[deliverable/linux.git] / net / netfilter / nfnetlink.c

/* Netfilter messages via netlink socket. Allows for user space
 * protocol helpers and general trouble making from userspace.
 *
 * (C) 2001 by Jay Schulist <jschlst@samba.org>,
 * (C) 2002-2005 by Harald Welte <laforge@gnumonks.org>
 * (C) 2005,2007 by Pablo Neira Ayuso <pablo@netfilter.org>
 *
 * Initial netfilter messages via netlink development funded and
 * generally made possible by Network Robots, Inc. (www.networkrobots.com)
 *
 * Further development of this code funded by Astaro AG (http://www.astaro.com)
 *
 * This software may be used and distributed according to the terms
 * of the GNU General Public License, incorporated herein by reference.
 */

#include <linux/module.h>
#include <linux/types.h>
#include <linux/socket.h>
#include <linux/kernel.h>
#include <linux/string.h>
#include <linux/sockios.h>
#include <linux/net.h>
#include <linux/skbuff.h>
#include <asm/uaccess.h>
#include <net/sock.h>
#include <linux/init.h>

#include <net/netlink.h>
#include <linux/netfilter/nfnetlink.h>

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NETFILTER);

static char __initdata nfversion[] = "0.30";

static struct {
	struct mutex				mutex;
	const struct nfnetlink_subsystem __rcu	*subsys;
} table[NFNL_SUBSYS_COUNT];

static const int nfnl_group2type[NFNLGRP_MAX+1] = {
	[NFNLGRP_CONNTRACK_NEW]		= NFNL_SUBSYS_CTNETLINK,
	[NFNLGRP_CONNTRACK_UPDATE]	= NFNL_SUBSYS_CTNETLINK,
	[NFNLGRP_CONNTRACK_DESTROY]	= NFNL_SUBSYS_CTNETLINK,
	[NFNLGRP_CONNTRACK_EXP_NEW]	= NFNL_SUBSYS_CTNETLINK_EXP,
	[NFNLGRP_CONNTRACK_EXP_UPDATE]	= NFNL_SUBSYS_CTNETLINK_EXP,
	[NFNLGRP_CONNTRACK_EXP_DESTROY] = NFNL_SUBSYS_CTNETLINK_EXP,
};

void nfnl_lock(__u8 subsys_id)
{
	mutex_lock(&table[subsys_id].mutex);
}
EXPORT_SYMBOL_GPL(nfnl_lock);

void nfnl_unlock(__u8 subsys_id)
{
	mutex_unlock(&table[subsys_id].mutex);
}
EXPORT_SYMBOL_GPL(nfnl_unlock);

#ifdef CONFIG_PROVE_LOCKING
int lockdep_nfnl_is_held(u8 subsys_id)
{
	return lockdep_is_held(&table[subsys_id].mutex);
}
EXPORT_SYMBOL_GPL(lockdep_nfnl_is_held);
#endif

int nfnetlink_subsys_register(const struct nfnetlink_subsystem *n)
{
	nfnl_lock(n->subsys_id);
	if (table[n->subsys_id].subsys) {
		nfnl_unlock(n->subsys_id);
		return -EBUSY;
	}
	rcu_assign_pointer(table[n->subsys_id].subsys, n);
	nfnl_unlock(n->subsys_id);

	return 0;
}
EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);

int nfnetlink_subsys_unregister(const struct nfnetlink_subsystem *n)
{
	nfnl_lock(n->subsys_id);
	table[n->subsys_id].subsys = NULL;
	nfnl_unlock(n->subsys_id);
	synchronize_rcu();
	return 0;
}
EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);

static inline const struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type)
{
	u_int8_t subsys_id = NFNL_SUBSYS_ID(type);

	if (subsys_id >= NFNL_SUBSYS_COUNT)
		return NULL;

	return rcu_dereference(table[subsys_id].subsys);
}

static inline const struct nfnl_callback *
nfnetlink_find_client(u_int16_t type, const struct nfnetlink_subsystem *ss)
{
	u_int8_t cb_id = NFNL_MSG_TYPE(type);

	if (cb_id >= ss->cb_count)
		return NULL;

	return &ss->cb[cb_id];
}

int nfnetlink_has_listeners(struct net *net, unsigned int group)
{
	return netlink_has_listeners(net->nfnl, group);
}
EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);

struct sk_buff *nfnetlink_alloc_skb(struct net *net, unsigned int size,
				    u32 dst_portid, gfp_t gfp_mask)
{
	return netlink_alloc_skb(net->nfnl, size, dst_portid, gfp_mask);
}
EXPORT_SYMBOL_GPL(nfnetlink_alloc_skb);

int nfnetlink_send(struct sk_buff *skb, struct net *net, u32 portid,
		   unsigned int group, int echo, gfp_t flags)
{
	return nlmsg_notify(net->nfnl, skb, portid, group, echo, flags);
}
EXPORT_SYMBOL_GPL(nfnetlink_send);

int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error)
{
	return netlink_set_err(net->nfnl, portid, group, error);
}
EXPORT_SYMBOL_GPL(nfnetlink_set_err);

int nfnetlink_unicast(struct sk_buff *skb, struct net *net, u32 portid,
		      int flags)
{
	return netlink_unicast(net->nfnl, skb, portid, flags);
}
EXPORT_SYMBOL_GPL(nfnetlink_unicast);

