Merge master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6: (25 commits)
  [XFRM]: Fix OOPSes in xfrm_audit_log().
  [TCP]: cleanup of htcp (resend)
  [TCP]: Use read mostly for CUBIC parameters.
  [NETFILTER]: nf_conntrack_tcp: make sysctl variables static
  [NETFILTER]: ip6t_mh: drop piggyback payload packet on MH packets
  [NETFILTER]: Fix whitespace errors
  [NETFILTER]: Kconfig: improve dependency handling
  [NETFILTER]: xt_mac/xt_CLASSIFY: use IPv6 hook names for IPv6 registration
  [NETFILTER]: nf_conntrack: change nf_conntrack_l[34]proto_unregister to void
  [NETFILTER]: nf_conntrack: properly use RCU for nf_conntrack_destroyed callback
  [NETFILTER]: ip_conntrack: properly use RCU for ip_conntrack_destroyed callback
  [NETFILTER]: nf_conntrack: fix invalid conntrack statistics RCU assumption
  [NETFILTER]: ip_conntrack: fix invalid conntrack statistics RCU assumption
  [NETFILTER]: nf_conntrack: properly use RCU API for nf_ct_protos/nf_ct_l3protos arrays
  [NETFILTER]: ip_conntrack: properly use RCU API for ip_ct_protos array
  [NETFILTER]: nf_nat: properly use RCU API for nf_nat_protos array
  [NETFILTER]: ip_nat: properly use RCU API for ip_nat_protos array
  [NETFILTER]: nf_log: minor cleanups
  [NETFILTER]: nf_log: switch logger registration/unregistration to mutex
  [NETFILTER]: nf_log: make nf_log_unregister_pf return void
  ...

include/linux/netfilter.h

@@ -172,8 +172,8 @@ struct nf_logger {
 
 /* Function to register/unregister log function. */
 int nf_log_register(int pf, struct nf_logger *logger);
-int nf_log_unregister_pf(int pf);
-void nf_log_unregister_logger(struct nf_logger *logger);
+void nf_log_unregister(struct nf_logger *logger);
+void nf_log_unregister_pf(int pf);
 
 /* Calls the registered backend logging function */
 void nf_log_packet(int pf,

include/linux/netfilter_ipv4/ip_conntrack.h

@@ -301,6 +301,12 @@ extern unsigned int ip_conntrack_htable_size;
 extern int ip_conntrack_checksum;
  
 #define CONNTRACK_STAT_INC(count) (__get_cpu_var(ip_conntrack_stat).count++)
+#define CONNTRACK_STAT_INC_ATOMIC(count)		\
+do {							\
+	local_bh_disable();				\
+	__get_cpu_var(ip_conntrack_stat).count++;	\
+	local_bh_enable();				\
+} while (0)
 
 #ifdef CONFIG_IP_NF_CONNTRACK_EVENTS
 #include <linux/notifier.h>

include/net/netfilter/nf_conntrack.h

@@ -257,6 +257,12 @@ 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)			\
+do {							\
+	local_bh_disable();				\
+	__get_cpu_var(nf_conntrack_stat).count++;	\
+	local_bh_enable();				\
+} while (0)
 
 /* no helper, no nat */
 #define	NF_CT_F_BASIC	0

include/net/netfilter/nf_conntrack_l3proto.h

@@ -89,7 +89,7 @@ extern struct nf_conntrack_l3proto *nf_ct_l3protos[AF_MAX];
 
 /* Protocol registration. */
 extern int nf_conntrack_l3proto_register(struct nf_conntrack_l3proto *proto);
-extern int nf_conntrack_l3proto_unregister(struct nf_conntrack_l3proto *proto);
+extern void nf_conntrack_l3proto_unregister(struct nf_conntrack_l3proto *proto);
 
 extern struct nf_conntrack_l3proto *
 nf_ct_l3proto_find_get(u_int16_t l3proto);
@@ -106,7 +106,7 @@ __nf_ct_l3proto_find(u_int16_t l3proto)
 {
 	if (unlikely(l3proto >= AF_MAX))
 		return &nf_conntrack_l3proto_generic;
-	return nf_ct_l3protos[l3proto];
+	return rcu_dereference(nf_ct_l3protos[l3proto]);
 }
 
 #endif /*_NF_CONNTRACK_L3PROTO_H*/

include/net/netfilter/nf_conntrack_l4proto.h

@@ -109,7 +109,7 @@ extern void nf_ct_l4proto_put(struct nf_conntrack_l4proto *p);
 
 /* Protocol registration. */
 extern int nf_conntrack_l4proto_register(struct nf_conntrack_l4proto *proto);
-extern int nf_conntrack_l4proto_unregister(struct nf_conntrack_l4proto *proto);
+extern void nf_conntrack_l4proto_unregister(struct nf_conntrack_l4proto *proto);
 
 /* Generic netlink helpers */
 extern int nf_ct_port_tuple_to_nfattr(struct sk_buff *skb,

net/bridge/netfilter/ebt_log.c

@@ -208,7 +208,7 @@ static int __init ebt_log_init(void)
 
 static void __exit ebt_log_fini(void)
 {
-	nf_log_unregister_logger(&ebt_log_logger);
+	nf_log_unregister(&ebt_log_logger);
 	ebt_unregister_watcher(&log);
 }
 

net/bridge/netfilter/ebt_ulog.c

@@ -323,7 +323,7 @@ static void __exit ebt_ulog_fini(void)
 	ebt_ulog_buff_t *ub;
 	int i;
 
