2 files changed, 213 insertions, 49 deletions

diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
index 2936fa3b6dc8..d0ff382c40ca 100644
--- a/net/netlink/af_netlink.c
+++ b/net/netlink/af_netlink.c
@@ -83,6 +83,11 @@ struct netlink_sock {
 	struct module		*module;
 };
 
+struct listeners_rcu_head {
+	struct rcu_head rcu_head;
+	void *ptr;
+};
+
 #define NETLINK_KERNEL_SOCKET	0x1
 #define NETLINK_RECV_PKTINFO	0x2
 #define NETLINK_BROADCAST_SEND_ERROR	0x4
@@ -1356,7 +1361,7 @@ static int netlink_recvmsg(struct kiocb *kiocb, struct socket *sock,
 	struct netlink_sock *nlk = nlk_sk(sk);
 	int noblock = flags&MSG_DONTWAIT;
 	size_t copied;
-	struct sk_buff *skb;
+	struct sk_buff *skb, *frag __maybe_unused = NULL;
 	int err;
 
 	if (flags&MSG_OOB)
@@ -1368,6 +1373,35 @@ static int netlink_recvmsg(struct kiocb *kiocb, struct socket *sock,
 	if (skb == NULL)
 		goto out;
 
+#ifdef CONFIG_COMPAT_NETLINK_MESSAGES
+	if (unlikely(skb_shinfo(skb)->frag_list)) {
+		bool need_compat = !!(flags & MSG_CMSG_COMPAT);
+
+		/*
+		 * If this skb has a frag_list, then here that means that
+		 * we will have to use the frag_list skb for compat tasks
+		 * and the regular skb for non-compat tasks.
+		 *
+		 * The skb might (and likely will) be cloned, so we can't
+		 * just reset frag_list and go on with things -- we need to
+		 * keep that. For the compat case that's easy -- simply get
+		 * a reference to the compat skb and free the regular one
+		 * including the frag. For the non-compat case, we need to
+		 * avoid sending the frag to the user -- so assign NULL but
+		 * restore it below before freeing the skb.
+		 */
+		if (need_compat) {
+			struct sk_buff *compskb = skb_shinfo(skb)->frag_list;
+			skb_get(compskb);
+			kfree_skb(skb);
+			skb = compskb;
+		} else {
+			frag = skb_shinfo(skb)->frag_list;
+			skb_shinfo(skb)->frag_list = NULL;
+		}
+	}
+#endif
+
 	msg->msg_namelen = 0;
 
 	copied = skb->len;
@@ -1398,6 +1432,11 @@ static int netlink_recvmsg(struct kiocb *kiocb, struct socket *sock,
 	siocb->scm->creds = *NETLINK_CREDS(skb);
 	if (flags & MSG_TRUNC)
 		copied = skb->len;
+
+#ifdef CONFIG_COMPAT_NETLINK_MESSAGES
+	skb_shinfo(skb)->frag_list = frag;
+#endif
+
 	skb_free_datagram(sk, skb);
 
 	if (nlk->cb && atomic_read(&sk->sk_rmem_alloc) <= sk->sk_rcvbuf / 2)
@@ -1453,7 +1492,8 @@ netlink_kernel_create(struct net *net, int unit, unsigned int groups,
 	if (groups < 32)
 		groups = 32;
 
-	listeners = kzalloc(NLGRPSZ(groups), GFP_KERNEL);
+	listeners = kzalloc(NLGRPSZ(groups) + sizeof(struct listeners_rcu_head),
+			    GFP_KERNEL);
 	if (!listeners)
 		goto out_sock_release;
 
@@ -1501,6 +1541,14 @@ netlink_kernel_release(struct sock *sk)
 EXPORT_SYMBOL(netlink_kernel_release);
 
 
+static void netlink_free_old_listeners(struct rcu_head *rcu_head)
+{
+	struct listeners_rcu_head *lrh;
+
+	lrh = container_of(rcu_head, struct listeners_rcu_head, rcu_head);
+	kfree(lrh->ptr);
+}
+
 /**
  * netlink_change_ngroups - change number of multicast groups
  *
@@ -1516,6 +1564,7 @@ EXPORT_SYMBOL(netlink_kernel_release);
 int netlink_change_ngroups(struct sock *sk, unsigned int groups)
 {
 	unsigned long *listeners, *old = NULL;
+	struct listeners_rcu_head *old_rcu_head;
 	struct netlink_table *tbl = &nl_table[sk->sk_protocol];
 	int err = 0;
 