/* Process one complete nfnetlink message. */
static int nfnetlink_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
{
	struct net *net = sock_net(skb->sk);
	const struct nfnl_callback *nc;
	const struct nfnetlink_subsystem *ss;
	int type, err;

	/* All the messages must at least contain nfgenmsg */
	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg))
		return 0;

	type = nlh->nlmsg_type;
replay:
	rcu_read_lock();
	ss = nfnetlink_get_subsys(type);
	if (!ss) {
#ifdef CONFIG_MODULES
		rcu_read_unlock();
		request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
		rcu_read_lock();
		ss = nfnetlink_get_subsys(type);
		if (!ss)
#endif
		{
			rcu_read_unlock();
			return -EINVAL;
		}
	}

	nc = nfnetlink_find_client(type, ss);
	if (!nc) {
		rcu_read_unlock();
		return -EINVAL;
	}

	{
		int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
		u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
		struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
		struct nlattr *attr = (void *)nlh + min_len;
		int attrlen = nlh->nlmsg_len - min_len;
		__u8 subsys_id = NFNL_SUBSYS_ID(type);

		err = nla_parse(cda, ss->cb[cb_id].attr_count,
				attr, attrlen, ss->cb[cb_id].policy);
		if (err < 0) {
			rcu_read_unlock();
			return err;
		}

		if (nc->call_rcu) {
			err = nc->call_rcu(net->nfnl, skb, nlh,
					   (const struct nlattr **)cda);
			rcu_read_unlock();
		} else {
			rcu_read_unlock();
			nfnl_lock(subsys_id);
			if (rcu_dereference_protected(table[subsys_id].subsys,
				lockdep_is_held(&table[subsys_id].mutex)) != ss ||
			    nfnetlink_find_client(type, ss) != nc)
				err = -EAGAIN;
			else if (nc->call)
				err = nc->call(net->nfnl, skb, nlh,
						   (const struct nlattr **)cda);
			else
				err = -EINVAL;
			nfnl_unlock(subsys_id);
		}
		if (err == -EAGAIN)
			goto replay;
		return err;
	}
}

struct nfnl_err {
	struct list_head	head;
	struct nlmsghdr		*nlh;
	int			err;
};

static int nfnl_err_add(struct list_head *list, struct nlmsghdr *nlh, int err)
{
	struct nfnl_err *nfnl_err;

	nfnl_err = kmalloc(sizeof(struct nfnl_err), GFP_KERNEL);
	if (nfnl_err == NULL)
		return -ENOMEM;

	nfnl_err->nlh = nlh;
	nfnl_err->err = err;
	list_add_tail(&nfnl_err->head, list);

	return 0;
}

static void nfnl_err_del(struct nfnl_err *nfnl_err)
{
	list_del(&nfnl_err->head);
	kfree(nfnl_err);
}

static void nfnl_err_reset(struct list_head *err_list)
{
	struct nfnl_err *nfnl_err, *next;

	list_for_each_entry_safe(nfnl_err, next, err_list, head)
		nfnl_err_del(nfnl_err);
}

static void nfnl_err_deliver(struct list_head *err_list, struct sk_buff *skb)
{
	struct nfnl_err *nfnl_err, *next;

	list_for_each_entry_safe(nfnl_err, next, err_list, head) {
		netlink_ack(skb, nfnl_err->nlh, nfnl_err->err);
		nfnl_err_del(nfnl_err);
	}
}

static void nfnetlink_rcv_batch(struct sk_buff *skb, struct nlmsghdr *nlh,
				u_int16_t subsys_id)
{
	struct sk_buff *nskb, *oskb = skb;
	struct net *net = sock_net(skb->sk);
	const struct nfnetlink_subsystem *ss;
	const struct nfnl_callback *nc;
	bool success = true, done = false;
	static LIST_HEAD(err_list);
	int err;

	if (subsys_id >= NFNL_SUBSYS_COUNT)
		return netlink_ack(skb, nlh, -EINVAL);
replay:
	nskb = netlink_skb_clone(oskb, GFP_KERNEL);
	if (!nskb)
		return netlink_ack(oskb, nlh, -ENOMEM);

	nskb->sk = oskb->sk;
	skb = nskb;

	nfnl_lock(subsys_id);
	ss = rcu_dereference_protected(table[subsys_id].subsys,
				       lockdep_is_held(&table[subsys_id].mutex));
	if (!ss) {
#ifdef CONFIG_MODULES
		nfnl_unlock(subsys_id);
		request_module("nfnetlink-subsys-%d", subsys_id);
		nfnl_lock(subsys_id);
		ss = rcu_dereference_protected(table[subsys_id].subsys,
					       lockdep_is_held(&table[subsys_id].mutex));
		if (!ss)
#endif
		{
			nfnl_unlock(subsys_id);
			netlink_ack(skb, nlh, -EOPNOTSUPP);
			return kfree_skb(nskb);
		}
	}

