[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);

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;
	}
}

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;
	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);
			kfree_skb(nskb);
			return netlink_ack(skb, nlh, -EOPNOTSUPP);
		}
	}

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

	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 */
			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) {
				ss->abort(skb);
				nfnl_unlock(subsys_id);
				kfree_skb(nskb);
				goto replay;
			}
		}
ack:
		if (nlh->nlmsg_flags & NLM_F_ACK || err) {
			/* We don't stop processing the batch on errors, thus,
			 * userspace gets all the errors that the batch
			 * triggers.
			 */
			netlink_ack(skb, nlh, err);
			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(skb);
	else
		ss->abort(skb);

	nfnl_unlock(subsys_id);
	kfree_skb(nskb);
}

static void nfnetlink_rcv(struct sk_buff *skb)
{
	struct nlmsghdr *nlh = nlmsg_hdr(skb);
	struct net *net = sock_net(skb->sk);
	int msglen;

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

	if (!ns_capable(net->user_ns, 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 void nfnetlink_bind(int group)
{
	const struct nfnetlink_subsystem *ss;
	int type = nfnl_group2type[group];

	rcu_read_lock();
	ss = nfnetlink_get_subsys(type);
	if (!ss) {
		rcu_read_unlock();
		request_module("nfnetlink-subsys-%d", type);
		return;
	}
	rcu_read_unlock();
}
#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
7c8d4cb4	64	int nfnetlink_subsys_register(const struct nfnetlink_subsystem *n)
f9e815b3	65	{
c14b78e7 PNA	66	nfnl_lock(n->subsys_id);
	67	if (table[n->subsys_id].subsys) {
	68	nfnl_unlock(n->subsys_id);
0ab43f84 HW	69	return -EBUSY;
0ab43f84 HW	70	}
c14b78e7 PNA	71	rcu_assign_pointer(table[n->subsys_id].subsys, n);
c14b78e7 PNA	72	nfnl_unlock(n->subsys_id);
f9e815b3 HW	73
	74	return 0;
	75	}
f4bc177f	76	EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);
f9e815b3	77
7c8d4cb4	78	int nfnetlink_subsys_unregister(const struct nfnetlink_subsystem *n)
f9e815b3	79	{
c14b78e7 PNA	80	nfnl_lock(n->subsys_id);
	81	table[n->subsys_id].subsys = NULL;
	82	nfnl_unlock(n->subsys_id);
6b75e3e8	83	synchronize_rcu();
f9e815b3 HW	84	return 0;
f9e815b3 HW	85	}
f4bc177f	86	EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);
f9e815b3	87
7c8d4cb4	88	static inline const struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type)
f9e815b3 HW	89	{
	90	u_int8_t subsys_id = NFNL_SUBSYS_ID(type);
	91
ac0f1d98	92	if (subsys_id >= NFNL_SUBSYS_COUNT)
f9e815b3 HW	93	return NULL;
f9e815b3 HW	94
c14b78e7	95	return rcu_dereference(table[subsys_id].subsys);
f9e815b3 HW	96	}
f9e815b3 HW	97
7c8d4cb4 PM	98	static inline const struct nfnl_callback *
7c8d4cb4 PM	99	nfnetlink_find_client(u_int16_t type, const struct nfnetlink_subsystem *ss)
f9e815b3 HW	100	{
f9e815b3 HW	101	u_int8_t cb_id = NFNL_MSG_TYPE(type);
601e68e1	102
67ca3966	103	if (cb_id >= ss->cb_count)
f9e815b3	104	return NULL;
f9e815b3 HW	105
	106	return &ss->cb[cb_id];
	107	}
	108
cd8c20b6	109	int nfnetlink_has_listeners(struct net *net, unsigned int group)
a2427692	110	{
cd8c20b6	111	return netlink_has_listeners(net->nfnl, group);
a2427692 PM	112	}
	113	EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);
	114
3ab1f683 PM	115	struct sk_buff nfnetlink_alloc_skb(struct net net, unsigned int size,
	116	u32 dst_portid, gfp_t gfp_mask)
	117	{
