[PATCH 1/2] netfilter: ctnetlink: helper modules load-on-demand support

[PATCH 1/2] netfilter: ctnetlink: helper modules load-on-demand support

[Pablo Neira Ayuso]

This patch adds module loading for helpers via ctnetlink.

* Creation path: We support explicit and implicit helper assignation. For
  the explicit case, we try to load the module. If the module is correctly
  loaded and the helper is present, we return EAGAIN to re-start the
  creation. Otherwise, we return EOPNOTSUPP.
* Update path: release the spin lock, load the module and check. If it is
  present, then return EAGAIN to re-start the update.

This patch provides a refactorized function to lookup-and-set the
connection tracking helper. The function removes the exported symbol
__nf_ct_helper_find as it has not clients anymore.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---

 include/net/netfilter/nf_conntrack_helper.h |    5 +-
 net/netfilter/nf_conntrack_core.c           |   28 +----------
 net/netfilter/nf_conntrack_helper.c         |   32 ++++++++++++
 net/netfilter/nf_conntrack_netlink.c        |   70 ++++++++++++++++++++++++---
 4 files changed, 95 insertions(+), 40 deletions(-)

diff --git a/include/net/netfilter/nf_conntrack_helper.h b/include/net/netfilter/nf_conntrack_helper.h
index f8060ab..66d65a7 100644
--- a/include/net/netfilter/nf_conntrack_helper.h
+++ b/include/net/netfilter/nf_conntrack_helper.h
@@ -39,9 +39,6 @@ struct nf_conntrack_helper
 };
 
 extern struct nf_conntrack_helper *
-__nf_ct_helper_find(const struct nf_conntrack_tuple *tuple);
-
-extern struct nf_conntrack_helper *
 __nf_conntrack_helper_find_byname(const char *name);
 
 extern int nf_conntrack_helper_register(struct nf_conntrack_helper *);
@@ -49,6 +46,8 @@ extern void nf_conntrack_helper_unregister(struct nf_conntrack_helper *);
 
 extern struct nf_conn_help *nf_ct_helper_ext_add(struct nf_conn *ct, gfp_t gfp);
 
+extern int __nf_ct_try_assign_helper(struct nf_conn *ct, gfp_t flags);
+
 static inline struct nf_conn_help *nfct_help(const struct nf_conn *ct)
 {
 	return nf_ct_ext_find(ct, NF_CT_EXT_HELPER);
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index 622d7c6..1e649fb 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -588,14 +588,7 @@ init_conntrack(struct net *net,
 		nf_conntrack_get(&ct->master->ct_general);
 		NF_CT_STAT_INC(net, expect_new);
 	} else {
-		struct nf_conntrack_helper *helper;
-
-		helper = __nf_ct_helper_find(&repl_tuple);
-		if (helper) {
-			help = nf_ct_helper_ext_add(ct, GFP_ATOMIC);
-			if (help)
-				rcu_assign_pointer(help->helper, helper);
-		}
+		__nf_ct_try_assign_helper(ct, GFP_ATOMIC);
 		NF_CT_STAT_INC(net, new);
 	}
 
@@ -772,7 +765,6 @@ void nf_conntrack_alter_reply(struct nf_conn *ct,
 			      const struct nf_conntrack_tuple *newreply)
 {
 	struct nf_conn_help *help = nfct_help(ct);
-	struct nf_conntrack_helper *helper;
 
 	/* Should be unconfirmed, so not in hash table yet */
 	NF_CT_ASSERT(!nf_ct_is_confirmed(ct));
@@ -785,23 +777,7 @@ void nf_conntrack_alter_reply(struct nf_conn *ct,
 		return;
 
 	rcu_read_lock();
-	helper = __nf_ct_helper_find(newreply);
-	if (helper == NULL) {
-		if (help)
-			rcu_assign_pointer(help->helper, NULL);
-		goto out;
-	}
-
-	if (help == NULL) {
-		help = nf_ct_helper_ext_add(ct, GFP_ATOMIC);
-		if (help == NULL)
-			goto out;
-	} else {
-		memset(&help->help, 0, sizeof(help->help));
-	}
-
-	rcu_assign_pointer(help->helper, helper);
-out:
+	__nf_ct_try_assign_helper(ct, GFP_ATOMIC);
 	rcu_read_unlock();
 }
 EXPORT_SYMBOL_GPL(nf_conntrack_alter_reply);
diff --git a/net/netfilter/nf_conntrack_helper.c b/net/netfilter/nf_conntrack_helper.c
index c39b6a9..a51bdac 100644
--- a/net/netfilter/nf_conntrack_helper.c
+++ b/net/netfilter/nf_conntrack_helper.c
@@ -45,7 +45,7 @@ static unsigned int helper_hash(const struct nf_conntrack_tuple *tuple)
 		(__force __u16)tuple->src.u.all) % nf_ct_helper_hsize;
 }
 