	if (!ss->commit || !ss->abort) {
		nfnl_unlock(subsys_id);
		netlink_ack(skb, nlh, -EOPNOTSUPP);
		return kfree_skb(skb);
	}

	while (skb->len >= nlmsg_total_size(0)) {
		int msglen, type;

		nlh = nlmsg_hdr(skb);
		err = 0;

		if (nlh->nlmsg_len < NLMSG_HDRLEN) {
			err = -EINVAL;
			goto ack;
		}

		/* Only requests are handled by the kernel */
		if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
			err = -EINVAL;
			goto ack;
		}

		type = nlh->nlmsg_type;
		if (type == NFNL_MSG_BATCH_BEGIN) {
			/* Malformed: Batch begin twice */
			nfnl_err_reset(&err_list);
			success = false;
			goto done;
		} else if (type == NFNL_MSG_BATCH_END) {
			done = true;
			goto done;
		} else if (type < NLMSG_MIN_TYPE) {
			err = -EINVAL;
			goto ack;
		}

		/* We only accept a batch with messages for the same
		 * subsystem.
		 */
		if (NFNL_SUBSYS_ID(type) != subsys_id) {
			err = -EINVAL;
			goto ack;
		}

		nc = nfnetlink_find_client(type, ss);
		if (!nc) {
			err = -EINVAL;
			goto ack;
		}

		{
			int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
			u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
			struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
			struct nlattr *attr = (void *)nlh + min_len;
			int attrlen = nlh->nlmsg_len - min_len;

			err = nla_parse(cda, ss->cb[cb_id].attr_count,
					attr, attrlen, ss->cb[cb_id].policy);
			if (err < 0)
				goto ack;

			if (nc->call_batch) {
				err = nc->call_batch(net->nfnl, skb, nlh,
						     (const struct nlattr **)cda);
			}

			/* The lock was released to autoload some module, we
			 * have to abort and start from scratch using the
			 * original skb.
			 */
			if (err == -EAGAIN) {
				nfnl_err_reset(&err_list);
				ss->abort(oskb);
				nfnl_unlock(subsys_id);
				kfree_skb(nskb);
				goto replay;
			}
		}
ack:
		if (nlh->nlmsg_flags & NLM_F_ACK || err) {
			/* Errors are delivered once the full batch has been
			 * processed, this avoids that the same error is
			 * reported several times when replaying the batch.
			 */
			if (nfnl_err_add(&err_list, nlh, err) < 0) {
				/* We failed to enqueue an error, reset the
				 * list of errors and send OOM to userspace
				 * pointing to the batch header.
				 */
				nfnl_err_reset(&err_list);
				netlink_ack(skb, nlmsg_hdr(oskb), -ENOMEM);
				success = false;
				goto done;
			}
			/* We don't stop processing the batch on errors, thus,
			 * userspace gets all the errors that the batch
			 * triggers.
			 */
			if (err)
				success = false;
		}

		msglen = NLMSG_ALIGN(nlh->nlmsg_len);
		if (msglen > skb->len)
			msglen = skb->len;
		skb_pull(skb, msglen);
	}
done:
	if (success && done)
		ss->commit(oskb);
	else
		ss->abort(oskb);

	nfnl_err_deliver(&err_list, oskb);
	nfnl_unlock(subsys_id);
	kfree_skb(nskb);
}

static void nfnetlink_rcv(struct sk_buff *skb)
{
	struct nlmsghdr *nlh = nlmsg_hdr(skb);
	int msglen;

	if (nlh->nlmsg_len < NLMSG_HDRLEN ||
	    skb->len < nlh->nlmsg_len)
		return;

	if (!netlink_net_capable(skb, CAP_NET_ADMIN)) {
		netlink_ack(skb, nlh, -EPERM);
		return;
	}

	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN) {
		struct nfgenmsg *nfgenmsg;

		msglen = NLMSG_ALIGN(nlh->nlmsg_len);
		if (msglen > skb->len)
			msglen = skb->len;

		if (nlh->nlmsg_len < NLMSG_HDRLEN ||
		    skb->len < NLMSG_HDRLEN + sizeof(struct nfgenmsg))
			return;

		nfgenmsg = nlmsg_data(nlh);
		skb_pull(skb, msglen);
		nfnetlink_rcv_batch(skb, nlh, nfgenmsg->res_id);
	} else {
		netlink_rcv_skb(skb, &nfnetlink_rcv_msg);
	}
}

