[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);

#define nfnl_dereference_protected(id) \
	rcu_dereference_protected(table[(id)].subsys, \
				  lockdep_nfnl_is_held((id)))

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,
	[NFNLGRP_NFTABLES]		= NFNL_SUBSYS_NFTABLES,
	[NFNLGRP_ACCT_QUOTA]		= NFNL_SUBSYS_ACCT,
	[NFNLGRP_NFTRACE]		= NFNL_SUBSYS_NFTABLES,
};

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
bool 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, net->nfnl, skb, nlh,
					   (const struct nlattr **)cda);
			rcu_read_unlock();
		} else {
			rcu_read_unlock();
			nfnl_lock(subsys_id);
			if (nfnl_dereference_protected(subsys_id) != ss ||
			    nfnetlink_find_client(type, ss) != nc)
				err = -EAGAIN;
			else if (nc->call)
				err = nc->call(net, 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);
	}
}

enum {
	NFNL_BATCH_FAILURE	= (1 << 0),
	NFNL_BATCH_DONE		= (1 << 1),
	NFNL_BATCH_REPLAY	= (1 << 2),
};

static void nfnetlink_rcv_batch(struct sk_buff *skb, struct nlmsghdr *nlh,
				u_int16_t subsys_id)
{
	struct sk_buff *oskb = skb;
	struct net *net = sock_net(skb->sk);
	const struct nfnetlink_subsystem *ss;
	const struct nfnl_callback *nc;
	static LIST_HEAD(err_list);
	u32 status;
	int err;

	if (subsys_id >= NFNL_SUBSYS_COUNT)
		return netlink_ack(skb, nlh, -EINVAL);
replay:
	status = 0;

	skb = netlink_skb_clone(oskb, GFP_KERNEL);
	if (!skb)
		return netlink_ack(oskb, nlh, -ENOMEM);

	nfnl_lock(subsys_id);
	ss = nfnl_dereference_protected(subsys_id);
	if (!ss) {
#ifdef CONFIG_MODULES
		nfnl_unlock(subsys_id);
		request_module("nfnetlink-subsys-%d", subsys_id);
		nfnl_lock(subsys_id);
		ss = nfnl_dereference_protected(subsys_id);
		if (!ss)
#endif
		{
			nfnl_unlock(subsys_id);
			netlink_ack(oskb, nlh, -EOPNOTSUPP);
			return kfree_skb(skb);
		}
	}

	if (!ss->commit || !ss->abort) {
		nfnl_unlock(subsys_id);
		netlink_ack(oskb, 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 ||
		    skb->len < nlh->nlmsg_len ||
		    nlmsg_len(nlh) < sizeof(struct nfgenmsg)) {
			nfnl_err_reset(&err_list);
			status |= NFNL_BATCH_FAILURE;
			goto done;
		}

		/* 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);
			status |= NFNL_BATCH_FAILURE;
			goto done;
		} else if (type == NFNL_MSG_BATCH_END) {
			status |= NFNL_BATCH_DONE;
			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, 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) {
				status |= NFNL_BATCH_REPLAY;
				goto next;
			}
		}
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(oskb, nlmsg_hdr(oskb), -ENOMEM);
				status |= NFNL_BATCH_FAILURE;
				goto done;
			}
			/* We don't stop processing the batch on errors, thus,
			 * userspace gets all the errors that the batch
			 * triggers.
			 */
			if (err)
				status |= NFNL_BATCH_FAILURE;
		}
next:
		msglen = NLMSG_ALIGN(nlh->nlmsg_len);
		if (msglen > skb->len)
			msglen = skb->len;
		skb_pull(skb, msglen);
	}
done:
	if (status & NFNL_BATCH_REPLAY) {
		ss->abort(net, oskb);
		nfnl_err_reset(&err_list);
		nfnl_unlock(subsys_id);
		kfree_skb(skb);
		goto replay;
	} else if (status == NFNL_BATCH_DONE) {
		ss->commit(net, oskb);
	} else {
		ss->abort(net, oskb);
	}