-	nf_log_unregister_logger(&ebt_ulog_logger);
+	nf_log_unregister(&ebt_ulog_logger);
 	ebt_unregister_watcher(&ulog);
 	for (i = 0; i < EBT_ULOG_MAXNLGROUPS; i++) {
 		ub = &ulog_buffers[i];

net/ipv4/netfilter/Kconfig

@@ -226,7 +226,7 @@ config IP_NF_QUEUE
 
 config IP_NF_IPTABLES
 	tristate "IP tables support (required for filtering/masq/NAT)"
-	depends on NETFILTER_XTABLES
+	select NETFILTER_XTABLES
 	help
 	  iptables is a general, extensible packet identification framework.
 	  The packet filtering and full NAT (masquerading, port forwarding,
@@ -606,7 +606,9 @@ config IP_NF_TARGET_TTL
 config IP_NF_TARGET_CLUSTERIP
 	tristate "CLUSTERIP target support (EXPERIMENTAL)"
 	depends on IP_NF_MANGLE && EXPERIMENTAL
-	depends on (IP_NF_CONNTRACK && IP_NF_CONNTRACK_MARK) || (NF_CONNTRACK_MARK && NF_CONNTRACK_IPV4)
+	depends on IP_NF_CONNTRACK || NF_CONNTRACK_IPV4
+	select IP_NF_CONNTRACK_MARK if IP_NF_CONNTRACK
+	select NF_CONNTRACK_MARK if NF_CONNTRACK_IPV4
 	help
 	  The CLUSTERIP target allows you to build load-balancing clusters of
 	  network servers without having a dedicated load-balancing
@@ -629,7 +631,7 @@ config IP_NF_RAW
 # ARP tables
 config IP_NF_ARPTABLES
 	tristate "ARP tables support"
-	depends on NETFILTER_XTABLES
+	select NETFILTER_XTABLES
 	help
 	  arptables is a general, extensible packet identification framework.
 	  The ARP packet filtering and mangling (manipulation)subsystems

net/ipv4/netfilter/ip_conntrack_core.c

@@ -303,6 +303,7 @@ destroy_conntrack(struct nf_conntrack *nfct)
 	struct ip_conntrack *ct = (struct ip_conntrack *)nfct;
 	struct ip_conntrack_protocol *proto;
 	struct ip_conntrack_helper *helper;
+	typeof(ip_conntrack_destroyed) destroyed;
 
 	DEBUGP("destroy_conntrack(%p)\n", ct);
 	IP_NF_ASSERT(atomic_read(&nfct->use) == 0);
@@ -318,12 +319,16 @@ destroy_conntrack(struct nf_conntrack *nfct)
 	/* To make sure we don't get any weird locking issues here:
 	 * destroy_conntrack() MUST NOT be called with a write lock
 	 * to ip_conntrack_lock!!! -HW */
+	rcu_read_lock();
 	proto = __ip_conntrack_proto_find(ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.protonum);
 	if (proto && proto->destroy)
 		proto->destroy(ct);
 
-	if (ip_conntrack_destroyed)
-		ip_conntrack_destroyed(ct);
+	destroyed = rcu_dereference(ip_conntrack_destroyed);
+	if (destroyed)
+		destroyed(ct);
+
+	rcu_read_unlock();
 
 	write_lock_bh(&ip_conntrack_lock);
 	/* Expectations will have been removed in clean_from_lists,
@@ -536,7 +541,7 @@ static int early_drop(struct list_head *chain)
 	if (del_timer(&ct->timeout)) {
 		death_by_timeout((unsigned long)ct);
 		dropped = 1;
-		CONNTRACK_STAT_INC(early_drop);
+		CONNTRACK_STAT_INC_ATOMIC(early_drop);
 	}
 	ip_conntrack_put(ct);
 	return dropped;
@@ -595,13 +600,13 @@ ip_conntrack_proto_find_get(u_int8_t protocol)
 {
 	struct ip_conntrack_protocol *p;
 
-	preempt_disable();
+	rcu_read_lock();
 	p = __ip_conntrack_proto_find(protocol);
 	if (p) {
 		if (!try_module_get(p->me))
 			p = &ip_conntrack_generic_protocol;
 	}
-	preempt_enable();
+	rcu_read_unlock();
 
 	return p;
 }
@@ -802,7 +807,7 @@ unsigned int ip_conntrack_in(unsigned int hooknum,
 
 	/* Previously seen (loopback or untracked)?  Ignore. */
 	if ((*pskb)->nfct) {
-		CONNTRACK_STAT_INC(ignore);
+		CONNTRACK_STAT_INC_ATOMIC(ignore);
 		return NF_ACCEPT;
 	}
 
@@ -830,6 +835,7 @@ unsigned int ip_conntrack_in(unsigned int hooknum,
 	}
 #endif
 
+	/* rcu_read_lock()ed by nf_hook_slow */
 	proto = __ip_conntrack_proto_find((*pskb)->nh.iph->protocol);
 
 	/* It may be an special packet, error, unclean...
@@ -837,20 +843,20 @@ unsigned int ip_conntrack_in(unsigned int hooknum,
 	 * core what to do with the packet. */
 	if (proto->error != NULL
 	    && (ret = proto->error(*pskb, &ctinfo, hooknum)) <= 0) {
-		CONNTRACK_STAT_INC(error);
-		CONNTRACK_STAT_INC(invalid);
+		CONNTRACK_STAT_INC_ATOMIC(error);
+		CONNTRACK_STAT_INC_ATOMIC(invalid);
 		return -ret;
 	}
 
 	if (!(ct = resolve_normal_ct(*pskb, proto,&set_reply,hooknum,&ctinfo))) {
 		/* Not valid part of a connection */
-		CONNTRACK_STAT_INC(invalid);
+		CONNTRACK_STAT_INC_ATOMIC(invalid);
 		return NF_ACCEPT;
 	}
 
 	if (IS_ERR(ct)) {
 		/* Too stressed to deal. */
-		CONNTRACK_STAT_INC(drop);
+		CONNTRACK_STAT_INC_ATOMIC(drop);
 		return NF_DROP;
 	}
 