@@ -1524,7 +1573,9 @@ int netlink_change_ngroups(struct sock *sk, unsigned int groups)
 
 	netlink_table_grab();
 	if (NLGRPSZ(tbl->groups) < NLGRPSZ(groups)) {
-		listeners = kzalloc(NLGRPSZ(groups), GFP_ATOMIC);
+		listeners = kzalloc(NLGRPSZ(groups) +
+				    sizeof(struct listeners_rcu_head),
+				    GFP_ATOMIC);
 		if (!listeners) {
 			err = -ENOMEM;
 			goto out_ungrab;
@@ -1532,16 +1583,24 @@ int netlink_change_ngroups(struct sock *sk, unsigned int groups)
 		old = tbl->listeners;
 		memcpy(listeners, old, NLGRPSZ(tbl->groups));
 		rcu_assign_pointer(tbl->listeners, listeners);
+		/*
+		 * Free the old memory after an RCU grace period so we
+		 * don't leak it. We use call_rcu() here in order to be
+		 * able to call this function from atomic contexts. The
+		 * allocation of this memory will have reserved enough
+		 * space for struct listeners_rcu_head at the end.
+		 */
+		old_rcu_head = (void *)(tbl->listeners +
+					NLGRPLONGS(tbl->groups));
+		old_rcu_head->ptr = old;
+		call_rcu(&old_rcu_head->rcu_head, netlink_free_old_listeners);
 	}
 	tbl->groups = groups;
 
  out_ungrab:
 	netlink_table_ungrab();
-	synchronize_rcu();
-	kfree(old);
 	return err;
 }
-EXPORT_SYMBOL(netlink_change_ngroups);
 
 /**
  * netlink_clear_multicast_users - kick off multicast listeners
@@ -1564,7 +1623,6 @@ void netlink_clear_multicast_users(struct sock *ksk, unsigned int group)
 
 	netlink_table_ungrab();
 }
-EXPORT_SYMBOL(netlink_clear_multicast_users);
 
 void netlink_set_nonroot(int protocol, unsigned int flags)
 {
@@ -1647,7 +1705,7 @@ errout:
 }
 
 int netlink_dump_start(struct sock *ssk, struct sk_buff *skb,
-		       struct nlmsghdr *nlh,
+		       const struct nlmsghdr *nlh,
 		       int (*dump)(struct sk_buff *skb,
 				   struct netlink_callback *),
 		       int (*done)(struct netlink_callback *))
diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c
index eed4c6a8afc0..66f6ba0bab11 100644
--- a/net/netlink/genetlink.c
+++ b/net/netlink/genetlink.c
@@ -18,8 +18,6 @@
 #include <net/sock.h>
 #include <net/genetlink.h>
 
-struct sock *genl_sock = NULL;
-
 static DEFINE_MUTEX(genl_mutex); /* serialization of message processing */
 
 static inline void genl_lock(void)
@@ -138,7 +136,7 @@ int genl_register_mc_group(struct genl_family *family,
 {
 	int id;
 	unsigned long *new_groups;
-	int err;
+	int err = 0;
 
 	BUG_ON(grp->name[0] == '\0');
 
@@ -175,10 +173,31 @@ int genl_register_mc_group(struct genl_family *family,
 		mc_groups_longs++;
 	}
 
-	err = netlink_change_ngroups(genl_sock,
-				     mc_groups_longs * BITS_PER_LONG);
-	if (err)
-		goto out;
+	if (family->netnsok) {
+		struct net *net;
+
+		rcu_read_lock();
+		for_each_net_rcu(net) {
+			err = netlink_change_ngroups(net->genl_sock,
+					mc_groups_longs * BITS_PER_LONG);
+			if (err) {
+				/*
+				 * No need to roll back, can only fail if
+				 * memory allocation fails and then the
+				 * number of _possible_ groups has been
+				 * increased on some sockets which is ok.
+				 */
+				rcu_read_unlock();
+				goto out;
+			}
+		}
+		rcu_read_unlock();
+	} else {
+		err = netlink_change_ngroups(init_net.genl_sock,
+					     mc_groups_longs * BITS_PER_LONG);
+		if (err)
+			goto out;
+	}
 
 	grp->id = id;
 	set_bit(id, mc_groups);
@@ -195,8 +214,14 @@ EXPORT_SYMBOL(genl_register_mc_group);
 static void __genl_unregister_mc_group(struct genl_family *family,
 				       struct genl_multicast_group *grp)
 {
+	struct net *net;
 	BUG_ON(grp->family != family);
-	netlink_clear_multicast_users(genl_sock, grp->id);
+
+	rcu_read_lock();
+	for_each_net_rcu(net)
+		netlink_clear_multicast_users(net->genl_sock, grp->id);
+	rcu_read_unlock();
+
 	clear_bit(grp->id, mc_groups);
 	list_del(&grp->list);
 	genl_ctrl_event(CTRL_CMD_DELMCAST_GRP, grp);
@@ -467,6 +492,7 @@ static int genl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
 {
 	struct genl_ops *ops;
 	struct genl_family *family;
+	struct net *net = sock_net(skb->sk);
 	struct genl_info info;
 	struct genlmsghdr *hdr = nlmsg_data(nlh);
 	int hdrlen, err;
@@ -475,6 +501,10 @@ static int genl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
 	if (family == NULL)
 		return -ENOENT;
 