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

static void nfnetlink_rcv(struct sk_buff *skb)
{
	struct nlmsghdr *nlh = nlmsg_hdr(skb);
	u_int16_t res_id;
	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);
		/* Work around old nft using host byte order */
		if (nfgenmsg->res_id == NFNL_SUBSYS_NFTABLES)
			res_id = NFNL_SUBSYS_NFTABLES;
		else
			res_id = ntohs(nfgenmsg->res_id);
		nfnetlink_rcv_batch(skb, nlh, res_id);
	} else {
		netlink_rcv_skb(skb, &nfnetlink_rcv_msg);
	}
}

#ifdef CONFIG_MODULES
static int nfnetlink_bind(struct net *net, int group)
{
	const struct nfnetlink_subsystem *ss;
	int type;

	if (group <= NFNLGRP_NONE || group > NFNLGRP_MAX)
		return 0;

	type = nfnl_group2type[group];

	rcu_read_lock();
	ss = nfnetlink_get_subsys(type << 8);
	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 = NFNLGRP_NONE + 1; i <= NFNLGRP_MAX; i++)
		BUG_ON(nfnl_group2type[i] == NFNL_SUBSYS_NONE);

	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	35
9c55d3b5 FW	36	#define nfnl_dereference_protected(id) \
	37	rcu_dereference_protected(table[(id)].subsys, \
	38	lockdep_nfnl_is_held((id)))
	39
f9e815b3 HW	40	static char __initdata nfversion[] = "0.30";
f9e815b3 HW	41
c14b78e7 PNA	42	static struct {
	43	struct mutex mutex;
	44	const struct nfnetlink_subsystem __rcu *subsys;
	45	} table[NFNL_SUBSYS_COUNT];
f9e815b3	46
03292745 PNA	47	static const int nfnl_group2type[NFNLGRP_MAX+1] = {
	48	[NFNLGRP_CONNTRACK_NEW] = NFNL_SUBSYS_CTNETLINK,
	49	[NFNLGRP_CONNTRACK_UPDATE] = NFNL_SUBSYS_CTNETLINK,
	50	[NFNLGRP_CONNTRACK_DESTROY] = NFNL_SUBSYS_CTNETLINK,
	51	[NFNLGRP_CONNTRACK_EXP_NEW] = NFNL_SUBSYS_CTNETLINK_EXP,
	52	[NFNLGRP_CONNTRACK_EXP_UPDATE] = NFNL_SUBSYS_CTNETLINK_EXP,
	53	[NFNLGRP_CONNTRACK_EXP_DESTROY] = NFNL_SUBSYS_CTNETLINK_EXP,
97840cb6 PNA	54	[NFNLGRP_NFTABLES] = NFNL_SUBSYS_NFTABLES,
97840cb6 PNA	55	[NFNLGRP_ACCT_QUOTA] = NFNL_SUBSYS_ACCT,
33d5a7b1	56	[NFNLGRP_NFTRACE] = NFNL_SUBSYS_NFTABLES,
03292745 PNA	57	};
03292745 PNA	58
c14b78e7	59	void nfnl_lock(__u8 subsys_id)
f9e815b3	60	{
c14b78e7	61	mutex_lock(&table[subsys_id].mutex);
f9e815b3	62	}
e6a7d3c0	63	EXPORT_SYMBOL_GPL(nfnl_lock);
f9e815b3	64
c14b78e7	65	void nfnl_unlock(__u8 subsys_id)
a3c5029c	66	{
c14b78e7	67	mutex_unlock(&table[subsys_id].mutex);
f9e815b3	68	}
e6a7d3c0	69	EXPORT_SYMBOL_GPL(nfnl_unlock);
f9e815b3	70
0eb5db7a	71	#ifdef CONFIG_PROVE_LOCKING
875e0829	72	bool lockdep_nfnl_is_held(u8 subsys_id)
0eb5db7a PM	73	{
	74	return lockdep_is_held(&table[subsys_id].mutex);
	75	}
	76	EXPORT_SYMBOL_GPL(lockdep_nfnl_is_held);
	77	#endif
	78