-struct nf_conntrack_helper *
+static struct nf_conntrack_helper *
 __nf_ct_helper_find(const struct nf_conntrack_tuple *tuple)
 {
 	struct nf_conntrack_helper *helper;
@@ -63,7 +63,6 @@ __nf_ct_helper_find(const struct nf_conntrack_tuple *tuple)
 	}
 	return NULL;
 }
-EXPORT_SYMBOL_GPL(__nf_ct_helper_find);
 
 struct nf_conntrack_helper *
 __nf_conntrack_helper_find_byname(const char *name)
@@ -95,6 +94,35 @@ struct nf_conn_help *nf_ct_helper_ext_add(struct nf_conn *ct, gfp_t gfp)
 }
 EXPORT_SYMBOL_GPL(nf_ct_helper_ext_add);
 
+int __nf_ct_try_assign_helper(struct nf_conn *ct, gfp_t flags)
+{
+	int ret = 0;
+	struct nf_conntrack_helper *helper;
+	struct nf_conn_help *help = nfct_help(ct);
+
+	helper = __nf_ct_helper_find(&ct->tuplehash[IP_CT_DIR_REPLY].tuple);
+	if (helper == NULL) {
+		if (help)
+			rcu_assign_pointer(help->helper, NULL);
+		goto out;
+	}
+
+	if (help == NULL) {
+		help = nf_ct_helper_ext_add(ct, flags);
+		if (help == NULL) {
+			ret = -ENOMEM;
+			goto out;
+		}
+	} else {
+		memset(&help->help, 0, sizeof(help->help));
+	}
+
+	rcu_assign_pointer(help->helper, helper);
+out:
+	return ret;
+}
+EXPORT_SYMBOL_GPL(__nf_ct_try_assign_helper);
+
 static inline int unhelp(struct nf_conntrack_tuple_hash *i,
 			 const struct nf_conntrack_helper *me)
 {
diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c
index 49a04fa..4f6486c 100644
--- a/net/netfilter/nf_conntrack_netlink.c
+++ b/net/netfilter/nf_conntrack_netlink.c
@@ -917,8 +917,22 @@ ctnetlink_change_helper(struct nf_conn *ct, struct nlattr *cda[])
 	}
 
 	helper = __nf_conntrack_helper_find_byname(helpname);
-	if (helper == NULL)
+	if (helper == NULL) {
+#ifdef CONFIG_MODULES
+		spin_unlock_bh(&nf_conntrack_lock);
+
+		if (request_module("nfct-helper-%s", helpname) < 0) {
+			spin_lock_bh(&nf_conntrack_lock);
+			return -EOPNOTSUPP;
+		}
+
+		spin_lock_bh(&nf_conntrack_lock);
+		helper = __nf_conntrack_helper_find_byname(helpname);
+		if (helper)
+			return -EAGAIN;
+#endif
 		return -EOPNOTSUPP;
+	}
 
 	if (help) {
 		if (help->helper == helper)
@@ -1082,7 +1096,6 @@ ctnetlink_create_conntrack(struct nlattr *cda[],
 {
 	struct nf_conn *ct;
 	int err = -EINVAL;
-	struct nf_conn_help *help;
 	struct nf_conntrack_helper *helper;
 
 	ct = nf_conntrack_alloc(&init_net, otuple, rtuple, GFP_KERNEL);
@@ -1097,16 +1110,55 @@ ctnetlink_create_conntrack(struct nlattr *cda[],
 	ct->status |= IPS_CONFIRMED;
 
