat_protocol ip_nat_protocol_udp
-= { { NULL, NULL }, "UDP", IPPROTO_UDP,
+= { "UDP", IPPROTO_UDP,
     udp_manip_pkt,
     udp_in_range,
     udp_unique_tuple,

net/ipv4/netfilter/ip_nat_proto_unknown.c

@@ -60,8 +60,8 @@ unknown_print_range(char *buffer, const struct ip_nat_range *range)
 	return 0;
 }
 
-struct ip_nat_protocol unknown_nat_protocol = {
-	{ NULL, NULL }, "unknown", 0,
+struct ip_nat_protocol ip_nat_unknown_protocol = {
+	"unknown", 0,
 	unknown_manip_pkt,
 	unknown_in_range,
 	unknown_unique_tuple,

net/ipv4/netfilter/ip_nat_standalone.c

@@ -283,18 +283,13 @@ static struct nf_hook_ops ip_nat_local_in_ops = {
 int ip_nat_protocol_register(struct ip_nat_protocol *proto)
 {
 	int ret = 0;
-	struct list_head *i;
 
 	WRITE_LOCK(&ip_nat_lock);
-	list_for_each(i, &protos) {
-		if (((struct ip_nat_protocol *)i)->protonum
-		    == proto->protonum) {
+	if (ip_nat_protos[proto->protonum] != &ip_nat_unknown_protocol) {
 		ret = -EBUSY;
 		goto out;
 	}
-	}
-
-	list_prepend(&protos, proto);
+	ip_nat_protos[proto->protonum] = proto;
  out:
 	WRITE_UNLOCK(&ip_nat_lock);
 	return ret;
@@ -304,7 +299,7 @@ int ip_nat_protocol_register(struct ip_nat_protocol *proto)
 void ip_nat_protocol_unregister(struct ip_nat_protocol *proto)
 {
 	WRITE_LOCK(&ip_nat_lock);
-	LIST_DELETE(&protos, proto);
+	ip_nat_protos[proto->protonum] = &ip_nat_unknown_protocol;
 	WRITE_UNLOCK(&ip_nat_lock);
 
 	/* Someone could be still looking at the proto in a bh. */

net/ipv4/netfilter/ipt_NOTRACK.c

@@ -23,7 +23,8 @@ target(struct sk_buff **pskb,
 	   If there is a real ct entry correspondig to this packet, 
 	   it'll hang aroun till timing out. We don't deal with it
 	   for performance reasons. JK */
-	(*pskb)->nfct = &ip_conntrack_untracked.infos[IP_CT_NEW];
+	(*pskb)->nfct = &ip_conntrack_untracked.ct_general;
+	(*pskb)->nfctinfo = IP_CT_NEW;
 	nf_conntrack_get((*pskb)->nfct);
 
 	return IPT_CONTINUE;

net/ipv4/netfilter/ipt_REJECT.c

@@ -41,14 +41,14 @@ MODULE_DESCRIPTION("iptables REJECT target module");
 /* If the original packet is part of a connection, but the connection
    is not confirmed, our manufactured reply will not be associated
    with it, so we need to do this manually. */
-static void connection_attach(struct sk_buff *new_skb, struct nf_ct_info *nfct)
+static void connection_attach(struct sk_buff *new_skb, struct sk_buff *skb)
 {
-	void (*attach)(struct sk_buff *, struct nf_ct_info *);
+	void (*attach)(struct sk_buff *, struct sk_buff *);
 
 	/* Avoid module unload race with ip_ct_attach being NULLed out */
-	if (nfct && (attach = ip_ct_attach) != NULL) {
+	if (skb->nfct && (attach = ip_ct_attach) != NULL) {
 		mb(); /* Just to be sure: must be read before executing this */
-		attach(new_skb, nfct);
+		attach(new_skb, skb);
 	}
 }
 
@@ -209,7 +209,7 @@ static void send_reset(struct sk_buff *oldskb, int hook)
 	if (nskb->len > dst_pmtu(nskb->dst))
 		goto free_nskb;
 
-	connection_attach(nskb, oldskb->nfct);
+	connection_attach(nskb, oldskb);
 