7c8d4cb4	79	int nfnetlink_subsys_register(const struct nfnetlink_subsystem *n)
f9e815b3	80	{
c14b78e7 PNA	81	nfnl_lock(n->subsys_id);
	82	if (table[n->subsys_id].subsys) {
	83	nfnl_unlock(n->subsys_id);
0ab43f84 HW	84	return -EBUSY;
0ab43f84 HW	85	}
c14b78e7 PNA	86	rcu_assign_pointer(table[n->subsys_id].subsys, n);
c14b78e7 PNA	87	nfnl_unlock(n->subsys_id);
f9e815b3 HW	88
	89	return 0;
	90	}
f4bc177f	91	EXPORT_SYMBOL_GPL(nfnetlink_subsys_register);
f9e815b3	92
7c8d4cb4	93	int nfnetlink_subsys_unregister(const struct nfnetlink_subsystem *n)
f9e815b3	94	{
c14b78e7 PNA	95	nfnl_lock(n->subsys_id);
	96	table[n->subsys_id].subsys = NULL;
	97	nfnl_unlock(n->subsys_id);
6b75e3e8	98	synchronize_rcu();
f9e815b3 HW	99	return 0;
f9e815b3 HW	100	}
f4bc177f	101	EXPORT_SYMBOL_GPL(nfnetlink_subsys_unregister);
f9e815b3	102
7c8d4cb4	103	static inline const struct nfnetlink_subsystem *nfnetlink_get_subsys(u_int16_t type)
f9e815b3 HW	104	{
	105	u_int8_t subsys_id = NFNL_SUBSYS_ID(type);
	106
ac0f1d98	107	if (subsys_id >= NFNL_SUBSYS_COUNT)
f9e815b3 HW	108	return NULL;
f9e815b3 HW	109
c14b78e7	110	return rcu_dereference(table[subsys_id].subsys);
f9e815b3 HW	111	}
f9e815b3 HW	112
7c8d4cb4 PM	113	static inline const struct nfnl_callback *
7c8d4cb4 PM	114	nfnetlink_find_client(u_int16_t type, const struct nfnetlink_subsystem *ss)
f9e815b3 HW	115	{
f9e815b3 HW	116	u_int8_t cb_id = NFNL_MSG_TYPE(type);
601e68e1	117
67ca3966	118	if (cb_id >= ss->cb_count)
f9e815b3	119	return NULL;
f9e815b3 HW	120
	121	return &ss->cb[cb_id];
	122	}
	123
cd8c20b6	124	int nfnetlink_has_listeners(struct net *net, unsigned int group)
a2427692	125	{
cd8c20b6	126	return netlink_has_listeners(net->nfnl, group);
a2427692 PM	127	}
	128	EXPORT_SYMBOL_GPL(nfnetlink_has_listeners);
	129
3ab1f683 PM	130	struct sk_buff nfnetlink_alloc_skb(struct net net, unsigned int size,
	131	u32 dst_portid, gfp_t gfp_mask)
	132	{
	133	return netlink_alloc_skb(net->nfnl, size, dst_portid, gfp_mask);
	134	}
	135	EXPORT_SYMBOL_GPL(nfnetlink_alloc_skb);
	136
ec464e5d	137	int nfnetlink_send(struct sk_buff skb, struct net net, u32 portid,
95c96174	138	unsigned int group, int echo, gfp_t flags)
f9e815b3	139	{
ec464e5d	140	return nlmsg_notify(net->nfnl, skb, portid, group, echo, flags);
f9e815b3	141	}
f4bc177f	142	EXPORT_SYMBOL_GPL(nfnetlink_send);
f9e815b3	143
ec464e5d	144	int nfnetlink_set_err(struct net *net, u32 portid, u32 group, int error)
dd5b6ce6	145	{
ec464e5d	146	return netlink_set_err(net->nfnl, portid, group, error);
dd5b6ce6 PNA	147	}
	148	EXPORT_SYMBOL_GPL(nfnetlink_set_err);
	149