#ifdef CONFIG_MODULES
static int nfnetlink_bind(int group)
{
	const struct nfnetlink_subsystem *ss;
	int type = nfnl_group2type[group];

	rcu_read_lock();
	ss = nfnetlink_get_subsys(type);
	rcu_read_unlock();
	if (!ss)
		request_module("nfnetlink-subsys-%d", type);
	return 0;
}
#endif

static int __net_init nfnetlink_net_init(struct net *net)
{
	struct sock *nfnl;
	struct netlink_kernel_cfg cfg = {
		.groups	= NFNLGRP_MAX,
		.input	= nfnetlink_rcv,
#ifdef CONFIG_MODULES
		.bind	= nfnetlink_bind,
#endif
	};

	nfnl = netlink_kernel_create(net, NETLINK_NETFILTER, &cfg);
	if (!nfnl)
		return -ENOMEM;
	net->nfnl_stash = nfnl;
	rcu_assign_pointer(net->nfnl, nfnl);
	return 0;
}

static void __net_exit nfnetlink_net_exit_batch(struct list_head *net_exit_list)
{
	struct net *net;

	list_for_each_entry(net, net_exit_list, exit_list)
		RCU_INIT_POINTER(net->nfnl, NULL);
	synchronize_net();
	list_for_each_entry(net, net_exit_list, exit_list)
		netlink_kernel_release(net->nfnl_stash);
}

static struct pernet_operations nfnetlink_net_ops = {
	.init		= nfnetlink_net_init,
	.exit_batch	= nfnetlink_net_exit_batch,
};

static int __init nfnetlink_init(void)
{
	int i;

	for (i=0; i<NFNL_SUBSYS_COUNT; i++)
		mutex_init(&table[i].mutex);

	pr_info("Netfilter messages via NETLINK v%s.\n", nfversion);
	return register_pernet_subsys(&nfnetlink_net_ops);
}