	118	return netlink_alloc_skb(net->nfnl, size, dst_portid, gfp_mask);
	119	}
	120	EXPORT_SYMBOL_GPL(nfnetlink_alloc_skb);
	121
ec464e5d	122	int nfnetlink_send(struct sk_buff skb, struct net net, u32 portid,
95c96174	123	unsigned int group, int echo, gfp_t flags)
f9e815b3	124	{
ec464e5d	125	return nlmsg_notify(net->nfnl, skb, portid, group, echo, flags);
f9e815b3	126	}
f4bc177f	127	EXPORT_SYMBOL_GPL(nfnetlink_send);
f9e815b3	128
ec464e5d	129	int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error)
dd5b6ce6	130	{
ec464e5d	131	return netlink_set_err(net->nfnl, portid, group, error);
dd5b6ce6 PNA	132	}
	133	EXPORT_SYMBOL_GPL(nfnetlink_set_err);
	134
ec464e5d PM	135	int nfnetlink_unicast(struct sk_buff skb, struct net net, u32 portid,
ec464e5d PM	136	int flags)
f9e815b3	137	{
ec464e5d	138	return netlink_unicast(net->nfnl, skb, portid, flags);
f9e815b3	139	}
f4bc177f	140	EXPORT_SYMBOL_GPL(nfnetlink_unicast);
f9e815b3 HW	141
f9e815b3 HW	142	/* Process one complete nfnetlink message. */
1d00a4eb	143	static int nfnetlink_rcv_msg(struct sk_buff skb, struct nlmsghdr nlh)
f9e815b3	144	{
cd8c20b6	145	struct net *net = sock_net(skb->sk);
7c8d4cb4 PM	146	const struct nfnl_callback *nc;
7c8d4cb4 PM	147	const struct nfnetlink_subsystem *ss;
1d00a4eb	148	int type, err;
f9e815b3	149
f9e815b3	150	/* All the messages must at least contain nfgenmsg */
573ce260	151	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg))
f9e815b3	152	return 0;
f9e815b3 HW	153
f9e815b3 HW	154	type = nlh->nlmsg_type;
e6a7d3c0	155	replay:
6b75e3e8	156	rcu_read_lock();
f9e815b3	157	ss = nfnetlink_get_subsys(type);
0ab43f84	158	if (!ss) {
95a5afca	159	#ifdef CONFIG_MODULES
6b75e3e8	160	rcu_read_unlock();
37d2e7a2	161	request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
6b75e3e8	162	rcu_read_lock();
37d2e7a2	163	ss = nfnetlink_get_subsys(type);
0ab43f84 HW	164	if (!ss)
0ab43f84 HW	165	#endif
6b75e3e8 ED	166	{
6b75e3e8 ED	167	rcu_read_unlock();
1d00a4eb	168	return -EINVAL;
6b75e3e8	169	}
0ab43f84	170	}
f9e815b3 HW	171
f9e815b3 HW	172	nc = nfnetlink_find_client(type, ss);
6b75e3e8 ED	173	if (!nc) {
6b75e3e8 ED	174	rcu_read_unlock();
1d00a4eb	175	return -EINVAL;
6b75e3e8	176	}
f9e815b3	177
f9e815b3	178	{
573ce260	179	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
e3730578	180	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
f49c857f PNA	181	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	182	struct nlattr attr = (void )nlh + min_len;
	183	int attrlen = nlh->nlmsg_len - min_len;
c14b78e7	184	__u8 subsys_id = NFNL_SUBSYS_ID(type);
f49c857f PNA	185
	186	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	187	attr, attrlen, ss->cb[cb_id].policy);
4009e188 TB	188	if (err < 0) {
4009e188 TB	189	rcu_read_unlock();
f49c857f	190	return err;
4009e188	191	}
601e68e1	192
6b75e3e8 ED	193	if (nc->call_rcu) {
	194	err = nc->call_rcu(net->nfnl, skb, nlh,
	195	(const struct nlattr **)cda);
	196	rcu_read_unlock();
	197	} else {
	198	rcu_read_unlock();
c14b78e7 PNA	199	nfnl_lock(subsys_id);
c14b78e7 PNA	200	if (rcu_dereference_protected(table[subsys_id].subsys,
9df9e783	201	lockdep_is_held(&table[subsys_id].mutex)) != ss \|\|
6b75e3e8 ED	202	nfnetlink_find_client(type, ss) != nc)
6b75e3e8 ED	203	err = -EAGAIN;
59560a38	204	else if (nc->call)
6b75e3e8 ED	205	err = nc->call(net->nfnl, skb, nlh,
6b75e3e8 ED	206	(const struct nlattr **)cda);
59560a38 TB	207	else
59560a38 TB	208	err = -EINVAL;
c14b78e7	209	nfnl_unlock(subsys_id);
6b75e3e8	210	}