ec464e5d PM	150	int nfnetlink_unicast(struct sk_buff skb, struct net net, u32 portid,
ec464e5d PM	151	int flags)
f9e815b3	152	{
ec464e5d	153	return netlink_unicast(net->nfnl, skb, portid, flags);
f9e815b3	154	}
f4bc177f	155	EXPORT_SYMBOL_GPL(nfnetlink_unicast);
f9e815b3 HW	156
f9e815b3 HW	157	/* Process one complete nfnetlink message. */
1d00a4eb	158	static int nfnetlink_rcv_msg(struct sk_buff skb, struct nlmsghdr nlh)
f9e815b3	159	{
cd8c20b6	160	struct net *net = sock_net(skb->sk);
7c8d4cb4 PM	161	const struct nfnl_callback *nc;
7c8d4cb4 PM	162	const struct nfnetlink_subsystem *ss;
1d00a4eb	163	int type, err;
f9e815b3	164
f9e815b3	165	/* All the messages must at least contain nfgenmsg */
573ce260	166	if (nlmsg_len(nlh) < sizeof(struct nfgenmsg))
f9e815b3	167	return 0;
f9e815b3 HW	168
f9e815b3 HW	169	type = nlh->nlmsg_type;
e6a7d3c0	170	replay:
6b75e3e8	171	rcu_read_lock();
f9e815b3	172	ss = nfnetlink_get_subsys(type);
0ab43f84	173	if (!ss) {
95a5afca	174	#ifdef CONFIG_MODULES
6b75e3e8	175	rcu_read_unlock();
37d2e7a2	176	request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
6b75e3e8	177	rcu_read_lock();
37d2e7a2	178	ss = nfnetlink_get_subsys(type);
0ab43f84 HW	179	if (!ss)
0ab43f84 HW	180	#endif
6b75e3e8 ED	181	{
6b75e3e8 ED	182	rcu_read_unlock();
1d00a4eb	183	return -EINVAL;
6b75e3e8	184	}
0ab43f84	185	}
f9e815b3 HW	186
f9e815b3 HW	187	nc = nfnetlink_find_client(type, ss);
6b75e3e8 ED	188	if (!nc) {
6b75e3e8 ED	189	rcu_read_unlock();
1d00a4eb	190	return -EINVAL;
6b75e3e8	191	}
f9e815b3	192
f9e815b3	193	{
573ce260	194	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
e3730578	195	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
f49c857f PNA	196	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	197	struct nlattr attr = (void )nlh + min_len;
	198	int attrlen = nlh->nlmsg_len - min_len;
c14b78e7	199	__u8 subsys_id = NFNL_SUBSYS_ID(type);
f49c857f PNA	200
	201	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	202	attr, attrlen, ss->cb[cb_id].policy);
4009e188 TB	203	if (err < 0) {
4009e188 TB	204	rcu_read_unlock();
f49c857f	205	return err;
4009e188	206	}
601e68e1	207
6b75e3e8	208	if (nc->call_rcu) {
7b8002a1	209	err = nc->call_rcu(net, net->nfnl, skb, nlh,
6b75e3e8 ED	210	(const struct nlattr **)cda);
	211	rcu_read_unlock();
	212	} else {
	213	rcu_read_unlock();
c14b78e7	214	nfnl_lock(subsys_id);
9c55d3b5	215	if (nfnl_dereference_protected(subsys_id) != ss \|\|
6b75e3e8 ED	216	nfnetlink_find_client(type, ss) != nc)
6b75e3e8 ED	217	err = -EAGAIN;
59560a38	218	else if (nc->call)
7b8002a1 PNA	219	err = nc->call(net, net->nfnl, skb, nlh,