 	NF_HOOK(PF_INET, NF_IP_LOCAL_OUT, nskb, NULL, nskb->dst->dev,
 		ip_finish_output);
@@ -360,7 +360,7 @@ static void send_unreach(struct sk_buff *skb_in, int code)
 	icmph->checksum = ip_compute_csum((unsigned char *)icmph,
 					  length - sizeof(struct iphdr));
 
-	connection_attach(nskb, skb_in->nfct);
+	connection_attach(nskb, skb_in);
 
 	NF_HOOK(PF_INET, NF_IP_LOCAL_OUT, nskb, NULL, nskb->dst->dev,
 		ip_finish_output);

net/ipv4/netfilter/ipt_comment.c

+/*
+ * Implements a dummy match to allow attaching comments to rules
+ *
+ * 2003-05-13 Brad Fisher (brad@info-link.net)
+ */
+
+#include <linux/module.h>
+#include <linux/skbuff.h>
+#include <linux/netfilter_ipv4/ip_tables.h>
+#include <linux/netfilter_ipv4/ipt_comment.h>
+
+MODULE_AUTHOR("Brad Fisher <brad@info-link.net>");
+MODULE_DESCRIPTION("iptables comment match module");
+MODULE_LICENSE("GPL");
+
+static int
+match(const struct sk_buff *skb,
+      const struct net_device *in,
+      const struct net_device *out,
+      const void *matchinfo,
+      int offset,
+      int *hotdrop)
+{
+	/* We always match */
+	return 1;
+}
+
+static int
+checkentry(const char *tablename,
+           const struct ipt_ip *ip,
+           void *matchinfo,
+           unsigned int matchsize,
+           unsigned int hook_mask)
+{
+	/* Check the size */
+	if (matchsize != IPT_ALIGN(sizeof(struct ipt_comment_info)))
+		return 0;
+	return 1;
+}
+
+static struct ipt_match comment_match = {
+	.name		= "comment",
+	.match		= match,
+	.checkentry	= checkentry,
+	.me		= THIS_MODULE
+};
+
+static int __init init(void)
+{
+	return ipt_register_match(&comment_match);
+}
+
+static void __exit fini(void)
+{
+	ipt_unregister_match(&comment_match);
+}
+
+module_init(init);
+module_exit(fini);

net/ipv4/netfilter/ipt_conntrack.c

@@ -35,7 +35,7 @@ match(const struct sk_buff *skb,
 
 #define FWINV(bool,invflg) ((bool) ^ !!(sinfo->invflags & invflg))
 
-	if (skb->nfct == &ip_conntrack_untracked.infos[IP_CT_NEW])
+	if (ct == &ip_conntrack_untracked)
 		statebit = IPT_CONNTRACK_STATE_UNTRACKED;
 	else if (ct)
  		statebit = IPT_CONNTRACK_STATE_BIT(ctinfo);

net/ipv4/netfilter/ipt_state.c

@@ -30,9 +30,9 @@ match(const struct sk_buff *skb,
 	enum ip_conntrack_info ctinfo;
 	unsigned int statebit;
 
-	if (skb->nfct == &ip_conntrack_untracked.infos[IP_CT_NEW])
+	if (skb->nfct == &ip_conntrack_untracked.ct_general)
 		statebit = IPT_STATE_UNTRACKED;
-	else if (!ip_conntrack_get((struct sk_buff *)skb, &ctinfo))
+	else if (!ip_conntrack_get(skb, &ctinfo))
 		statebit = IPT_STATE_INVALID;
 	else
 		statebit = IPT_STATE_BIT(ctinfo);

net/ipv6/ip6_output.c

@@ -477,6 +477,7 @@ static void ip6_copy_metadata(struct sk_buff *to, struct sk_buff *from)
 	/* Connection association is same as pre-frag packet */
 	to->nfct = from->nfct;
 	nf_conntrack_get(to->nfct);
+	to->nfctinfo = from->nfctinfo;
 #ifdef CONFIG_BRIDGE_NETFILTER
 	nf_bridge_put(to->nf_bridge);
 	to->nf_bridge = from->nf_bridge;