@@ -862,7 +868,7 @@ unsigned int ip_conntrack_in(unsigned int hooknum,
 		 * the netfilter core what to do*/
 		nf_conntrack_put((*pskb)->nfct);
 		(*pskb)->nfct = NULL;
-		CONNTRACK_STAT_INC(invalid);
+		CONNTRACK_STAT_INC_ATOMIC(invalid);
 		return -ret;
 	}
 
@@ -875,8 +881,15 @@ unsigned int ip_conntrack_in(unsigned int hooknum,
 int invert_tuplepr(struct ip_conntrack_tuple *inverse,
 		   const struct ip_conntrack_tuple *orig)
 {
-	return ip_ct_invert_tuple(inverse, orig,
-				  __ip_conntrack_proto_find(orig->dst.protonum));
+	struct ip_conntrack_protocol *proto;
+	int ret;
+
+	rcu_read_lock();
+	proto = __ip_conntrack_proto_find(orig->dst.protonum);
+	ret = ip_ct_invert_tuple(inverse, orig, proto);
+	rcu_read_unlock();
+
+	return ret;
 }
 
 /* Would two expected things clash? */
@@ -1354,7 +1367,7 @@ static void free_conntrack_hash(struct list_head *hash, int vmalloced,int size)
    supposed to kill the mall. */
 void ip_conntrack_cleanup(void)
 {
-	ip_ct_attach = NULL;
+	rcu_assign_pointer(ip_ct_attach, NULL);
 
 	/* This makes sure all current packets have passed through
 	   netfilter framework.  Roll on, two-stage module
@@ -1507,15 +1520,15 @@ int __init ip_conntrack_init(void)
 	/* Don't NEED lock here, but good form anyway. */
 	write_lock_bh(&ip_conntrack_lock);
 	for (i = 0; i < MAX_IP_CT_PROTO; i++)
-		ip_ct_protos[i] = &ip_conntrack_generic_protocol;
+		rcu_assign_pointer(ip_ct_protos[i], &ip_conntrack_generic_protocol);
 	/* Sew in builtin protocols. */
-	ip_ct_protos[IPPROTO_TCP] = &ip_conntrack_protocol_tcp;
-	ip_ct_protos[IPPROTO_UDP] = &ip_conntrack_protocol_udp;
-	ip_ct_protos[IPPROTO_ICMP] = &ip_conntrack_protocol_icmp;
+	rcu_assign_pointer(ip_ct_protos[IPPROTO_TCP], &ip_conntrack_protocol_tcp);
+	rcu_assign_pointer(ip_ct_protos[IPPROTO_UDP], &ip_conntrack_protocol_udp);
+	rcu_assign_pointer(ip_ct_protos[IPPROTO_ICMP], &ip_conntrack_protocol_icmp);
 	write_unlock_bh(&ip_conntrack_lock);
 
 	/* For use by ipt_REJECT */
-	ip_ct_attach = ip_conntrack_attach;
+	rcu_assign_pointer(ip_ct_attach, ip_conntrack_attach);
 
 	/* Set up fake conntrack:
 	    - to never be deleted, not in any hashes */

net/ipv4/netfilter/ip_conntrack_standalone.c

@@ -796,7 +796,7 @@ int ip_conntrack_protocol_register(struct ip_conntrack_protocol *proto)
 		ret = -EBUSY;
 		goto out;
 	}
-	ip_ct_protos[proto->proto] = proto;
+	rcu_assign_pointer(ip_ct_protos[proto->proto], proto);
  out:
 	write_unlock_bh(&ip_conntrack_lock);
 	return ret;
@@ -805,11 +805,10 @@ int ip_conntrack_protocol_register(struct ip_conntrack_protocol *proto)
 void ip_conntrack_protocol_unregister(struct ip_conntrack_protocol *proto)
 {
 	write_lock_bh(&ip_conntrack_lock);
-	ip_ct_protos[proto->proto] = &ip_conntrack_generic_protocol;
+	rcu_assign_pointer(ip_ct_protos[proto->proto],
+			   &ip_conntrack_generic_protocol);
 	write_unlock_bh(&ip_conntrack_lock);
-
-	/* Somebody could be still looking at the proto in bh. */
-	synchronize_net();
+	synchronize_rcu();
 
 	/* Remove all contrack entries for this protocol */
 	ip_ct_iterate_cleanup(kill_proto, &proto->proto);

net/ipv4/netfilter/ip_nat_core.c

@@ -50,7 +50,7 @@ static struct ip_nat_protocol *ip_nat_protos[MAX_IP_NAT_PROTO];
 static inline struct ip_nat_protocol *
 __ip_nat_proto_find(u_int8_t protonum)
 {
-	return ip_nat_protos[protonum];
+	return rcu_dereference(ip_nat_protos[protonum]);
 }
 
 struct ip_nat_protocol *
@@ -58,13 +58,11 @@ ip_nat_proto_find_get(u_int8_t protonum)
 {
 	struct ip_nat_protocol *p;
 
-	/* we need to disable preemption to make sure 'p' doesn't get
-	 * removed until we've grabbed the reference */
-	preempt_disable();
+	rcu_read_lock();
 	p = __ip_nat_proto_find(protonum);
 	if (!try_module_get(p->me))
 		p = &ip_nat_unknown_protocol;
-	preempt_enable();
+	rcu_read_unlock();
 
 	return p;
 }
@@ -120,8 +118,8 @@ static int
 in_range(const struct ip_conntrack_tuple *tuple,
 	 const struct ip_nat_range *range)
 {
-	struct ip_nat_protocol *proto =
-				__ip_nat_proto_find(tuple->dst.protonum);
+	struct ip_nat_protocol *proto;
+	int ret = 0;
 
 	/* If we are supposed to map IPs, then we must be in the
 	   range specified, otherwise let this drag us onto a new src IP. */
@@ -131,12 +129,15 @@ in_range(const struct ip_conntrack_tuple *tuple,
 			return 0;
 	}
 