+	/* this family doesn't exist in this netns */
+	if (!family->netnsok && !net_eq(net, &init_net))
+		return -ENOENT;
+
 	hdrlen = GENL_HDRLEN + family->hdrsize;
 	if (nlh->nlmsg_len < nlmsg_msg_size(hdrlen))
 		return -EINVAL;
@@ -492,7 +522,7 @@ static int genl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
 			return -EOPNOTSUPP;
 
 		genl_unlock();
-		err = netlink_dump_start(genl_sock, skb, nlh,
+		err = netlink_dump_start(net->genl_sock, skb, nlh,
 					 ops->dumpit, ops->done);
 		genl_lock();
 		return err;
@@ -514,6 +544,7 @@ static int genl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
 	info.genlhdr = nlmsg_data(nlh);
 	info.userhdr = nlmsg_data(nlh) + GENL_HDRLEN;
 	info.attrs = family->attrbuf;
+	genl_info_net_set(&info, net);
 
 	return ops->doit(skb, &info);
 }
@@ -534,6 +565,7 @@ static struct genl_family genl_ctrl = {
 	.name = "nlctrl",
 	.version = 0x2,
 	.maxattr = CTRL_ATTR_MAX,
+	.netnsok = true,
 };
 
 static int ctrl_fill_info(struct genl_family *family, u32 pid, u32 seq,
@@ -650,6 +682,7 @@ static int ctrl_dumpfamily(struct sk_buff *skb, struct netlink_callback *cb)
 
 	int i, n = 0;
 	struct genl_family *rt;
+	struct net *net = sock_net(skb->sk);
 	int chains_to_skip = cb->args[0];
 	int fams_to_skip = cb->args[1];
 
@@ -658,6 +691,8 @@ static int ctrl_dumpfamily(struct sk_buff *skb, struct netlink_callback *cb)
 			continue;
 		n = 0;
 		list_for_each_entry(rt, genl_family_chain(i), family_list) {
+			if (!rt->netnsok && !net_eq(net, &init_net))
+				continue;
 			if (++n < fams_to_skip)
 				continue;
 			if (ctrl_fill_info(rt, NETLINK_CB(cb->skb).pid,
@@ -729,6 +764,7 @@ static int ctrl_getfamily(struct sk_buff *skb, struct genl_info *info)
 	if (info->attrs[CTRL_ATTR_FAMILY_ID]) {
 		u16 id = nla_get_u16(info->attrs[CTRL_ATTR_FAMILY_ID]);
 		res = genl_family_find_byid(id);
+		err = -ENOENT;
 	}
 
 	if (info->attrs[CTRL_ATTR_FAMILY_NAME]) {
@@ -736,49 +772,61 @@ static int ctrl_getfamily(struct sk_buff *skb, struct genl_info *info)
 
 		name = nla_data(info->attrs[CTRL_ATTR_FAMILY_NAME]);
 		res = genl_family_find_byname(name);
+		err = -ENOENT;
 	}
 
-	if (res == NULL) {
-		err = -ENOENT;
-		goto errout;
+	if (res == NULL)
+		return err;
+
+	if (!res->netnsok && !net_eq(genl_info_net(info), &init_net)) {
+		/* family doesn't exist here */
+		return -ENOENT;
 	}
 
 	msg = ctrl_build_family_msg(res, info->snd_pid, info->snd_seq,
 				    CTRL_CMD_NEWFAMILY);
-	if (IS_ERR(msg)) {
-		err = PTR_ERR(msg);
-		goto errout;
-	}
+	if (IS_ERR(msg))
+		return PTR_ERR(msg);
 
-	err = genlmsg_reply(msg, info);
-errout:
-	return err;
+	return genlmsg_reply(msg, info);
 }
 
 static int genl_ctrl_event(int event, void *data)
 {
 	struct sk_buff *msg;
+	struct genl_family *family;
+	struct genl_multicast_group *grp;
 