7b8002a1 PNA	220	(const struct nlattr **)cda);
59560a38 TB	221	else
59560a38 TB	222	err = -EINVAL;
c14b78e7	223	nfnl_unlock(subsys_id);
6b75e3e8	224	}
e6a7d3c0 PNA	225	if (err == -EAGAIN)
	226	goto replay;
	227	return err;
f9e815b3	228	}
f9e815b3 HW	229	}
f9e815b3 HW	230
cbb8125e PNA	231	struct nfnl_err {
	232	struct list_head head;
	233	struct nlmsghdr *nlh;
	234	int err;
	235	};
	236
	237	static int nfnl_err_add(struct list_head list, struct nlmsghdr nlh, int err)
	238	{
	239	struct nfnl_err *nfnl_err;
	240
	241	nfnl_err = kmalloc(sizeof(struct nfnl_err), GFP_KERNEL);
	242	if (nfnl_err == NULL)
	243	return -ENOMEM;
	244
	245	nfnl_err->nlh = nlh;
	246	nfnl_err->err = err;
	247	list_add_tail(&nfnl_err->head, list);
	248
	249	return 0;
	250	}
	251
	252	static void nfnl_err_del(struct nfnl_err *nfnl_err)
	253	{
	254	list_del(&nfnl_err->head);
	255	kfree(nfnl_err);
	256	}
	257
	258	static void nfnl_err_reset(struct list_head *err_list)
	259	{
	260	struct nfnl_err nfnl_err, next;
	261
	262	list_for_each_entry_safe(nfnl_err, next, err_list, head)
	263	nfnl_err_del(nfnl_err);
	264	}
	265
	266	static void nfnl_err_deliver(struct list_head err_list, struct sk_buff skb)
	267	{
	268	struct nfnl_err nfnl_err, next;
	269
	270	list_for_each_entry_safe(nfnl_err, next, err_list, head) {
	271	netlink_ack(skb, nfnl_err->nlh, nfnl_err->err);
	272	nfnl_err_del(nfnl_err);
	273	}
	274	}
	275
6742b9e3 PNA	276	enum {
	277	NFNL_BATCH_FAILURE = (1 << 0),
	278	NFNL_BATCH_DONE = (1 << 1),
	279	NFNL_BATCH_REPLAY = (1 << 2),
	280	};
	281
0628b123 PNA	282	static void nfnetlink_rcv_batch(struct sk_buff skb, struct nlmsghdr nlh,
	283	u_int16_t subsys_id)
	284	{
0f816232	285	struct sk_buff *oskb = skb;
0628b123 PNA	286	struct net *net = sock_net(skb->sk);
	287	const struct nfnetlink_subsystem *ss;
	288	const struct nfnl_callback *nc;
cbb8125e	289	static LIST_HEAD(err_list);
6742b9e3	290	u32 status;
0628b123 PNA	291	int err;
	292
	293	if (subsys_id >= NFNL_SUBSYS_COUNT)
	294	return netlink_ack(skb, nlh, -EINVAL);
	295	replay:
6742b9e3 PNA	296	status = 0;
6742b9e3 PNA	297
0f816232 DJ	298	skb = netlink_skb_clone(oskb, GFP_KERNEL);
0f816232 DJ	299	if (!skb)
0628b123 PNA	300	return netlink_ack(oskb, nlh, -ENOMEM);
0628b123 PNA	301
0628b123	302	nfnl_lock(subsys_id);
9c55d3b5	303	ss = nfnl_dereference_protected(subsys_id);
0628b123 PNA	304	if (!ss) {
	305	#ifdef CONFIG_MODULES
	306	nfnl_unlock(subsys_id);
	307	request_module("nfnetlink-subsys-%d", subsys_id);
	308	nfnl_lock(subsys_id);
9c55d3b5	309	ss = nfnl_dereference_protected(subsys_id);
0628b123 PNA	310	if (!ss)
	311	#endif
	312	{