static void __exit nfnetlink_exit(void)
{
	pr_info("Removing netfilter NETLINK layer.\n");
	unregister_pernet_subsys(&nfnetlink_net_ops);
}
module_init(nfnetlink_init);
module_exit(nfnetlink_exit);
Commit	Line	Data
f9e815b3 HW	1	/* Netfilter messages via netlink socket. Allows for user space
	2	* protocol helpers and general trouble making from userspace.
	3	*
	4	* (C) 2001 by Jay Schulist <jschlst@samba.org>,
	5	* (C) 2002-2005 by Harald Welte <laforge@gnumonks.org>
67ca3966	6	* (C) 2005,2007 by Pablo Neira Ayuso <pablo@netfilter.org>
f9e815b3 HW	7	*
	8	* Initial netfilter messages via netlink development funded and
	9	* generally made possible by Network Robots, Inc. (www.networkrobots.com)
	10	*
	11	* Further development of this code funded by Astaro AG (http://www.astaro.com)
	12	*
	13	* This software may be used and distributed according to the terms
	14	* of the GNU General Public License, incorporated herein by reference.
	15	*/
	16
f9e815b3 HW	17	#include <linux/module.h>
	18	#include <linux/types.h>
	19	#include <linux/socket.h>
	20	#include <linux/kernel.h>
f9e815b3 HW	21	#include <linux/string.h>
	22	#include <linux/sockios.h>
	23	#include <linux/net.h>
f9e815b3 HW	24	#include <linux/skbuff.h>
f9e815b3 HW	25	#include <asm/uaccess.h>
f9e815b3 HW	26	#include <net/sock.h>
f9e815b3 HW	27	#include <linux/init.h>
f9e815b3	28
573ce260	29	#include <net/netlink.h>
f9e815b3 HW	30	#include <linux/netfilter/nfnetlink.h>
	31
	32	MODULE_LICENSE("GPL");
4fdb3bb7 HW	33	MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>");
4fdb3bb7 HW	34	MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NETFILTER);
f9e815b3 HW	35
	36	static char __initdata nfversion[] = "0.30";
	37
c14b78e7 PNA	38	static struct {
	39	struct mutex mutex;
	40	const struct nfnetlink_subsystem __rcu *subsys;
	41	} table[NFNL_SUBSYS_COUNT];
f9e815b3	42
03292745 PNA	43	static const int nfnl_group2type[NFNLGRP_MAX+1] = {
	44	[NFNLGRP_CONNTRACK_NEW] = NFNL_SUBSYS_CTNETLINK,
	45	[NFNLGRP_CONNTRACK_UPDATE] = NFNL_SUBSYS_CTNETLINK,
	46	[NFNLGRP_CONNTRACK_DESTROY] = NFNL_SUBSYS_CTNETLINK,
	47	[NFNLGRP_CONNTRACK_EXP_NEW] = NFNL_SUBSYS_CTNETLINK_EXP,
	48	[NFNLGRP_CONNTRACK_EXP_UPDATE] = NFNL_SUBSYS_CTNETLINK_EXP,
	49	[NFNLGRP_CONNTRACK_EXP_DESTROY] = NFNL_SUBSYS_CTNETLINK_EXP,
	50	};
	51
c14b78e7	52	void nfnl_lock(__u8 subsys_id)
f9e815b3	53	{
c14b78e7	54	mutex_lock(&table[subsys_id].mutex);
f9e815b3	55	}
e6a7d3c0	56	EXPORT_SYMBOL_GPL(nfnl_lock);
f9e815b3	57
c14b78e7	58	void nfnl_unlock(__u8 subsys_id)
a3c5029c	59	{
c14b78e7	60	mutex_unlock(&table[subsys_id].mutex);
f9e815b3	61	}
e6a7d3c0	62	EXPORT_SYMBOL_GPL(nfnl_unlock);
f9e815b3	63
0eb5db7a PM	64	#ifdef CONFIG_PROVE_LOCKING
	65	int lockdep_nfnl_is_held(u8 subsys_id)
	66	{
	67	return lockdep_is_held(&table[subsys_id].mutex);
	68	}
	69	EXPORT_SYMBOL_GPL(lockdep_nfnl_is_held);
	70	#endif
	71
7c8d4cb4	72	int nfnetlink_subsys_register(const struct nfnetlink_subsystem *n)
f9e815b3	73	{
c14b78e7 PNA	74	nfnl_lock(n->subsys_id);
	75	if (table[n->subsys_id].subsys) {
	76	nfnl_unlock(n->subsys_id);
0ab43f84 HW	77	return -EBUSY;
0ab43f84 HW	78	}
c14b78e7 PNA	79	rcu_assign_pointer(table[n->subsys_id].subsys, n);
c14b78e7 PNA	80	nfnl_unlock(n->subsys_id);
f9e815b3 HW	81
	82	return 0;
	83	}
f4bc177f	84	EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);
f9e815b3	85
7c8d4cb4	86	int nfnetlink_subsys_unregister(const struct nfnetlink_subsystem *n)
f9e815b3	87	{
c14b78e7 PNA	88	nfnl_lock(n->subsys_id);
	89	table[n->subsys_id].subsys = NULL;
	90	nfnl_unlock(n->subsys_id);
6b75e3e8	91	synchronize_rcu();
f9e815b3 HW	92	return 0;
f9e815b3 HW	93	}
f4bc177f	94	EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);
f9e815b3	95
7c8d4cb4	96	static inline const struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type)
f9e815b3 HW	97	{
	98	u_int8_t subsys_id = NFNL_SUBSYS_ID(type);
	99
ac0f1d98	100	if (subsys_id >= NFNL_SUBSYS_COUNT)
f9e815b3 HW	101	return NULL;
f9e815b3 HW	102
c14b78e7	103	return rcu_dereference(table[subsys_id].subsys);
f9e815b3 HW	104	}
f9e815b3 HW	105
7c8d4cb4 PM	106	static inline const struct nfnl_callback *
7c8d4cb4 PM	107	nfnetlink_find_client(u_int16_t type, const struct nfnetlink_subsystem *ss)
f9e815b3 HW	108	{
f9e815b3 HW	109	u_int8_t cb_id = NFNL_MSG_TYPE(type);
601e68e1	110
67ca3966	111	if (cb_id >= ss->cb_count)