+	rcu_read_lock();
+	proto = __ip_nat_proto_find(tuple->dst.protonum);
 	if (!(range->flags & IP_NAT_RANGE_PROTO_SPECIFIED)
 	    || proto->in_range(tuple, IP_NAT_MANIP_SRC,
 			       &range->min, &range->max))
-		return 1;
+		ret = 1;
+	rcu_read_unlock();
 
-	return 0;
+	return ret;
 }
 
 static inline int
@@ -260,27 +261,25 @@ get_unique_tuple(struct ip_conntrack_tuple *tuple,
 	/* 3) The per-protocol part of the manip is made to map into
 	   the range to make a unique tuple. */
 
-	proto = ip_nat_proto_find_get(orig_tuple->dst.protonum);
+	rcu_read_lock();
+	proto = __ip_nat_proto_find(orig_tuple->dst.protonum);
 
 	/* Change protocol info to have some randomization */
 	if (range->flags & IP_NAT_RANGE_PROTO_RANDOM) {
 		proto->unique_tuple(tuple, range, maniptype, conntrack);
-		ip_nat_proto_put(proto);
-		return;
+		goto out;
 	}
 
 	/* Only bother mapping if it's not already in range and unique */
 	if ((!(range->flags & IP_NAT_RANGE_PROTO_SPECIFIED)
 	     || proto->in_range(tuple, maniptype, &range->min, &range->max))
-	    && !ip_nat_used_tuple(tuple, conntrack)) {
-		ip_nat_proto_put(proto);
-		return;
-	}
+	    && !ip_nat_used_tuple(tuple, conntrack))
+		goto out;
 
 	/* Last change: get protocol to try to obtain unique tuple. */
 	proto->unique_tuple(tuple, range, maniptype, conntrack);
-
-	ip_nat_proto_put(proto);
+out:
+	rcu_read_unlock();
 }
 
 unsigned int
@@ -360,12 +359,11 @@ manip_pkt(u_int16_t proto,
 	iph = (void *)(*pskb)->data + iphdroff;
 
 	/* Manipulate protcol part. */
-	p = ip_nat_proto_find_get(proto);
-	if (!p->manip_pkt(pskb, iphdroff, target, maniptype)) {
-		ip_nat_proto_put(p);
+
+	/* rcu_read_lock()ed by nf_hook_slow */
+	p = __ip_nat_proto_find(proto);
+	if (!p->manip_pkt(pskb, iphdroff, target, maniptype))
 		return 0;
-	}
-	ip_nat_proto_put(p);
 
 	iph = (void *)(*pskb)->data + iphdroff;
 
@@ -422,6 +420,7 @@ int ip_nat_icmp_reply_translation(struct ip_conntrack *ct,
 		struct icmphdr icmp;
 		struct iphdr ip;
 	} *inside;
+	struct ip_conntrack_protocol *proto;
 	struct ip_conntrack_tuple inner, target;
 	int hdrlen = (*pskb)->nh.iph->ihl * 4;
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
@@ -457,10 +456,11 @@ int ip_nat_icmp_reply_translation(struct ip_conntrack *ct,
 	DEBUGP("icmp_reply_translation: translating error %p manp %u dir %s\n",
 	       *pskb, manip, dir == IP_CT_DIR_ORIGINAL ? "ORIG" : "REPLY");
 
+	/* rcu_read_lock()ed by nf_hook_slow */
+	proto = __ip_conntrack_proto_find(inside->ip.protocol);
 	if (!ip_ct_get_tuple(&inside->ip, *pskb, (*pskb)->nh.iph->ihl*4 +
 			     sizeof(struct icmphdr) + inside->ip.ihl*4,
-			     &inner,
-			     __ip_conntrack_proto_find(inside->ip.protocol)))
+			     &inner, proto))
 		return 0;
 
 	/* Change inner back to look like incoming packet.  We do the
@@ -515,7 +515,7 @@ int ip_nat_protocol_register(struct ip_nat_protocol *proto)
 		ret = -EBUSY;
 		goto out;
 	}
-	ip_nat_protos[proto->protonum] = proto;
+	rcu_assign_pointer(ip_nat_protos[proto->protonum], proto);
  out:
 	write_unlock_bh(&ip_nat_lock);
 	return ret;
@@ -526,11 +526,10 @@ EXPORT_SYMBOL(ip_nat_protocol_register);
 void ip_nat_protocol_unregister(struct ip_nat_protocol *proto)
 {
 	write_lock_bh(&ip_nat_lock);
-	ip_nat_protos[proto->protonum] = &ip_nat_unknown_protocol;
+	rcu_assign_pointer(ip_nat_protos[proto->protonum],
+			   &ip_nat_unknown_protocol);
 	write_unlock_bh(&ip_nat_lock);
-
-	/* Someone could be still looking at the proto in a bh. */
-	synchronize_net();
+	synchronize_rcu();
 }
 EXPORT_SYMBOL(ip_nat_protocol_unregister);
 
@@ -594,10 +593,10 @@ static int __init ip_nat_init(void)
 	/* Sew in builtin protocols. */
 	write_lock_bh(&ip_nat_lock);
 	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;
+		rcu_assign_pointer(ip_nat_protos[i], &ip_nat_unknown_protocol);
+	rcu_assign_pointer(ip_nat_protos[IPPROTO_TCP], &ip_nat_protocol_tcp);
+	rcu_assign_pointer(ip_nat_protos[IPPROTO_UDP], &ip_nat_protocol_udp);
+	rcu_assign_pointer(ip_nat_protos[IPPROTO_ICMP], &ip_nat_protocol_icmp);
 	write_unlock_bh(&ip_nat_lock);
 
 	for (i = 0; i < ip_nat_htable_size; i++) {
@@ -605,8 +604,8 @@ static int __init ip_nat_init(void)
 	}
 