	313	nfnl_unlock(subsys_id);
7c7bdf35	314	netlink_ack(oskb, nlh, -EOPNOTSUPP);
0f816232	315	return kfree_skb(skb);
0628b123 PNA	316	}
	317	}
	318
	319	if (!ss->commit \|\| !ss->abort) {
	320	nfnl_unlock(subsys_id);
7c7bdf35	321	netlink_ack(oskb, nlh, -EOPNOTSUPP);
ecd15dd7	322	return kfree_skb(skb);
0628b123 PNA	323	}
	324
	325	while (skb->len >= nlmsg_total_size(0)) {
	326	int msglen, type;
	327
	328	nlh = nlmsg_hdr(skb);
	329	err = 0;
	330
c58d6c93 PT	331	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	332	skb->len < nlh->nlmsg_len \|\|
	333	nlmsg_len(nlh) < sizeof(struct nfgenmsg)) {
	334	nfnl_err_reset(&err_list);
	335	status \|= NFNL_BATCH_FAILURE;
	336	goto done;
0628b123 PNA	337	}
	338
	339	/* Only requests are handled by the kernel */
	340	if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) {
	341	err = -EINVAL;
	342	goto ack;
	343	}
	344
	345	type = nlh->nlmsg_type;
	346	if (type == NFNL_MSG_BATCH_BEGIN) {
	347	/* Malformed: Batch begin twice */
cbb8125e	348	nfnl_err_reset(&err_list);
6742b9e3	349	status \|= NFNL_BATCH_FAILURE;
0628b123 PNA	350	goto done;
0628b123 PNA	351	} else if (type == NFNL_MSG_BATCH_END) {
6742b9e3	352	status \|= NFNL_BATCH_DONE;
0628b123 PNA	353	goto done;
	354	} else if (type < NLMSG_MIN_TYPE) {
	355	err = -EINVAL;
	356	goto ack;
	357	}
	358
	359	/* We only accept a batch with messages for the same
	360	* subsystem.
	361	*/
	362	if (NFNL_SUBSYS_ID(type) != subsys_id) {
	363	err = -EINVAL;
	364	goto ack;
	365	}
	366
	367	nc = nfnetlink_find_client(type, ss);
	368	if (!nc) {
	369	err = -EINVAL;
	370	goto ack;
	371	}
	372
	373	{
	374	int min_len = nlmsg_total_size(sizeof(struct nfgenmsg));
	375	u_int8_t cb_id = NFNL_MSG_TYPE(nlh->nlmsg_type);
	376	struct nlattr *cda[ss->cb[cb_id].attr_count + 1];
	377	struct nlattr attr = (void )nlh + min_len;
	378	int attrlen = nlh->nlmsg_len - min_len;
	379
	380	err = nla_parse(cda, ss->cb[cb_id].attr_count,
	381	attr, attrlen, ss->cb[cb_id].policy);
	382	if (err < 0)
	383	goto ack;
	384
	385	if (nc->call_batch) {
633c9a84	386	err = nc->call_batch(net, net->nfnl, skb, nlh,
0628b123 PNA	387	(const struct nlattr **)cda);
	388	}
	389
	390	/* The lock was released to autoload some module, we
	391	* have to abort and start from scratch using the
	392	* original skb.
	393	*/
	394	if (err == -EAGAIN) {
6742b9e3 PNA	395	status \|= NFNL_BATCH_REPLAY;
6742b9e3 PNA	396	goto next;
0628b123 PNA	397	}
	398	}
	399	ack:
	400	if (nlh->nlmsg_flags & NLM_F_ACK \|\| err) {
cbb8125e PNA	401	/* Errors are delivered once the full batch has been
	402	* processed, this avoids that the same error is