 	rcu_read_lock();
-	helper = __nf_ct_helper_find(rtuple);
-	if (helper) {
-		help = nf_ct_helper_ext_add(ct, GFP_ATOMIC);
-		if (help == NULL) {
+ 	if (cda[CTA_HELP]) {
+ 		char *helpname;
+ 
+ 		err = ctnetlink_parse_help(cda[CTA_HELP], &helpname);
+ 		if (err < 0) {
+			rcu_read_unlock();
+			goto err;
+		}
+
+		helper = __nf_conntrack_helper_find_byname(helpname);
+		if (helper == NULL) {
+			rcu_read_unlock();
+#ifdef CONFIG_MODULES
+			if (request_module("nfct-helper-%s", helpname) < 0) {
+				err = -EOPNOTSUPP;
+				goto err;
+			}
+
+			rcu_read_lock();
+			helper = __nf_conntrack_helper_find_byname(helpname);
+			if (helper) {
+				rcu_read_unlock();
+				err = -EAGAIN;
+				goto err;
+			}
+			rcu_read_unlock();
+#endif
+			err = -EOPNOTSUPP;
+			goto err;
+		} else {
+			struct nf_conn_help *help;
+
+			help = nf_ct_helper_ext_add(ct, GFP_ATOMIC);
+			if (help == NULL) {
+				rcu_read_unlock();
+				err = -ENOMEM;
+				goto err;
+			}
+
+			/* not in hash table yet so not strictly necessary */
+			rcu_assign_pointer(help->helper, helper);
+		}
+	} else {
+		/* try an implicit helper assignation */
+		err = __nf_ct_try_assign_helper(ct, GFP_ATOMIC);
+		if (err < 0) {
 			rcu_read_unlock();
-			err = -ENOMEM;
 			goto err;
 		}
-		/* not in hash table yet so not strictly necessary */
-		rcu_assign_pointer(help->helper, helper);
 	}
 
 	if (cda[CTA_STATUS]) {

[PATCH 2/2] netfilter: ctnetlink: deliver events for conntracks changed from userspace

[Pablo Neira Ayuso]

As for now, the creation and update of conntracks via ctnetlink do not
propagate an event to userspace. This can result in inconsistent situations
if several userspace processes modify the connection tracking table by means
of ctnetlink at the same time. Specifically, using the conntrack command
line tool and conntrackd at the same time can trigger unconsistencies.

This patch also modifies the event cache infrastructure to pass the
process PID and the ECHO flag to nfnetlink_send() to report back
to userspace if the process that triggered the change needs so.
Based on a suggestion from Patrick McHardy.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---

 include/net/netfilter/nf_conntrack.h        |    2 -
 include/net/netfilter/nf_conntrack_ecache.h |   57 ++++++++++++++++-
 include/net/netfilter/nf_conntrack_expect.h |    2 +
 net/netfilter/nf_conntrack_core.c           |   25 ++++++--
 net/netfilter/nf_conntrack_ecache.c         |   14 +++-
 net/netfilter/nf_conntrack_expect.c         |   43 +++++++++++--
 net/netfilter/nf_conntrack_netlink.c        |   88 ++++++++++++++++++++++-----
 7 files changed, 197 insertions(+), 34 deletions(-)

diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h
index f11255e..2e0c536 100644
--- a/include/net/netfilter/nf_conntrack.h
+++ b/include/net/netfilter/nf_conntrack.h
@@ -199,7 +199,7 @@ __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(struct net *net);
+extern void nf_conntrack_flush(struct net *net, u32 pid, int report);
 
 extern bool nf_ct_get_tuplepr(const struct sk_buff *skb,
 			      unsigned int nhoff, u_int16_t l3num,
diff --git a/include/net/netfilter/nf_conntrack_ecache.h b/include/net/netfilter/nf_conntrack_ecache.h
index 1285ff2..0ff0dc6 100644
--- a/include/net/netfilter/nf_conntrack_ecache.h
+++ b/include/net/netfilter/nf_conntrack_ecache.h
@@ -17,6 +17,13 @@ struct nf_conntrack_ecache {
 	unsigned int events;
 };
 
+/* This structure is passed to event handler */
+struct nf_ct_event {
+	struct nf_conn *ct;
+	u32 pid;
+	int report;
+};
+
 extern struct atomic_notifier_head nf_conntrack_chain;
 extern int nf_conntrack_register_notifier(struct notifier_block *nb);
 extern int nf_conntrack_unregister_notifier(struct notifier_block *nb);
@@ -39,22 +46,56 @@ nf_conntrack_event_cache(enum ip_conntrack_events event, struct nf_conn *ct)
 	local_bh_enable();
 }
 