f9e815b3	112	return NULL;
f9e815b3 HW	113
	114	return &ss->cb[cb_id];
	115	}
	116
cd8c20b6	117	int nfnetlink_has_listeners(struct net *net, unsigned int group)
a2427692	118	{
cd8c20b6	119	return netlink_has_listeners(net->nfnl, group);
a2427692 PM	120	}
	121	EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);
	122
3ab1f683 PM	123	struct sk_buff nfnetlink_alloc_skb(struct net net, unsigned int size,
	124	u32 dst_portid, gfp_t gfp_mask)
	125	{
	126	return netlink_alloc_skb(net->nfnl, size, dst_portid, gfp_mask);
	127	}
	128	EXPORT_SYMBOL_GPL(nfnetlink_alloc_skb);
	129
ec464e5d	130	int nfnetlink_send(struct sk_buff skb, struct net net, u32 portid,
95c96174	131	unsigned int group, int echo, gfp_t flags)
f9e815b3	132	{
ec464e5d	133	return nlmsg_notify(net->nfnl, skb, portid, group, echo, flags);
f9e815b3	134	}
f4bc177f	135	EXPORT_SYMBOL_GPL(nfnetlink_send);
f9e815b3	136
ec464e5d	137	int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error)
dd5b6ce6	138	{
ec464e5d	139	return netlink_set_err(net->nfnl, portid, group, error);
dd5b6ce6 PNA	140	}
	141	EXPORT_SYMBOL_GPL(nfnetlink_set_err);
	142
ec464e5d PM	143	int nfnetlink_unicast(struct sk_buff skb, struct net net, u32 portid,
ec464e5d PM	144	int flags)
f9e815b3	145	{
ec464e5d	146	return netlink_unicast(net->nfnl, skb, portid, flags);
f9e815b3	147	}
f4bc177f	148	EXPORT_SYMBOL_GPL(nfnetlink_unicast);
f9e815b3 HW	149
f9e815b3 HW	150	/* Process one complete nfnetlink message. */
1d00a4eb	151	static int nfnetlink_rcv_msg(struct sk_buff skb, struct nlmsghdr nlh)
f9e815b3	152	{
cd8c20b6	153	struct net *net = sock_net(skb->sk);
7c8d4cb4 PM	154	const struct nfnl_callback *nc;
7c8d4cb4 PM	155	const struct nfnetlink_subsystem *ss;
1d00a4eb	156	int type, err;
f9e815b3	157
f9e815b3	158	/* All the messages must at least contain nfgenmsg */
573ce260	159	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg))
f9e815b3	160	return 0;
f9e815b3 HW	161
f9e815b3 HW	162	type = nlh->nlmsg_type;
e6a7d3c0	163	replay:
6b75e3e8	164	rcu_read_lock();
f9e815b3	165	ss = nfnetlink_get_subsys(type);
0ab43f84	166	if (!ss) {
95a5afca	167	#ifdef CONFIG_MODULES
6b75e3e8	168	rcu_read_unlock();
37d2e7a2	169	request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
6b75e3e8	170	rcu_read_lock();
37d2e7a2	171	ss = nfnetlink_get_subsys(type);
0ab43f84 HW	172	if (!ss)
0ab43f84 HW	173	#endif
6b75e3e8 ED	174	{
6b75e3e8 ED	175	rcu_read_unlock();
1d00a4eb	176	return -EINVAL;
6b75e3e8	177	}
0ab43f84	178	}
f9e815b3 HW	179
f9e815b3 HW	180	nc = nfnetlink_find_client(type, ss);
6b75e3e8 ED	181	if (!nc) {
6b75e3e8 ED	182	rcu_read_unlock();
1d00a4eb	183	return -EINVAL;
6b75e3e8	184	}
f9e815b3	185
f9e815b3	186	{
573ce260	187	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
e3730578	188	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
f49c857f PNA	189	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	190	struct nlattr attr = (void )nlh + min_len;
	191	int attrlen = nlh->nlmsg_len - min_len;
c14b78e7	192	__u8 subsys_id = NFNL_SUBSYS_ID(type);
f49c857f PNA	193
	194	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	195	attr, attrlen, ss->cb[cb_id].policy);
4009e188 TB	196	if (err < 0) {
4009e188 TB	197	rcu_read_unlock();
f49c857f	198	return err;
4009e188	199	}
601e68e1	200
6b75e3e8 ED	201	if (nc->call_rcu) {
	202	err = nc->call_rcu(net->nfnl, skb, nlh,
	203	(const struct nlattr **)cda);
	204	rcu_read_unlock();
	205	} else {
	206	rcu_read_unlock();
c14b78e7 PNA	207	nfnl_lock(subsys_id);
c14b78e7 PNA	208	if (rcu_dereference_protected(table[subsys_id].subsys,
9df9e783	209	lockdep_is_held(&table[subsys_id].mutex)) != ss \|\|
6b75e3e8 ED	210	nfnetlink_find_client(type, ss) != nc)
6b75e3e8 ED	211	err = -EAGAIN;
59560a38	212	else if (nc->call)
6b75e3e8 ED	213	err = nc->call(net->nfnl, skb, nlh,
6b75e3e8 ED	214	(const struct nlattr **)cda);
59560a38 TB	215	else
59560a38 TB	216	err = -EINVAL;
c14b78e7	217	nfnl_unlock(subsys_id);
6b75e3e8	218	}
e6a7d3c0 PNA	219	if (err == -EAGAIN)
	220	goto replay;
	221	return err;
f9e815b3	222	}
f9e815b3 HW	223	}
f9e815b3 HW	224
cbb8125e PNA	225	struct nfnl_err {
	226	struct list_head head;
	227	struct nlmsghdr *nlh;
	228	int err;
	229	};
	230
	231	static int nfnl_err_add(struct list_head list, struct nlmsghdr nlh, int err)
	232	{
	233	struct nfnl_err *nfnl_err;
	234