	403	* reported several times when replaying the batch.
	404	*/
	405	if (nfnl_err_add(&err_list, nlh, err) < 0) {
	406	/* We failed to enqueue an error, reset the
	407	* list of errors and send OOM to userspace
	408	* pointing to the batch header.
	409	*/
	410	nfnl_err_reset(&err_list);
7c7bdf35	411	netlink_ack(oskb, nlmsg_hdr(oskb), -ENOMEM);
6742b9e3	412	status \|= NFNL_BATCH_FAILURE;
cbb8125e PNA	413	goto done;
cbb8125e PNA	414	}
0628b123 PNA	415	/* We don't stop processing the batch on errors, thus,
	416	* userspace gets all the errors that the batch
	417	* triggers.
	418	*/
0628b123	419	if (err)
6742b9e3	420	status \|= NFNL_BATCH_FAILURE;
0628b123	421	}
6742b9e3	422	next:
0628b123 PNA	423	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	424	if (msglen > skb->len)
	425	msglen = skb->len;
	426	skb_pull(skb, msglen);
	427	}
	428	done:
6742b9e3	429	if (status & NFNL_BATCH_REPLAY) {
5913beaf	430	ss->abort(net, oskb);
6742b9e3 PNA	431	nfnl_err_reset(&err_list);
	432	nfnl_unlock(subsys_id);
	433	kfree_skb(skb);
	434	goto replay;
	435	} else if (status == NFNL_BATCH_DONE) {
5913beaf	436	ss->commit(net, oskb);
6742b9e3	437	} else {
5913beaf	438	ss->abort(net, oskb);
6742b9e3	439	}
0628b123	440
cbb8125e	441	nfnl_err_deliver(&err_list, oskb);
0628b123	442	nfnl_unlock(subsys_id);
0f816232	443	kfree_skb(skb);
0628b123 PNA	444	}
0628b123 PNA	445
cd40b7d3	446	static void nfnetlink_rcv(struct sk_buff *skb)
f9e815b3	447	{
0628b123	448	struct nlmsghdr *nlh = nlmsg_hdr(skb);
a9de9777	449	u_int16_t res_id;
0628b123 PNA	450	int msglen;
0628b123 PNA	451
0628b123 PNA	452	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	453	skb->len < nlh->nlmsg_len)
	454	return;
	455
90f62cf3	456	if (!netlink_net_capable(skb, CAP_NET_ADMIN)) {
cdbe7c2d JB	457	netlink_ack(skb, nlh, -EPERM);
	458	return;
	459	}
	460
0628b123 PNA	461	if (nlh->nlmsg_type == NFNL_MSG_BATCH_BEGIN) {
	462	struct nfgenmsg *nfgenmsg;
	463
	464	msglen = NLMSG_ALIGN(nlh->nlmsg_len);
	465	if (msglen > skb->len)
	466	msglen = skb->len;
	467
	468	if (nlh->nlmsg_len < NLMSG_HDRLEN \|\|
	469	skb->len < NLMSG_HDRLEN + sizeof(struct nfgenmsg))
	470	return;
	471
	472	nfgenmsg = nlmsg_data(nlh);
	473	skb_pull(skb, msglen);
a9de9777 PNA	474	/* Work around old nft using host byte order */
	475	if (nfgenmsg->res_id == NFNL_SUBSYS_NFTABLES)
	476	res_id = NFNL_SUBSYS_NFTABLES;
	477	else
	478	res_id = ntohs(nfgenmsg->res_id);
	479	nfnetlink_rcv_batch(skb, nlh, res_id);
0628b123 PNA	480	} else {
	481	netlink_rcv_skb(skb, &nfnetlink_rcv_msg);
	482	}
f9e815b3 HW	483	}
f9e815b3 HW	484