-static inline void nf_conntrack_event(enum ip_conntrack_events event,
-				      struct nf_conn *ct)
+static inline void
+nf_conntrack_event_report(enum ip_conntrack_events event,
+			  struct nf_conn *ct,
+			  u32 pid,
+			  int report)
 {
+	struct nf_ct_event item = {
+		.ct 	= ct,
+		.pid	= pid,
+		.report = report
+	};
 	if (nf_ct_is_confirmed(ct) && !nf_ct_is_dying(ct))
-		atomic_notifier_call_chain(&nf_conntrack_chain, event, ct);
+		atomic_notifier_call_chain(&nf_conntrack_chain, event, &item);
 }
 
+static inline void
+nf_conntrack_event(enum ip_conntrack_events event, struct nf_conn *ct)
+{
+	nf_conntrack_event_report(event, ct, 0, 0);
+}
+
+struct nf_exp_event {
+	struct nf_conntrack_expect *exp;
+	u32 pid;
+	int report;
+};
+
 extern struct atomic_notifier_head nf_ct_expect_chain;
 extern int nf_ct_expect_register_notifier(struct notifier_block *nb);
 extern int nf_ct_expect_unregister_notifier(struct notifier_block *nb);
 
 static inline void
+nf_ct_expect_event_report(enum ip_conntrack_expect_events event,
+			  struct nf_conntrack_expect *exp,
+			  u32 pid,
+			  int report)
+{
+	struct nf_exp_event item = {
+		.exp	= exp,
+		.pid	= pid,
+		.report = report
+	};
+	atomic_notifier_call_chain(&nf_ct_expect_chain, event, &item);
+}
+
+static inline void
 nf_ct_expect_event(enum ip_conntrack_expect_events event,
 		   struct nf_conntrack_expect *exp)
 {
-	atomic_notifier_call_chain(&nf_ct_expect_chain, event, exp);
+	nf_ct_expect_event_report(event, exp, 0, 0);
 }
 
 extern int nf_conntrack_ecache_init(struct net *net);
@@ -66,9 +107,17 @@ static inline void nf_conntrack_event_cache(enum ip_conntrack_events event,
 					    struct nf_conn *ct) {}
 static inline void nf_conntrack_event(enum ip_conntrack_events event,
 				      struct nf_conn *ct) {}
+static inline void nf_conntrack_event_report(enum ip_conntrack_events event,
+					     struct nf_conn *ct,
+					     u32 pid,
+					     int report) {}
 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_expect_event_report(enum ip_conntrack_expect_events e,
+					     struct nf_conntrack_expect *exp,
+ 					     u32 pid,
+ 					     int report) {}
 static inline void nf_ct_event_cache_flush(struct net *net) {}
 
 static inline int nf_conntrack_ecache_init(struct net *net)
diff --git a/include/net/netfilter/nf_conntrack_expect.h b/include/net/netfilter/nf_conntrack_expect.h
index 37a7fc1..ab17a15 100644
--- a/include/net/netfilter/nf_conntrack_expect.h
+++ b/include/net/netfilter/nf_conntrack_expect.h
@@ -100,6 +100,8 @@ void nf_ct_expect_init(struct nf_conntrack_expect *, unsigned int, u_int8_t,
 		       u_int8_t, const __be16 *, const __be16 *);
 void nf_ct_expect_put(struct nf_conntrack_expect *exp);
 int nf_ct_expect_related(struct nf_conntrack_expect *expect);
+int nf_ct_expect_related_report(struct nf_conntrack_expect *expect, 
+				u32 pid, int report);
 
 #endif /*_NF_CONNTRACK_EXPECT_H*/
 
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index 1e649fb..dc3fea0 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -181,7 +181,8 @@ destroy_conntrack(struct nf_conntrack *nfct)
 	NF_CT_ASSERT(atomic_read(&nfct->use) == 0);
 	NF_CT_ASSERT(!timer_pending(&ct->timeout));
 