 	/* FIXME: Man, this is a hack.  <SIGH> */
-	IP_NF_ASSERT(ip_conntrack_destroyed == NULL);
-	ip_conntrack_destroyed = &ip_nat_cleanup_conntrack;
+	IP_NF_ASSERT(rcu_dereference(ip_conntrack_destroyed) == NULL);
+	rcu_assign_pointer(ip_conntrack_destroyed, ip_nat_cleanup_conntrack);
 
 	/* Initialize fake conntrack so that NAT will skip it */
 	ip_conntrack_untracked.status |= IPS_NAT_DONE_MASK;
@@ -624,7 +623,8 @@ static int clean_nat(struct ip_conntrack *i, void *data)
 static void __exit ip_nat_cleanup(void)
 {
 	ip_ct_iterate_cleanup(&clean_nat, NULL);
-	ip_conntrack_destroyed = NULL;
+	rcu_assign_pointer(ip_conntrack_destroyed, NULL);
+	synchronize_rcu();
 	vfree(bysource);
 }
 

net/ipv4/netfilter/ipt_LOG.c

@@ -489,7 +489,7 @@ static int __init ipt_log_init(void)
 
 static void __exit ipt_log_fini(void)
 {
-	nf_log_unregister_logger(&ipt_log_logger);
+	nf_log_unregister(&ipt_log_logger);
 	xt_unregister_target(&ipt_log_reg);
 }
 

net/ipv4/netfilter/ipt_ULOG.c

@@ -419,7 +419,7 @@ static void __exit ipt_ulog_fini(void)
 	DEBUGP("ipt_ULOG: cleanup_module\n");
 
 	if (nflog)
-		nf_log_unregister_logger(&ipt_ulog_logger);
+		nf_log_unregister(&ipt_ulog_logger);
 	xt_unregister_target(&ipt_ulog_reg);
 	sock_release(nflognl->sk_socket);
 

net/ipv4/netfilter/nf_conntrack_proto_icmp.c

@@ -170,7 +170,9 @@ icmp_error_message(struct sk_buff *skb,
 		return -NF_ACCEPT;
 	}
 
+	/* rcu_read_lock()ed by nf_hook_slow */
 	innerproto = __nf_ct_l4proto_find(PF_INET, inside->ip.protocol);
+
 	dataoff = skb->nh.iph->ihl*4 + sizeof(inside->icmp);
 	/* Are they talking about one of our connections? */
 	if (!nf_ct_get_tuple(skb, dataoff, dataoff + inside->ip.ihl*4, PF_INET,

net/ipv4/netfilter/nf_nat_core.c

@@ -53,7 +53,7 @@ static struct nf_nat_protocol *nf_nat_protos[MAX_IP_NAT_PROTO];
 static inline struct nf_nat_protocol *
 __nf_nat_proto_find(u_int8_t protonum)
 {
-	return nf_nat_protos[protonum];
+	return rcu_dereference(nf_nat_protos[protonum]);
 }
 
 struct nf_nat_protocol *
@@ -61,13 +61,11 @@ nf_nat_proto_find_get(u_int8_t protonum)
 {
 	struct nf_nat_protocol *p;
 
-	/* we need to disable preemption to make sure 'p' doesn't get
-	 * removed until we've grabbed the reference */
-	preempt_disable();
+	rcu_read_lock();
 	p = __nf_nat_proto_find(protonum);
 	if (!try_module_get(p->me))
 		p = &nf_nat_unknown_protocol;
-	preempt_enable();
+	rcu_read_unlock();
 
 	return p;
 }
@@ -126,8 +124,8 @@ in_range(const struct nf_conntrack_tuple *tuple,
 	 const struct nf_nat_range *range)
 {
 	struct nf_nat_protocol *proto;
+	int ret = 0;
 
-	proto = __nf_nat_proto_find(tuple->dst.protonum);
 	/* If we are supposed to map IPs, then we must be in the
 	   range specified, otherwise let this drag us onto a new src IP. */
 	if (range->flags & IP_NAT_RANGE_MAP_IPS) {
@@ -136,12 +134,15 @@ in_range(const struct nf_conntrack_tuple *tuple,
 			return 0;
 	}
 
+	rcu_read_lock();
+	proto = __nf_nat_proto_find(tuple->dst.protonum);
 	if (!(range->flags & IP_NAT_RANGE_PROTO_SPECIFIED) ||
 	    proto->in_range(tuple, IP_NAT_MANIP_SRC,
 			    &range->min, &range->max))
-		return 1;
+		ret = 1;
+	rcu_read_unlock();
 
-	return 0;
+	return ret;
 }
 
 static inline int
@@ -268,27 +269,25 @@ get_unique_tuple(struct nf_conntrack_tuple *tuple,
 	/* 3) The per-protocol part of the manip is made to map into
 	   the range to make a unique tuple. */
 
-	proto = nf_nat_proto_find_get(orig_tuple->dst.protonum);
+	rcu_read_lock();
+	proto = __nf_nat_proto_find(orig_tuple->dst.protonum);
 
 	/* Change protocol info to have some randomization */
 	if (range->flags & IP_NAT_RANGE_PROTO_RANDOM) {
 		proto->unique_tuple(tuple, range, maniptype, ct);
-		nf_nat_proto_put(proto);
-		return;
+		goto out;
 	}
 
 	/* Only bother mapping if it's not already in range and unique */
 	if ((!(range->flags & IP_NAT_RANGE_PROTO_SPECIFIED) ||
 	     proto->in_range(tuple, maniptype, &range->min, &range->max)) &&
-	    !nf_nat_used_tuple(tuple, ct)) {
-		nf_nat_proto_put(proto);
-		return;
-	}
+	    !nf_nat_used_tuple(tuple, ct))
+		goto out;
 