03292745	485	#ifdef CONFIG_MODULES
023e2cfa	486	static int nfnetlink_bind(struct net *net, int group)
03292745 PNA	487	{
03292745 PNA	488	const struct nfnetlink_subsystem *ss;
97840cb6 PNA	489	int type;
	490
	491	if (group <= NFNLGRP_NONE \|\| group > NFNLGRP_MAX)
62924af2	492	return 0;
97840cb6 PNA	493
97840cb6 PNA	494	type = nfnl_group2type[group];
03292745 PNA	495
03292745 PNA	496	rcu_read_lock();
dbc3617f	497	ss = nfnetlink_get_subsys(type << 8);
03292745	498	rcu_read_unlock();
bfe4bc71 RGB	499	if (!ss)
bfe4bc71 RGB	500	request_module("nfnetlink-subsys-%d", type);
4f520900	501	return 0;
03292745 PNA	502	}
	503	#endif
	504
cd8c20b6	505	static int __net_init nfnetlink_net_init(struct net *net)
f9e815b3	506	{
cd8c20b6	507	struct sock *nfnl;
a31f2d17 PNA	508	struct netlink_kernel_cfg cfg = {
	509	.groups = NFNLGRP_MAX,
	510	.input = nfnetlink_rcv,
03292745 PNA	511	#ifdef CONFIG_MODULES
	512	.bind = nfnetlink_bind,
	513	#endif
a31f2d17	514	};
cd8c20b6	515
9f00d977	516	nfnl = netlink_kernel_create(net, NETLINK_NETFILTER, &cfg);
cd8c20b6 AD	517	if (!nfnl)
	518	return -ENOMEM;
	519	net->nfnl_stash = nfnl;
cf778b00	520	rcu_assign_pointer(net->nfnl, nfnl);
cd8c20b6	521	return 0;
f9e815b3 HW	522	}
f9e815b3 HW	523
cd8c20b6	524	static void __net_exit nfnetlink_net_exit_batch(struct list_head *net_exit_list)
f9e815b3	525	{
cd8c20b6	526	struct net *net;
f9e815b3	527
cd8c20b6	528	list_for_each_entry(net, net_exit_list, exit_list)
a9b3cd7f	529	RCU_INIT_POINTER(net->nfnl, NULL);
cd8c20b6 AD	530	synchronize_net();
	531	list_for_each_entry(net, net_exit_list, exit_list)
	532	netlink_kernel_release(net->nfnl_stash);
	533	}
f9e815b3	534
cd8c20b6 AD	535	static struct pernet_operations nfnetlink_net_ops = {
	536	.init = nfnetlink_net_init,
	537	.exit_batch = nfnetlink_net_exit_batch,
	538	};
	539
	540	static int __init nfnetlink_init(void)
	541	{
c14b78e7 PNA	542	int i;
c14b78e7 PNA	543
97840cb6 PNA	544	for (i = NFNLGRP_NONE + 1; i <= NFNLGRP_MAX; i++)
	545	BUG_ON(nfnl_group2type[i] == NFNL_SUBSYS_NONE);
	546
c14b78e7 PNA	547	for (i=0; i<NFNL_SUBSYS_COUNT; i++)
	548	mutex_init(&table[i].mutex);
	549
654d0fbd	550	pr_info("Netfilter messages via NETLINK v%s.\n", nfversion);
cd8c20b6	551	return register_pernet_subsys(&nfnetlink_net_ops);
f9e815b3 HW	552	}
f9e815b3 HW	553
cd8c20b6 AD	554	static void __exit nfnetlink_exit(void)
cd8c20b6 AD	555	{
654d0fbd	556	pr_info("Removing netfilter NETLINK layer.\n");
cd8c20b6 AD	557	unregister_pernet_subsys(&nfnetlink_net_ops);
cd8c20b6 AD	558	}
f9e815b3 HW	559	module_init(nfnetlink_init);
f9e815b3 HW	560	module_exit(nfnetlink_exit);