-	nf_conntrack_event(IPCT_DESTROY, ct);
+	if (!test_bit(IPS_DYING_BIT, &ct->status))
+		nf_conntrack_event(IPCT_DESTROY, ct);
 	set_bit(IPS_DYING_BIT, &ct->status);
 
 	/* To make sure we don't get any weird locking issues here:
@@ -972,8 +973,20 @@ void nf_ct_iterate_cleanup(struct net *net,
 }
 EXPORT_SYMBOL_GPL(nf_ct_iterate_cleanup);
 
+struct __nf_ct_flush_report {
+	u32 pid;
+	int report;
+};
+
 static int kill_all(struct nf_conn *i, void *data)
 {
+	struct __nf_ct_flush_report *fr = (struct __nf_ct_flush_report *)data;
+
+	/* get_next_corpse sets the dying bit for us */
+	nf_conntrack_event_report(IPCT_DESTROY,
+				  i,
+				  fr->pid,
+				  fr->report);
 	return 1;
 }
 
@@ -987,9 +1000,13 @@ void nf_ct_free_hashtable(struct hlist_head *hash, int vmalloced, unsigned int s
 }
 EXPORT_SYMBOL_GPL(nf_ct_free_hashtable);
 
-void nf_conntrack_flush(struct net *net)
+void nf_conntrack_flush(struct net *net, u32 pid, int report)
 {
-	nf_ct_iterate_cleanup(net, kill_all, NULL);
+	struct __nf_ct_flush_report fr = {
+		.pid 	= pid,
+		.report = report,
+	};
+	nf_ct_iterate_cleanup(net, kill_all, &fr);
 }
 EXPORT_SYMBOL_GPL(nf_conntrack_flush);
 
@@ -1005,7 +1022,7 @@ static void nf_conntrack_cleanup_net(struct net *net)
 	nf_ct_event_cache_flush(net);
 	nf_conntrack_ecache_fini(net);
  i_see_dead_people:
-	nf_conntrack_flush(net);
+	nf_conntrack_flush(net, 0, 0);
 	if (atomic_read(&net->ct.count) != 0) {
 		schedule();
 		goto i_see_dead_people;
diff --git a/net/netfilter/nf_conntrack_ecache.c b/net/netfilter/nf_conntrack_ecache.c
index a5f5e2e..dee4190 100644
--- a/net/netfilter/nf_conntrack_ecache.c
+++ b/net/netfilter/nf_conntrack_ecache.c
@@ -35,9 +35,17 @@ static inline void
 __nf_ct_deliver_cached_events(struct nf_conntrack_ecache *ecache)
 {
 	if (nf_ct_is_confirmed(ecache->ct) && !nf_ct_is_dying(ecache->ct)
-	    && ecache->events)
-		atomic_notifier_call_chain(&nf_conntrack_chain, ecache->events,
-				    ecache->ct);
+	    && ecache->events) {
+		struct nf_ct_event item = {
+			.ct 	= ecache->ct,
+			.pid	= 0,
+			.report	= 0
+		};
+
+		atomic_notifier_call_chain(&nf_conntrack_chain,
+					   ecache->events,
+					   &item);
+	}
 
 	ecache->events = 0;
 	nf_ct_put(ecache->ct);
diff --git a/net/netfilter/nf_conntrack_expect.c b/net/netfilter/nf_conntrack_expect.c
index 37a703b..3a8a34a 100644
--- a/net/netfilter/nf_conntrack_expect.c
+++ b/net/netfilter/nf_conntrack_expect.c
@@ -362,7 +362,7 @@ static inline int refresh_timer(struct nf_conntrack_expect *i)
 	return 1;
 }
 
-int nf_ct_expect_related(struct nf_conntrack_expect *expect)
+static inline int __nf_ct_expect_check(struct nf_conntrack_expect *expect)
 {
 	const struct nf_conntrack_expect_policy *p;
 	struct nf_conntrack_expect *i;
@@ -371,11 +371,8 @@ int nf_ct_expect_related(struct nf_conntrack_expect *expect)
 	struct net *net = nf_ct_exp_net(expect);
 	struct hlist_node *n;
 	unsigned int h;
-	int ret;
-
-	NF_CT_ASSERT(master_help);
+	int ret = 0;
 