e6a7d3c0 PNA	211	if (err == -EAGAIN)
	212	goto replay;
	213	return err;
f9e815b3	214	}
f9e815b3 HW	215	}
f9e815b3 HW	216
0628b123 PNA	217	static void nfnetlink_rcv_batch(struct sk_buff skb, struct nlmsghdr nlh,
	218	u_int16_t subsys_id)
	219	{
	220	struct sk_buff nskb, oskb = skb;
	221	struct net *net = sock_net(skb->sk);
	222	const struct nfnetlink_subsystem *ss;
	223	const struct nfnl_callback *nc;
	224	bool success = true, done = false;
	225	int err;
	226
	227	if (subsys_id >= NFNL_SUBSYS_COUNT)
	228	return netlink_ack(skb, nlh, -EINVAL);
	229	replay:
	230	nskb = netlink_skb_clone(oskb, GFP_KERNEL);
	231	if (!nskb)
	232	return netlink_ack(oskb, nlh, -ENOMEM);
	233
	234	nskb->sk = oskb->sk;
	235	skb = nskb;
	236
	237	nfnl_lock(subsys_id);
	238	ss = rcu_dereference_protected(table[subsys_id].subsys,
	239	lockdep_is_held(&table[subsys_id].mutex));
	240	if (!ss) {
	241	#ifdef CONFIG_MODULES
	242	nfnl_unlock(subsys_id);
	243	request_module("nfnetlink-subsys-%d", subsys_id);
	244	nfnl_lock(subsys_id);
	245	ss = rcu_dereference_protected(table[subsys_id].subsys,
	246	lockdep_is_held(&table[subsys_id].mutex));
	247	if (!ss)
	248	#endif
	249	{
	250	nfnl_unlock(subsys_id);
	251	kfree_skb(nskb);
	252	return netlink_ack(skb, nlh, -EOPNOTSUPP);
	253	}
	254	}
	255
	256	if (!ss->commit \|\| !ss->abort) {
	257	nfnl_unlock(subsys_id);
	258	kfree_skb(nskb);
	259	return netlink_ack(skb, nlh, -EOPNOTSUPP);
	260	}
	261
	262	while (skb->len >= nlmsg_total_size(0)) {
	263	int msglen, type;
	264
	265	nlh = nlmsg_hdr(skb);
	266	err = 0;
	267
	268	if (nlh->nlmsg_len < NLMSG_HDRLEN) {
	269	err = -EINVAL;
	270	goto ack;
	271	}
	272
	273	/* Only requests are handled by the kernel */
	274	if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
	275	err = -EINVAL;
	276	goto ack;
	277	}
	278
	279	type = nlh->nlmsg_type;
	280	if (type == NFNL_MSG_BATCH_BEGIN) {
281	/* Malformed: Batch begin twice */
282	success = false;
283	goto done;
284	} else if (type == NFNL_MSG_BATCH_END) {
285	done = true;
286	goto done;
287	} else if (type < NLMSG_MIN_TYPE) {
288	err = -EINVAL;
289	goto ack;
290	}
291
292	/* We only accept a batch with messages for the same
293	* subsystem.
294	*/
295	if (NFNL_SUBSYS_ID(type) != subsys_id) {
296	err = -EINVAL;
297	goto ack;
298	}
299
300	nc = nfnetlink_find_client(type, ss);
301	if (!nc) {
302	err = -EINVAL;
303	goto ack;
304	}
305
306	{
307	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
308	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
309	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
310	struct nlattr attr = (void )nlh + min_len;
311	int attrlen = nlh->nlmsg_len - min_len;
312
313	err = nla_parse(cda, ss->cb[cb_id].attr_count,
314	attr, attrlen, ss->cb[cb_id].policy);
315	if (err < 0)
316	goto ack;
317
318	if (nc->call_batch) {
319	err = nc->call_batch(net->nfnl, skb, nlh,
320	(const struct nlattr **)cda);
321	}
322
323	/* The lock was released to autoload some module, we
324	* have to abort and start from scratch using the
325	* original skb.
326	*/
327	if (err == -EAGAIN) {
328	ss->abort(skb);
329	nfnl_unlock(subsys_id);
330	kfree_skb(nskb);
331	goto replay;
332	}
333	}
334	ack:
335	if (nlh->nlmsg_flags & NLM_F_ACK \|\| err) {
336	/* We don't stop processing the batch on errors, thus,
337	* userspace gets all the errors that the batch
338	* triggers.
339	*/
340	netlink_ack(skb, nlh, err);
341	if (err)
342	success = false;
343	}
344
345	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