-	if (genl_sock == NULL)
+	/* genl is still initialising */
+	if (!init_net.genl_sock)
 		return 0;
 
 	switch (event) {
 	case CTRL_CMD_NEWFAMILY:
 	case CTRL_CMD_DELFAMILY:
-		msg = ctrl_build_family_msg(data, 0, 0, event);
-		if (IS_ERR(msg))
-			return PTR_ERR(msg);
-
-		genlmsg_multicast(msg, 0, GENL_ID_CTRL, GFP_KERNEL);
+		family = data;
+		msg = ctrl_build_family_msg(family, 0, 0, event);
 		break;
 	case CTRL_CMD_NEWMCAST_GRP:
 	case CTRL_CMD_DELMCAST_GRP:
+		grp = data;
+		family = grp->family;
 		msg = ctrl_build_mcgrp_msg(data, 0, 0, event);
-		if (IS_ERR(msg))
-			return PTR_ERR(msg);
-
-		genlmsg_multicast(msg, 0, GENL_ID_CTRL, GFP_KERNEL);
 		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (IS_ERR(msg))
+		return PTR_ERR(msg);
+
+	if (!family->netnsok) {
+		genlmsg_multicast_netns(&init_net, msg, 0,
+					GENL_ID_CTRL, GFP_KERNEL);
+	} else {
+		rcu_read_lock();
+		genlmsg_multicast_allns(msg, 0, GENL_ID_CTRL, GFP_ATOMIC);
+		rcu_read_unlock();
 	}
 
 	return 0;
@@ -795,6 +843,33 @@ static struct genl_multicast_group notify_grp = {
 	.name		= "notify",
 };
 
+static int __net_init genl_pernet_init(struct net *net)
+{
+	/* we'll bump the group number right afterwards */
+	net->genl_sock = netlink_kernel_create(net, NETLINK_GENERIC, 0,
+					       genl_rcv, &genl_mutex,
+					       THIS_MODULE);
+
+	if (!net->genl_sock && net_eq(net, &init_net))
+		panic("GENL: Cannot initialize generic netlink\n");
+
+	if (!net->genl_sock)
+		return -ENOMEM;
+
+	return 0;
+}
+
+static void __net_exit genl_pernet_exit(struct net *net)
+{
+	netlink_kernel_release(net->genl_sock);
+	net->genl_sock = NULL;
+}
+
+static struct pernet_operations genl_pernet_ops = {
+	.init = genl_pernet_init,
+	.exit = genl_pernet_exit,
+};
+
 static int __init genl_init(void)
 {
 	int i, err;
@@ -804,36 +879,67 @@ static int __init genl_init(void)
 
 	err = genl_register_family(&genl_ctrl);
 	if (err < 0)
-		goto errout;
+		goto problem;
 
 	err = genl_register_ops(&genl_ctrl, &genl_ctrl_ops);
 	if (err < 0)
-		goto errout_register;
+		goto problem;
 
 	netlink_set_nonroot(NETLINK_GENERIC, NL_NONROOT_RECV);
 
-	/* we'll bump the group number right afterwards */
-	genl_sock = netlink_kernel_create(&init_net, NETLINK_GENERIC, 0,
-					  genl_rcv, &genl_mutex, THIS_MODULE);
-	if (genl_sock == NULL)
-		panic("GENL: Cannot initialize generic netlink\n");
+	err = register_pernet_subsys(&genl_pernet_ops);
+	if (err)
+		goto problem;
 
 	err = genl_register_mc_group(&genl_ctrl, &notify_grp);
 	if (err < 0)
-		goto errout_register;
+		goto problem;
 
 	return 0;
 
-errout_register:
-	genl_unregister_family(&genl_ctrl);
-errout:
+problem:
 	panic("GENL: Cannot register controller: %d\n", err);
 }
 
 subsys_initcall(genl_init);
 
-EXPORT_SYMBOL(genl_sock);
 EXPORT_SYMBOL(genl_register_ops);
 EXPORT_SYMBOL(genl_unregister_ops);
 EXPORT_SYMBOL(genl_register_family);
 EXPORT_SYMBOL(genl_unregister_family);
+
+static int genlmsg_mcast(struct sk_buff *skb, u32 pid, unsigned long group,
+			 gfp_t flags)
+{
+	struct sk_buff *tmp;
+	struct net *net, *prev = NULL;
+	int err;
+
+	for_each_net_rcu(net) {
+		if (prev) {
+			tmp = skb_clone(skb, flags);
+			if (!tmp) {
+				err = -ENOMEM;
+				goto error;
+			}
+			err = nlmsg_multicast(prev->genl_sock, tmp,
+					      pid, group, flags);
+			if (err)
+				goto error;
+		}
+
+		prev = net;
+	}
+
+	return nlmsg_multicast(prev->genl_sock, skb, pid, group, flags);
+ error:
+	kfree_skb(skb);
+	return err;
+}
+
+int genlmsg_multicast_allns(struct sk_buff *skb, u32 pid, unsigned int group,
+			    gfp_t flags)
+{
+	return genlmsg_mcast(skb, pid, group, flags);
+}
+EXPORT_SYMBOL(genlmsg_multicast_allns);