-	spin_lock_bh(&nf_conntrack_lock);
 	if (!master_help->helper) {
 		ret = -ESHUTDOWN;
 		goto out;
@@ -409,18 +406,50 @@ int nf_ct_expect_related(struct nf_conntrack_expect *expect)
 			printk(KERN_WARNING
 			       "nf_conntrack: expectation table full\n");
 		ret = -EMFILE;
-		goto out;
 	}
+out:
+	return ret;
+}
+
+int nf_ct_expect_related(struct nf_conntrack_expect *expect)
+{
+	int ret;
+
+	spin_lock_bh(&nf_conntrack_lock);
+	ret = __nf_ct_expect_check(expect);
+	if (ret < 0)
+		goto out;
 
 	nf_ct_expect_insert(expect);
+	atomic_inc(&expect->use);
+	spin_unlock_bh(&nf_conntrack_lock);
 	nf_ct_expect_event(IPEXP_NEW, expect);
-	ret = 0;
+	nf_ct_expect_put(expect);
+	return ret;
 out:
 	spin_unlock_bh(&nf_conntrack_lock);
 	return ret;
 }
 EXPORT_SYMBOL_GPL(nf_ct_expect_related);
 
+int nf_ct_expect_related_report(struct nf_conntrack_expect *expect, 
+				u32 pid, int report)
+{
+	int ret;
+
+	spin_lock_bh(&nf_conntrack_lock);
+	ret = __nf_ct_expect_check(expect);
+	if (ret < 0)
+		goto out;
+	nf_ct_expect_insert(expect);
+out:
+	spin_unlock_bh(&nf_conntrack_lock);
+	if (ret == 0)
+		nf_ct_expect_event_report(IPEXP_NEW, expect, pid, report);
+	return ret;
+}
+EXPORT_SYMBOL_GPL(nf_ct_expect_related_report);
+
 #ifdef CONFIG_PROC_FS
 struct ct_expect_iter_state {
 	struct seq_net_private p;
diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c
index 4f6486c..ccc5ef1 100644
--- a/net/netfilter/nf_conntrack_netlink.c
+++ b/net/netfilter/nf_conntrack_netlink.c
@@ -410,7 +410,8 @@ static int ctnetlink_conntrack_event(struct notifier_block *this,
 	struct nlmsghdr *nlh;
 	struct nfgenmsg *nfmsg;
 	struct nlattr *nest_parms;
-	struct nf_conn *ct = (struct nf_conn *)ptr;
+	struct nf_ct_event *item = (struct nf_ct_event *)ptr;
+	struct nf_conn *ct = item->ct;
 	struct sk_buff *skb;
 	unsigned int type;
 	sk_buff_data_t b;
@@ -443,7 +444,7 @@ static int ctnetlink_conntrack_event(struct notifier_block *this,
 	b = skb->tail;
 
 	type |= NFNL_SUBSYS_CTNETLINK << 8;
-	nlh   = NLMSG_PUT(skb, 0, 0, type, sizeof(struct nfgenmsg));
+	nlh   = NLMSG_PUT(skb, item->pid, 0, type, sizeof(struct nfgenmsg));
 	nfmsg = NLMSG_DATA(nlh);
 
 	nlh->nlmsg_flags    = flags;
@@ -511,7 +512,7 @@ static int ctnetlink_conntrack_event(struct notifier_block *this,
 	rcu_read_unlock();
 
 	nlh->nlmsg_len = skb->tail - b;
-	nfnetlink_send(skb, 0, group, 0);
+	nfnetlink_send(skb, item->pid, group, item->report);
 	return NOTIFY_DONE;
 
 nla_put_failure:
@@ -722,7 +723,9 @@ ctnetlink_del_conntrack(struct sock *ctnl, struct sk_buff *skb,
 		err = ctnetlink_parse_tuple(cda, &tuple, CTA_TUPLE_REPLY, u3);
 	else {
 		/* Flush the whole table */
-		nf_conntrack_flush(&init_net);
+		nf_conntrack_flush(&init_net, 
+				   NETLINK_CB(skb).pid, 
+				   nlmsg_report(nlh));
 		return 0;
 	}
 
@@ -743,6 +746,14 @@ ctnetlink_del_conntrack(struct sock *ctnl, struct sk_buff *skb,
 		}
 	}
 