	235	nfnl_err = kmalloc(sizeof(struct nfnl_err), GFP_KERNEL);
	236	if (nfnl_err == NULL)
	237	return -ENOMEM;
	238
	239	nfnl_err->nlh = nlh;
	240	nfnl_err->err = err;
	241	list_add_tail(&nfnl_err->head, list);
	242
	243	return 0;
	244	}
	245
	246	static void nfnl_err_del(struct nfnl_err *nfnl_err)
	247	{
	248	list_del(&nfnl_err->head);
	249	kfree(nfnl_err);
	250	}
	251
	252	static void nfnl_err_reset(struct list_head *err_list)
	253	{
	254	struct nfnl_err nfnl_err, next;
	255
	256	list_for_each_entry_safe(nfnl_err, next, err_list, head)
	257	nfnl_err_del(nfnl_err);
	258	}
	259
	260	static void nfnl_err_deliver(struct list_head err_list, struct sk_buff skb)
	261	{
	262	struct nfnl_err nfnl_err, next;
	263
	264	list_for_each_entry_safe(nfnl_err, next, err_list, head) {
	265	netlink_ack(skb, nfnl_err->nlh, nfnl_err->err);
	266	nfnl_err_del(nfnl_err);
	267	}
	268	}
	269
0628b123 PNA	270	static void nfnetlink_rcv_batch(struct sk_buff skb, struct nlmsghdr nlh,
	271	u_int16_t subsys_id)
	272	{
	273	struct sk_buff nskb, oskb = skb;
	274	struct net *net = sock_net(skb->sk);
	275	const struct nfnetlink_subsystem *ss;
	276	const struct nfnl_callback *nc;
	277	bool success = true, done = false;
cbb8125e	278	static LIST_HEAD(err_list);
0628b123 PNA	279	int err;
	280
	281	if (subsys_id >= NFNL_SUBSYS_COUNT)
	282	return netlink_ack(skb, nlh, -EINVAL);
	283	replay:
	284	nskb = netlink_skb_clone(oskb, GFP_KERNEL);
	285	if (!nskb)
	286	return netlink_ack(oskb, nlh, -ENOMEM);
	287
	288	nskb->sk = oskb->sk;
	289	skb = nskb;
	290
	291	nfnl_lock(subsys_id);
	292	ss = rcu_dereference_protected(table[subsys_id].subsys,
	293	lockdep_is_held(&table[subsys_id].mutex));
	294	if (!ss) {
	295	#ifdef CONFIG_MODULES
	296	nfnl_unlock(subsys_id);
	297	request_module("nfnetlink-subsys-%d", subsys_id);
	298	nfnl_lock(subsys_id);
	299	ss = rcu_dereference_protected(table[subsys_id].subsys,
	300	lockdep_is_held(&table[subsys_id].mutex));
	301	if (!ss)
	302	#endif
	303	{
	304	nfnl_unlock(subsys_id);
ecd15dd7 DF	305	netlink_ack(skb, nlh, -EOPNOTSUPP);
ecd15dd7 DF	306	return kfree_skb(nskb);
0628b123 PNA	307	}
	308	}
	309
	310	if (!ss->commit \|\| !ss->abort) {
	311	nfnl_unlock(subsys_id);
ecd15dd7 DF	312	netlink_ack(skb, nlh, -EOPNOTSUPP);
ecd15dd7 DF	313	return kfree_skb(skb);
0628b123 PNA	314	}
	315
	316	while (skb->len >= nlmsg_total_size(0)) {
	317	int msglen, type;
	318
	319	nlh = nlmsg_hdr(skb);
	320	err = 0;
	321
	322	if (nlh->nlmsg_len < NLMSG_HDRLEN) {
	323	err = -EINVAL;
	324	goto ack;
	325	}
	326
	327	/* Only requests are handled by the kernel */
	328	if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
	329	err = -EINVAL;
	330	goto ack;
	331	}
	332
	333	type = nlh->nlmsg_type;
	334	if (type == NFNL_MSG_BATCH_BEGIN) {
	335	/* Malformed: Batch begin twice */
cbb8125e	336	nfnl_err_reset(&err_list);
0628b123 PNA	337	success = false;
	338	goto done;
	339	} else if (type == NFNL_MSG_BATCH_END) {
	340	done = true;
	341	goto done;
	342	} else if (type < NLMSG_MIN_TYPE) {
	343	err = -EINVAL;
	344	goto ack;
	345	}
	346
	347	/* We only accept a batch with messages for the same
	348	* subsystem.
	349	*/
	350	if (NFNL_SUBSYS_ID(type) != subsys_id) {
	351	err = -EINVAL;
	352	goto ack;
	353	}
	354
	355	nc = nfnetlink_find_client(type, ss);
	356	if (!nc) {
	357	err = -EINVAL;
	358	goto ack;
	359	}
	360
	361	{
	362	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
	363	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
	364	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	365	struct nlattr attr = (void )nlh + min_len;
	366	int attrlen = nlh->nlmsg_len - min_len;
	367
	368	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	369	attr, attrlen, ss->cb[cb_id].policy);
	370	if (err < 0)
	371	goto ack;
	372
	373	if (nc->call_batch) {
	374	err = nc->call_batch(net->nfnl, skb, nlh,
	375	(const struct nlattr **)cda);
	376	}
	377
	378	/* The lock was released to autoload some module, we
	379	* have to abort and start from scratch using the
	380	* original skb.
	381	*/
	382	if (err == -EAGAIN) {
cbb8125e	383	nfnl_err_reset(&err_list);
fc04733a	384	ss->abort(oskb);
0628b123 PNA	385	nfnl_unlock(subsys_id);
	386	kfree_skb(nskb);
	387	goto replay;
	388	}
	389	}
	390	ack:
	391	if (nlh->nlmsg_flags & NLM_F_ACK \|\| err) {
cbb8125e PNA	392	/* Errors are delivered once the full batch has been