 	/* Last change: get protocol to try to obtain unique tuple. */
 	proto->unique_tuple(tuple, range, maniptype, ct);
-
-	nf_nat_proto_put(proto);
+out:
+	rcu_read_unlock();
 }
 
 unsigned int
@@ -369,12 +368,11 @@ manip_pkt(u_int16_t proto,
 	iph = (void *)(*pskb)->data + iphdroff;
 
 	/* Manipulate protcol part. */
-	p = nf_nat_proto_find_get(proto);
-	if (!p->manip_pkt(pskb, iphdroff, target, maniptype)) {
-		nf_nat_proto_put(p);
+
+	/* rcu_read_lock()ed by nf_hook_slow */
+	p = __nf_nat_proto_find(proto);
+	if (!p->manip_pkt(pskb, iphdroff, target, maniptype))
 		return 0;
-	}
-	nf_nat_proto_put(p);
 
 	iph = (void *)(*pskb)->data + iphdroff;
 
@@ -431,6 +429,7 @@ int nf_nat_icmp_reply_translation(struct nf_conn *ct,
 		struct icmphdr icmp;
 		struct iphdr ip;
 	} *inside;
+	struct nf_conntrack_l4proto *l4proto;
 	struct nf_conntrack_tuple inner, target;
 	int hdrlen = (*pskb)->nh.iph->ihl * 4;
 	enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo);
@@ -466,16 +465,16 @@ int nf_nat_icmp_reply_translation(struct nf_conn *ct,
 	DEBUGP("icmp_reply_translation: translating error %p manp %u dir %s\n",
 	       *pskb, manip, dir == IP_CT_DIR_ORIGINAL ? "ORIG" : "REPLY");
 
+	/* rcu_read_lock()ed by nf_hook_slow */
+	l4proto = __nf_ct_l4proto_find(PF_INET, inside->ip.protocol);
+
 	if (!nf_ct_get_tuple(*pskb,
 			     (*pskb)->nh.iph->ihl*4 + sizeof(struct icmphdr),
 			     (*pskb)->nh.iph->ihl*4 +
 			     sizeof(struct icmphdr) + inside->ip.ihl*4,
 			     (u_int16_t)AF_INET,
 			     inside->ip.protocol,
-			     &inner,
-			     l3proto,
-			     __nf_ct_l4proto_find((u_int16_t)PF_INET,
-						  inside->ip.protocol)))
+			     &inner, l3proto, l4proto))
 		return 0;
 
 	/* Change inner back to look like incoming packet.  We do the
@@ -529,7 +528,7 @@ int nf_nat_protocol_register(struct nf_nat_protocol *proto)
 		ret = -EBUSY;
 		goto out;
 	}
-	nf_nat_protos[proto->protonum] = proto;
+	rcu_assign_pointer(nf_nat_protos[proto->protonum], proto);
  out:
 	write_unlock_bh(&nf_nat_lock);
 	return ret;
@@ -540,11 +539,10 @@ EXPORT_SYMBOL(nf_nat_protocol_register);
 void nf_nat_protocol_unregister(struct nf_nat_protocol *proto)
 {
 	write_lock_bh(&nf_nat_lock);
-	nf_nat_protos[proto->protonum] = &nf_nat_unknown_protocol;
+	rcu_assign_pointer(nf_nat_protos[proto->protonum],
+			   &nf_nat_unknown_protocol);
 	write_unlock_bh(&nf_nat_lock);
-
-	/* Someone could be still looking at the proto in a bh. */
-	synchronize_net();
+	synchronize_rcu();
 }
 EXPORT_SYMBOL(nf_nat_protocol_unregister);
 
@@ -608,10 +606,10 @@ static int __init nf_nat_init(void)
 	/* Sew in builtin protocols. */
 	write_lock_bh(&nf_nat_lock);
 	for (i = 0; i < MAX_IP_NAT_PROTO; i++)
-		nf_nat_protos[i] = &nf_nat_unknown_protocol;
-	nf_nat_protos[IPPROTO_TCP] = &nf_nat_protocol_tcp;
-	nf_nat_protos[IPPROTO_UDP] = &nf_nat_protocol_udp;
-	nf_nat_protos[IPPROTO_ICMP] = &nf_nat_protocol_icmp;
+		rcu_assign_pointer(nf_nat_protos[i], &nf_nat_unknown_protocol);
+	rcu_assign_pointer(nf_nat_protos[IPPROTO_TCP], &nf_nat_protocol_tcp);
+	rcu_assign_pointer(nf_nat_protos[IPPROTO_UDP], &nf_nat_protocol_udp);
+	rcu_assign_pointer(nf_nat_protos[IPPROTO_ICMP], &nf_nat_protocol_icmp);
 	write_unlock_bh(&nf_nat_lock);
 
 	for (i = 0; i < nf_nat_htable_size; i++) {
@@ -619,8 +617,8 @@ static int __init nf_nat_init(void)
 	}
 
 	/* FIXME: Man, this is a hack.  <SIGH> */
-	NF_CT_ASSERT(nf_conntrack_destroyed == NULL);
-	nf_conntrack_destroyed = &nf_nat_cleanup_conntrack;
+	NF_CT_ASSERT(rcu_dereference(nf_conntrack_destroyed) == NULL);
+	rcu_assign_pointer(nf_conntrack_destroyed, nf_nat_cleanup_conntrack);
 