+	nf_conntrack_event_report(IPCT_DESTROY,
+				  ct,
+				  NETLINK_CB(skb).pid,
+				  nlmsg_report(nlh));
+
+	/* death_by_timeout would report the event again */
+	set_bit(IPS_DYING_BIT, &ct->status);
+
 	nf_ct_kill(ct);
 	nf_ct_put(ct);
 
@@ -1088,11 +1099,35 @@ ctnetlink_change_conntrack(struct nf_conn *ct, struct nlattr *cda[])
 	return 0;
 }
 
+static inline void
+ctnetlink_event_report(struct nf_conn *ct, u32 pid, int report)
+{
+	unsigned int events = 0;
+
+	if (test_bit(IPS_EXPECTED_BIT, &ct->status))
+		events |= IPCT_RELATED;
+	else
+		events |= IPCT_NEW;
+
+	nf_conntrack_event_report(IPCT_STATUS |
+				  IPCT_HELPER |
+				  IPCT_REFRESH |
+				  IPCT_PROTOINFO |
+				  IPCT_NATSEQADJ |
+				  IPCT_MARK |
+				  events,
+				  ct,
+				  pid,
+				  report);
+}
+
 static int
 ctnetlink_create_conntrack(struct nlattr *cda[],
 			   struct nf_conntrack_tuple *otuple,
 			   struct nf_conntrack_tuple *rtuple,
-			   struct nf_conn *master_ct)
+			   struct nf_conn *master_ct,
+			   u32 pid,
+			   int report)
 {
 	struct nf_conn *ct;
 	int err = -EINVAL;
@@ -1198,9 +1233,12 @@ ctnetlink_create_conntrack(struct nlattr *cda[],
 		ct->master = master_ct;
 	}
 
+	nf_conntrack_get(&ct->ct_general);
 	add_timer(&ct->timeout);
 	nf_conntrack_hash_insert(ct);
 	rcu_read_unlock();
+	ctnetlink_event_report(ct, pid, report);
+	nf_ct_put(ct);
 
 	return 0;
 
@@ -1265,7 +1303,9 @@ ctnetlink_new_conntrack(struct sock *ctnl, struct sk_buff *skb,
 			err = ctnetlink_create_conntrack(cda,
 							 &otuple,
 							 &rtuple,
-							 master_ct);
+							 master_ct,
+							 NETLINK_CB(skb).pid,
+							 nlmsg_report(nlh));
 		if (err < 0 && master_ct)
 			nf_ct_put(master_ct);
 
@@ -1277,6 +1317,8 @@ ctnetlink_new_conntrack(struct sock *ctnl, struct sk_buff *skb,
 	 * so there's no need to increase the refcount */
 	err = -EEXIST;
 	if (!(nlh->nlmsg_flags & NLM_F_EXCL)) {
+		struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h);
+
 		/* we only allow nat config for new conntracks */
 		if (cda[CTA_NAT_SRC] || cda[CTA_NAT_DST]) {
 			err = -EOPNOTSUPP;
@@ -1287,8 +1329,19 @@ ctnetlink_new_conntrack(struct sock *ctnl, struct sk_buff *skb,
 			err = -EOPNOTSUPP;
 			goto out_unlock;
 		}
-		err = ctnetlink_change_conntrack(nf_ct_tuplehash_to_ctrack(h),
-						 cda);
+
+		err = ctnetlink_change_conntrack(ct, cda);
+		if (err == 0) {
+			nf_conntrack_get(&ct->ct_general);
+			spin_unlock_bh(&nf_conntrack_lock);
+			ctnetlink_event_report(ct,
+					       NETLINK_CB(skb).pid,
+					       nlmsg_report(nlh));
+			nf_ct_put(ct);
+		} else
+			spin_unlock_bh(&nf_conntrack_lock);
+
+		return err;
 	}
 
 out_unlock:
@@ -1423,7 +1476,8 @@ static int ctnetlink_expect_event(struct notifier_block *this,
 {
 	struct nlmsghdr *nlh;
 	struct nfgenmsg *nfmsg;
-	struct nf_conntrack_expect *exp = (struct nf_conntrack_expect *)ptr;
+	struct nf_exp_event *item = (struct nf_exp_event *)ptr;
+	struct nf_conntrack_expect *exp = item->exp;
 	struct sk_buff *skb;
 	unsigned int type;
 	sk_buff_data_t b;
@@ -1445,7 +1499,7 @@ static int ctnetlink_expect_event(struct notifier_block *this,
 	b = skb->tail;
 