	393	* processed, this avoids that the same error is
	394	* reported several times when replaying the batch.
	395	*/
	396	if (nfnl_err_add(&err_list, nlh, err) < 0) {
	397	/* We failed to enqueue an error, reset the
	398	* list of errors and send OOM to userspace
	399	* pointing to the batch header.
	400	*/
	401	nfnl_err_reset(&err_list);
	402	netlink_ack(skb, nlmsg_hdr(oskb), -ENOMEM);
	403	success = false;
	404	goto done;
	405	}
0628b123 PNA	406	/* We don't stop processing the batch on errors, thus,
	407	* userspace gets all the errors that the batch
	408	* triggers.
	409	*/
0628b123 PNA	410	if (err)
	411	success = false;
	412	}
	413
	414	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	415	if (msglen > skb->len)
	416	msglen = skb->len;
	417	skb_pull(skb, msglen);
	418	}
	419	done:
	420	if (success && done)
fc04733a	421	ss->commit(oskb);
0628b123	422	else
fc04733a	423	ss->abort(oskb);
0628b123	424
cbb8125e	425	nfnl_err_deliver(&err_list, oskb);
0628b123 PNA	426	nfnl_unlock(subsys_id);
	427	kfree_skb(nskb);
	428	}
	429
cd40b7d3	430	static void nfnetlink_rcv(struct sk_buff *skb)
f9e815b3	431	{
0628b123	432	struct nlmsghdr *nlh = nlmsg_hdr(skb);
0628b123 PNA	433	int msglen;
0628b123 PNA	434
0628b123 PNA	435	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	436	skb->len < nlh->nlmsg_len)
	437	return;
	438
90f62cf3	439	if (!netlink_net_capable(skb, CAP_NET_ADMIN)) {
cdbe7c2d JB	440	netlink_ack(skb, nlh, -EPERM);
	441	return;
	442	}
	443
0628b123 PNA	444	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN) {
	445	struct nfgenmsg *nfgenmsg;
	446
	447	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	448	if (msglen > skb->len)
	449	msglen = skb->len;
	450
	451	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	452	skb->len < NLMSG_HDRLEN + sizeof(struct nfgenmsg))
	453	return;
	454
	455	nfgenmsg = nlmsg_data(nlh);
	456	skb_pull(skb, msglen);
	457	nfnetlink_rcv_batch(skb, nlh, nfgenmsg->res_id);
	458	} else {
	459	netlink_rcv_skb(skb, &nfnetlink_rcv_msg);
	460	}
f9e815b3 HW	461	}
f9e815b3 HW	462
03292745	463	#ifdef CONFIG_MODULES
4f520900	464	static int nfnetlink_bind(int group)
03292745 PNA	465	{
	466	const struct nfnetlink_subsystem *ss;
	467	int type = nfnl_group2type[group];
	468
	469	rcu_read_lock();
	470	ss = nfnetlink_get_subsys(type);
03292745	471	rcu_read_unlock();
bfe4bc71 RGB	472	if (!ss)
bfe4bc71 RGB	473	request_module("nfnetlink-subsys-%d", type);
4f520900	474	return 0;
03292745 PNA	475	}
	476	#endif
	477
cd8c20b6	478	static int __net_init nfnetlink_net_init(struct net *net)
f9e815b3	479	{
cd8c20b6	480	struct sock *nfnl;
a31f2d17 PNA	481	struct netlink_kernel_cfg cfg = {
	482	.groups = NFNLGRP_MAX,
	483	.input = nfnetlink_rcv,
03292745 PNA	484	#ifdef CONFIG_MODULES
	485	.bind = nfnetlink_bind,
	486	#endif
a31f2d17	487	};
cd8c20b6	488
9f00d977	489	nfnl = netlink_kernel_create(net, NETLINK_NETFILTER, &cfg);
cd8c20b6 AD	490	if (!nfnl)
	491	return -ENOMEM;
	492	net->nfnl_stash = nfnl;
cf778b00	493	rcu_assign_pointer(net->nfnl, nfnl);
cd8c20b6	494	return 0;
f9e815b3 HW	495	}
f9e815b3 HW	496
cd8c20b6	497	static void __net_exit nfnetlink_net_exit_batch(struct list_head *net_exit_list)
f9e815b3	498	{
cd8c20b6	499	struct net *net;
f9e815b3	500
cd8c20b6	501	list_for_each_entry(net, net_exit_list, exit_list)
a9b3cd7f	502	RCU_INIT_POINTER(net->nfnl, NULL);
cd8c20b6 AD	503	synchronize_net();
	504	list_for_each_entry(net, net_exit_list, exit_list)
	505	netlink_kernel_release(net->nfnl_stash);
	506	}
f9e815b3	507
cd8c20b6 AD	508	static struct pernet_operations nfnetlink_net_ops = {
	509	.init = nfnetlink_net_init,
	510	.exit_batch = nfnetlink_net_exit_batch,
	511	};
	512
	513	static int __init nfnetlink_init(void)
	514	{
c14b78e7 PNA	515	int i;
	516
	517	for (i=0; i<NFNL_SUBSYS_COUNT; i++)
	518	mutex_init(&table[i].mutex);
	519
654d0fbd	520	pr_info("Netfilter messages via NETLINK v%s.\n", nfversion);
cd8c20b6	521	return register_pernet_subsys(&nfnetlink_net_ops);
f9e815b3 HW	522	}
f9e815b3 HW	523
cd8c20b6 AD	524	static void __exit nfnetlink_exit(void)
cd8c20b6 AD	525	{
654d0fbd	526	pr_info("Removing netfilter NETLINK layer.\n");
cd8c20b6 AD	527	unregister_pernet_subsys(&nfnetlink_net_ops);
cd8c20b6 AD	528	}
f9e815b3 HW	529	module_init(nfnetlink_init);
f9e815b3 HW	530	module_exit(nfnetlink_exit);