 	/* Initialize fake conntrack so that NAT will skip it */
 	nf_conntrack_untracked.status |= IPS_NAT_DONE_MASK;
@@ -644,7 +642,8 @@ static int clean_nat(struct nf_conn *i, void *data)
 static void __exit nf_nat_cleanup(void)
 {
 	nf_ct_iterate_cleanup(&clean_nat, NULL);
-	nf_conntrack_destroyed = NULL;
+	rcu_assign_pointer(nf_conntrack_destroyed, NULL);
+	synchronize_rcu();
 	vfree(bysource);
 	nf_ct_l3proto_put(l3proto);
 }

net/ipv4/tcp_cubic.c

@@ -26,16 +26,16 @@
 					  */
 #define	BICTCP_HZ		10	/* BIC HZ 2^10 = 1024 */
 
-static int fast_convergence = 1;
-static int max_increment = 16;
-static int beta = 819;		/* = 819/1024 (BICTCP_BETA_SCALE) */
-static int initial_ssthresh = 100;
-static int bic_scale = 41;
-static int tcp_friendliness = 1;
+static int fast_convergence __read_mostly = 1;
+static int max_increment __read_mostly = 16;
+static int beta __read_mostly = 819;	/* = 819/1024 (BICTCP_BETA_SCALE) */
+static int initial_ssthresh __read_mostly = 100;
+static int bic_scale __read_mostly = 41;
+static int tcp_friendliness __read_mostly = 1;
 
-static u32 cube_rtt_scale;
-static u32 beta_scale;
-static u64 cube_factor;
+static u32 cube_rtt_scale __read_mostly;
+static u32 beta_scale __read_mostly;
+static u64 cube_factor __read_mostly;
 
 /* Note parameters that are used for precomputing scale factors are read-only */
 module_param(fast_convergence, int, 0644);

net/ipv4/tcp_htcp.c

@@ -14,18 +14,19 @@
 #define BETA_MIN	(1<<6)	/* 0.5 with shift << 7 */
 #define BETA_MAX	102	/* 0.8 with shift << 7 */
 
-static int use_rtt_scaling = 1;
+static int use_rtt_scaling __read_mostly = 1;
 module_param(use_rtt_scaling, int, 0644);
 MODULE_PARM_DESC(use_rtt_scaling, "turn on/off RTT scaling");
 
-static int use_bandwidth_switch = 1;
+static int use_bandwidth_switch __read_mostly = 1;
 module_param(use_bandwidth_switch, int, 0644);
 MODULE_PARM_DESC(use_bandwidth_switch, "turn on/off bandwidth switcher");
 
 struct htcp {
 	u32	alpha;		/* Fixed point arith, << 7 */
 	u8	beta;           /* Fixed point arith, << 7 */
-	u8	modeswitch;     /* Delay modeswitch until we had at least one congestion event */
+	u8	modeswitch;	/* Delay modeswitch
+				   until we had at least one congestion event */
 	u16	pkts_acked;
 	u32	packetcount;
 	u32	minRTT;
@@ -44,14 +45,14 @@ struct htcp {
 	u32	lasttime;
 };
 
-static inline u32 htcp_cong_time(struct htcp *ca)
+static inline u32 htcp_cong_time(const struct htcp *ca)
 {
 	return jiffies - ca->last_cong;
 }
 
-static inline u32 htcp_ccount(struct htcp *ca)
+static inline u32 htcp_ccount(const struct htcp *ca)
 {
-	return htcp_cong_time(ca)/ca->minRTT;
+	return htcp_cong_time(ca) / ca->minRTT;
 }
 
 static inline void htcp_reset(struct htcp *ca)
@@ -67,10 +68,12 @@ static u32 htcp_cwnd_undo(struct sock *sk)
 {
 	const struct tcp_sock *tp = tcp_sk(sk);
 	struct htcp *ca = inet_csk_ca(sk);
+
 	ca->last_cong = ca->undo_last_cong;
 	ca->maxRTT = ca->undo_maxRTT;
 	ca->old_maxB = ca->undo_old_maxB;
-	return max(tp->snd_cwnd, (tp->snd_ssthresh<<7)/ca->beta);
+
+	return max(tp->snd_cwnd, (tp->snd_ssthresh << 7) / ca->beta);
 }
 
 static inline void measure_rtt(struct sock *sk)
@@ -78,17 +81,19 @@ static inline void measure_rtt(struct sock *sk)
 	const struct inet_connection_sock *icsk = inet_csk(sk);
 	const struct tcp_sock *tp = tcp_sk(sk);
 	struct htcp *ca = inet_csk_ca(sk);
-	u32 srtt = tp->srtt>>3;
+	u32 srtt = tp->srtt >> 3;
 
 	/* keep track of minimum RTT seen so far, minRTT is zero at first */
 	if (ca->minRTT > srtt || !ca->minRTT)
 		ca->minRTT = srtt;
 
 	/* max RTT */
-	if (icsk->icsk_ca_state == TCP_CA_Open && tp->snd_ssthresh < 0xFFFF && htcp_ccount(ca) > 3) {
+	if (icsk->icsk_ca_state == TCP_CA_Open
+	    && tp->snd_ssthresh < 0xFFFF && htcp_ccount(ca) > 3) {
 		if (ca->maxRTT < ca->minRTT)
 			ca->maxRTT = ca->minRTT;
-		if (ca->maxRTT < srtt && srtt <= ca->maxRTT+msecs_to_jiffies(20))
+		if (ca->maxRTT < srtt
+		    && srtt <= ca->maxRTT + msecs_to_jiffies(20))
 			ca->maxRTT = srtt;
 	}
 }
@@ -116,15 +121,16 @@ static void measure_achieved_throughput(struct sock *sk, u32 pkts_acked)
 
 	ca->packetcount += pkts_acked;
 