346	if (msglen > skb->len)
347	msglen = skb->len;
348	skb_pull(skb, msglen);
349	}
350	done:
351	if (success && done)
352	ss->commit(skb);
353	else
354	ss->abort(skb);
355
356	nfnl_unlock(subsys_id);
357	kfree_skb(nskb);
358	}
359
cd40b7d3	360	static void nfnetlink_rcv(struct sk_buff *skb)
f9e815b3	361	{
0628b123 PNA	362	struct nlmsghdr *nlh = nlmsg_hdr(skb);
	363	struct net *net = sock_net(skb->sk);
	364	int msglen;
	365
0628b123 PNA	366	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	367	skb->len < nlh->nlmsg_len)
	368	return;
	369
cdbe7c2d JB	370	if (!ns_capable(net->user_ns, CAP_NET_ADMIN)) {
	371	netlink_ack(skb, nlh, -EPERM);
	372	return;
	373	}
	374
0628b123 PNA	375	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN) {
	376	struct nfgenmsg *nfgenmsg;
	377
	378	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	379	if (msglen > skb->len)
	380	msglen = skb->len;
	381
	382	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	383	skb->len < NLMSG_HDRLEN + sizeof(struct nfgenmsg))
	384	return;
	385
	386	nfgenmsg = nlmsg_data(nlh);
	387	skb_pull(skb, msglen);
	388	nfnetlink_rcv_batch(skb, nlh, nfgenmsg->res_id);
	389	} else {
	390	netlink_rcv_skb(skb, &nfnetlink_rcv_msg);
	391	}
f9e815b3 HW	392	}
f9e815b3 HW	393
03292745 PNA	394	#ifdef CONFIG_MODULES
	395	static void nfnetlink_bind(int group)
	396	{
	397	const struct nfnetlink_subsystem *ss;
	398	int type = nfnl_group2type[group];
	399
	400	rcu_read_lock();
	401	ss = nfnetlink_get_subsys(type);
	402	if (!ss) {
	403	rcu_read_unlock();
	404	request_module("nfnetlink-subsys-%d", type);
	405	return;
	406	}
	407	rcu_read_unlock();
	408	}
	409	#endif
	410
cd8c20b6	411	static int __net_init nfnetlink_net_init(struct net *net)
f9e815b3	412	{
cd8c20b6	413	struct sock *nfnl;
a31f2d17 PNA	414	struct netlink_kernel_cfg cfg = {
	415	.groups = NFNLGRP_MAX,
	416	.input = nfnetlink_rcv,
03292745 PNA	417	#ifdef CONFIG_MODULES
	418	.bind = nfnetlink_bind,
	419	#endif
a31f2d17	420	};
cd8c20b6	421
9f00d977	422	nfnl = netlink_kernel_create(net, NETLINK_NETFILTER, &cfg);
cd8c20b6 AD	423	if (!nfnl)
	424	return -ENOMEM;
	425	net->nfnl_stash = nfnl;
cf778b00	426	rcu_assign_pointer(net->nfnl, nfnl);
cd8c20b6	427	return 0;
f9e815b3 HW	428	}
f9e815b3 HW	429
cd8c20b6	430	static void __net_exit nfnetlink_net_exit_batch(struct list_head *net_exit_list)
f9e815b3	431	{
cd8c20b6	432	struct net *net;
f9e815b3	433
cd8c20b6	434	list_for_each_entry(net, net_exit_list, exit_list)
a9b3cd7f	435	RCU_INIT_POINTER(net->nfnl, NULL);
cd8c20b6 AD	436	synchronize_net();
	437	list_for_each_entry(net, net_exit_list, exit_list)
	438	netlink_kernel_release(net->nfnl_stash);
	439	}
f9e815b3	440
cd8c20b6 AD	441	static struct pernet_operations nfnetlink_net_ops = {
	442	.init = nfnetlink_net_init,
	443	.exit_batch = nfnetlink_net_exit_batch,
	444	};
	445
	446	static int __init nfnetlink_init(void)
	447	{
c14b78e7 PNA	448	int i;
	449
	450	for (i=0; i<NFNL_SUBSYS_COUNT; i++)
	451	mutex_init(&table[i].mutex);
	452
654d0fbd	453	pr_info("Netfilter messages via NETLINK v%s.\n", nfversion);
cd8c20b6	454	return register_pernet_subsys(&nfnetlink_net_ops);
f9e815b3 HW	455	}
f9e815b3 HW	456
cd8c20b6 AD	457	static void __exit nfnetlink_exit(void)
cd8c20b6 AD	458	{
654d0fbd	459	pr_info("Removing netfilter NETLINK layer.\n");
cd8c20b6 AD	460	unregister_pernet_subsys(&nfnetlink_net_ops);
cd8c20b6 AD	461	}
f9e815b3 HW	462	module_init(nfnetlink_init);
f9e815b3 HW	463	module_exit(nfnetlink_exit);