 	type |= NFNL_SUBSYS_CTNETLINK_EXP << 8;
-	nlh   = NLMSG_PUT(skb, 0, 0, type, sizeof(struct nfgenmsg));
+	nlh   = NLMSG_PUT(skb, item->pid, 0, type, sizeof(struct nfgenmsg));
 	nfmsg = NLMSG_DATA(nlh);
 
 	nlh->nlmsg_flags    = flags;
@@ -1459,7 +1513,7 @@ static int ctnetlink_expect_event(struct notifier_block *this,
 	rcu_read_unlock();
 
 	nlh->nlmsg_len = skb->tail - b;
-	nfnetlink_send(skb, 0, NFNLGRP_CONNTRACK_EXP_NEW, 0);
+	nfnetlink_send(skb, item->pid, NFNLGRP_CONNTRACK_EXP_NEW, item->report);
 	return NOTIFY_DONE;
 
 nla_put_failure:
@@ -1673,7 +1727,7 @@ ctnetlink_change_expect(struct nf_conntrack_expect *x, struct nlattr *cda[])
 }
 
 static int
-ctnetlink_create_expect(struct nlattr *cda[], u_int8_t u3)
+ctnetlink_create_expect(struct nlattr *cda[], u_int8_t u3, u32 pid, int report)
 {
 	struct nf_conntrack_tuple tuple, mask, master_tuple;
 	struct nf_conntrack_tuple_hash *h = NULL;
@@ -1720,7 +1774,7 @@ ctnetlink_create_expect(struct nlattr *cda[], u_int8_t u3)
 	memcpy(&exp->mask.src.u3, &mask.src.u3, sizeof(exp->mask.src.u3));
 	exp->mask.src.u.all = mask.src.u.all;
 
-	err = nf_ct_expect_related(exp);
+	err = nf_ct_expect_related_report(exp, pid, report);
 	nf_ct_expect_put(exp);
 
 out:
@@ -1753,8 +1807,12 @@ ctnetlink_new_expect(struct sock *ctnl, struct sk_buff *skb,
 	if (!exp) {
 		spin_unlock_bh(&nf_conntrack_lock);
 		err = -ENOENT;
-		if (nlh->nlmsg_flags & NLM_F_CREATE)
-			err = ctnetlink_create_expect(cda, u3);
+		if (nlh->nlmsg_flags & NLM_F_CREATE) {
+			err = ctnetlink_create_expect(cda,
+						      u3,
+						      NETLINK_CB(skb).pid,
+						      nlmsg_report(nlh));
+		}
 		return err;
 	}
 

Re: [PATCH 1/2] netfilter: ctnetlink: helper modules load-on-demand support

[Patrick McHardy]

Pablo Neira Ayuso wrote:
> This patch adds module loading for helpers via ctnetlink.
> 
> * Creation path: We support explicit and implicit helper assignation. For
>   the explicit case, we try to load the module. If the module is correctly
>   loaded and the helper is present, we return EAGAIN to re-start the
>   creation. Otherwise, we return EOPNOTSUPP.
> * Update path: release the spin lock, load the module and check. If it is
>   present, then return EAGAIN to re-start the update.
> 
> This patch provides a refactorized function to lookup-and-set the
> connection tracking helper. The function removes the exported symbol
> __nf_ct_helper_find as it has not clients anymore.

Applied, thanks Pablo.

Re: [PATCH 2/2] netfilter: ctnetlink: deliver events for conntracks changed from userspace

[Patrick McHardy]

Pablo Neira Ayuso wrote:
> As for now, the creation and update of conntracks via ctnetlink do not
> propagate an event to userspace. This can result in inconsistent situations
> if several userspace processes modify the connection tracking table by means
> of ctnetlink at the same time. Specifically, using the conntrack command
> line tool and conntrackd at the same time can trigger unconsistencies.
> 
> This patch also modifies the event cache infrastructure to pass the
> process PID and the ECHO flag to nfnetlink_send() to report back
> to userspace if the process that triggered the change needs so.
> Based on a suggestion from Patrick McHardy.

Applied, thanks.