-	if (ca->packetcount >= tp->snd_cwnd - (ca->alpha>>7? : 1)
+	if (ca->packetcount >= tp->snd_cwnd - (ca->alpha >> 7 ? : 1)
 	    && now - ca->lasttime >= ca->minRTT
 	    && ca->minRTT > 0) {
-		__u32 cur_Bi = ca->packetcount*HZ/(now - ca->lasttime);
+		__u32 cur_Bi = ca->packetcount * HZ / (now - ca->lasttime);
+
 		if (htcp_ccount(ca) <= 3) {
 			/* just after backoff */
 			ca->minB = ca->maxB = ca->Bi = cur_Bi;
 		} else {
-			ca->Bi = (3*ca->Bi + cur_Bi)/4;
+			ca->Bi = (3 * ca->Bi + cur_Bi) / 4;
 			if (ca->Bi > ca->maxB)
 				ca->maxB = ca->Bi;
 			if (ca->minB > ca->maxB)
@@ -142,7 +148,7 @@ static inline void htcp_beta_update(struct htcp *ca, u32 minRTT, u32 maxRTT)
 		u32 old_maxB = ca->old_maxB;
 		ca->old_maxB = ca->maxB;
 
-		if (!between(5*maxB, 4*old_maxB, 6*old_maxB)) {
+		if (!between(5 * maxB, 4 * old_maxB, 6 * old_maxB)) {
 			ca->beta = BETA_MIN;
 			ca->modeswitch = 0;
 			return;
@@ -150,7 +156,7 @@ static inline void htcp_beta_update(struct htcp *ca, u32 minRTT, u32 maxRTT)
 	}
 
 	if (ca->modeswitch && minRTT > msecs_to_jiffies(10) && maxRTT) {
-		ca->beta = (minRTT<<7)/maxRTT;
+		ca->beta = (minRTT << 7) / maxRTT;
 		if (ca->beta < BETA_MIN)
 			ca->beta = BETA_MIN;
 		else if (ca->beta > BETA_MAX)
@@ -169,23 +175,26 @@ static inline void htcp_alpha_update(struct htcp *ca)
 
 	if (diff > HZ) {
 		diff -= HZ;
-		factor = 1+ ( 10*diff + ((diff/2)*(diff/2)/HZ) )/HZ;
+		factor = 1 + (10 * diff + ((diff / 2) * (diff / 2) / HZ)) / HZ;
 	}
 
 	if (use_rtt_scaling && minRTT) {
-		u32 scale = (HZ<<3)/(10*minRTT);
-		scale = min(max(scale, 1U<<2), 10U<<3); /* clamping ratio to interval [0.5,10]<<3 */
-		factor = (factor<<3)/scale;
+		u32 scale = (HZ << 3) / (10 * minRTT);
+
+		/* clamping ratio to interval [0.5,10]<<3 */
+		scale = min(max(scale, 1U << 2), 10U << 3);
+		factor = (factor << 3) / scale;
 		if (!factor)
 			factor = 1;
 	}
 
-	ca->alpha = 2*factor*((1<<7)-ca->beta);
+	ca->alpha = 2 * factor * ((1 << 7) - ca->beta);
 	if (!ca->alpha)
 		ca->alpha = ALPHA_BASE;
 }
 
-/* After we have the rtt data to calculate beta, we'd still prefer to wait one
+/*
+ * After we have the rtt data to calculate beta, we'd still prefer to wait one
  * rtt before we adjust our beta to ensure we are working from a consistent
  * data.
  *
@@ -202,15 +211,16 @@ static void htcp_param_update(struct sock *sk)
 	htcp_beta_update(ca, minRTT, maxRTT);
 	htcp_alpha_update(ca);
 
-	/* add slowly fading memory for maxRTT to accommodate routing changes etc */
+	/* add slowly fading memory for maxRTT to accommodate routing changes */
 	if (minRTT > 0 && maxRTT > minRTT)
-		ca->maxRTT = minRTT + ((maxRTT-minRTT)*95)/100;
+		ca->maxRTT = minRTT + ((maxRTT - minRTT) * 95) / 100;
 }
 
 static u32 htcp_recalc_ssthresh(struct sock *sk)
 {
 	const struct tcp_sock *tp = tcp_sk(sk);
 	const struct htcp *ca = inet_csk_ca(sk);
+
 	htcp_param_update(sk);
 	return max((tp->snd_cwnd * ca->beta) >> 7, 2U);
 }
@@ -227,7 +237,6 @@ static void htcp_cong_avoid(struct sock *sk, u32 ack, u32 rtt,
 	if (tp->snd_cwnd <= tp->snd_ssthresh)
 		tcp_slow_start(tp);
 	else {
-
 		measure_rtt(sk);
 
 		/* In dangerous area, increase slowly.

net/ipv6/netfilter/Kconfig

@@ -42,7 +42,8 @@ config IP6_NF_QUEUE
 
 config IP6_NF_IPTABLES
 	tristate "IP6 tables support (required for filtering)"
-	depends on INET && IPV6 && EXPERIMENTAL && NETFILTER_XTABLES
+	depends on INET && IPV6 && EXPERIMENTAL
+	select NETFILTER_XTABLES
 	help
 	  ip6tables is a general, extensible packet identification framework.
 	  Currently only the packet filtering and packet mangling subsystem

net/ipv6/netfilter/ip6t_LOG.c

@@ -501,7 +501,7 @@ static int __init ip6t_log_init(void)
 
 static void __exit ip6t_log_fini(void)
 {
-	nf_log_unregister_logger(&ip6t_logger);
+	nf_log_unregister(&ip6t_logger);
 	xt_unregister_target(&ip6t_log_reg);
 }
 

net/ipv6/netfilter/ip6t_mh.c

@@ -66,6 +66,13 @@ match(const struct sk_buff *skb,
 		return 0;
 	}
 
+	if (mh->ip6mh_proto != IPPROTO_NONE) {
+		duprintf("Dropping invalid MH Payload Proto: %u\n",
+			 mh->ip6mh_proto);
+		*hotdrop = 1;
+		return 0;
+	}
+
 	return type_match(mhinfo->types[0], mhinfo->types[1], mh->ip6mh_type,
 			  !!(mhinfo->invflags & IP6T_MH_INV_TYPE));
 }