[mirror_ubuntu-bionic-kernel.git] / net / netfilter / core.c

/* netfilter.c: look after the filters for various protocols.
 * Heavily influenced by the old firewall.c by David Bonn and Alan Cox.
 *
 * Thanks to Rob `CmdrTaco' Malda for not influencing this code in any
 * way.
 *
 * Rusty Russell (C)2000 -- This code is GPL.
 * Patrick McHardy (c) 2006-2012
 */
#include <linux/kernel.h>
#include <linux/netfilter.h>
#include <net/protocol.h>
#include <linux/init.h>
#include <linux/skbuff.h>
#include <linux/wait.h>
#include <linux/module.h>
#include <linux/interrupt.h>
#include <linux/if.h>
#include <linux/netdevice.h>
#include <linux/netfilter_ipv6.h>
#include <linux/inetdevice.h>
#include <linux/proc_fs.h>
#include <linux/mutex.h>
#include <linux/slab.h>
#include <linux/rcupdate.h>
#include <net/net_namespace.h>
#include <net/sock.h>

#include "nf_internals.h"

static DEFINE_MUTEX(afinfo_mutex);

const struct nf_afinfo __rcu *nf_afinfo[NFPROTO_NUMPROTO] __read_mostly;
EXPORT_SYMBOL(nf_afinfo);
const struct nf_ipv6_ops __rcu *nf_ipv6_ops __read_mostly;
EXPORT_SYMBOL_GPL(nf_ipv6_ops);

DEFINE_PER_CPU(bool, nf_skb_duplicated);
EXPORT_SYMBOL_GPL(nf_skb_duplicated);

int nf_register_afinfo(const struct nf_afinfo *afinfo)
{
	mutex_lock(&afinfo_mutex);
	RCU_INIT_POINTER(nf_afinfo[afinfo->family], afinfo);
	mutex_unlock(&afinfo_mutex);
	return 0;
}
EXPORT_SYMBOL_GPL(nf_register_afinfo);

void nf_unregister_afinfo(const struct nf_afinfo *afinfo)
{
	mutex_lock(&afinfo_mutex);
	RCU_INIT_POINTER(nf_afinfo[afinfo->family], NULL);
	mutex_unlock(&afinfo_mutex);
	synchronize_rcu();
}
EXPORT_SYMBOL_GPL(nf_unregister_afinfo);

#ifdef HAVE_JUMP_LABEL
struct static_key nf_hooks_needed[NFPROTO_NUMPROTO][NF_MAX_HOOKS];
EXPORT_SYMBOL(nf_hooks_needed);
#endif

static DEFINE_MUTEX(nf_hook_mutex);
#define nf_entry_dereference(e) \
	rcu_dereference_protected(e, lockdep_is_held(&nf_hook_mutex))

static struct nf_hook_entry __rcu **nf_hook_entry_head(struct net *net, const struct nf_hook_ops *reg)
{
	if (reg->pf != NFPROTO_NETDEV)
		return net->nf.hooks[reg->pf]+reg->hooknum;

#ifdef CONFIG_NETFILTER_INGRESS
	if (reg->hooknum == NF_NETDEV_INGRESS) {
		if (reg->dev && dev_net(reg->dev) == net)
			return &reg->dev->nf_hooks_ingress;
	}
#endif
	return NULL;
}

int nf_register_net_hook(struct net *net, const struct nf_hook_ops *reg)
{
	struct nf_hook_entry __rcu **pp;
	struct nf_hook_entry *entry, *p;

	if (reg->pf == NFPROTO_NETDEV) {
#ifndef CONFIG_NETFILTER_INGRESS
		if (reg->hooknum == NF_NETDEV_INGRESS)
			return -EOPNOTSUPP;
#endif
		if (reg->hooknum != NF_NETDEV_INGRESS ||
		    !reg->dev || dev_net(reg->dev) != net)
			return -EINVAL;
	}

	pp = nf_hook_entry_head(net, reg);
	if (!pp)
		return -EINVAL;

	entry = kmalloc(sizeof(*entry), GFP_KERNEL);
	if (!entry)
		return -ENOMEM;

	entry->orig_ops	= reg;
	entry->ops	= *reg;
	entry->next	= NULL;

	mutex_lock(&nf_hook_mutex);

	/* Find the spot in the list */
	while ((p = nf_entry_dereference(*pp)) != NULL) {
		if (reg->priority < p->orig_ops->priority)
			break;
		pp = &p->next;
	}
	rcu_assign_pointer(entry->next, p);
	rcu_assign_pointer(*pp, entry);

	mutex_unlock(&nf_hook_mutex);
#ifdef CONFIG_NETFILTER_INGRESS
	if (reg->pf == NFPROTO_NETDEV && reg->hooknum == NF_NETDEV_INGRESS)
		net_inc_ingress_queue();
#endif
#ifdef HAVE_JUMP_LABEL
	static_key_slow_inc(&nf_hooks_needed[reg->pf][reg->hooknum]);
#endif
	return 0;
}
EXPORT_SYMBOL(nf_register_net_hook);

void nf_unregister_net_hook(struct net *net, const struct nf_hook_ops *reg)
{
	struct nf_hook_entry __rcu **pp;
	struct nf_hook_entry *p;

	pp = nf_hook_entry_head(net, reg);
	if (WARN_ON_ONCE(!pp))
		return;

	mutex_lock(&nf_hook_mutex);
	while ((p = nf_entry_dereference(*pp)) != NULL) {
		if (p->orig_ops == reg) {
			rcu_assign_pointer(*pp, p->next);
			break;
		}
		pp = &p->next;
	}
	mutex_unlock(&nf_hook_mutex);
	if (!p) {
		WARN(1, "nf_unregister_net_hook: hook not found!\n");
		return;
	}
#ifdef CONFIG_NETFILTER_INGRESS
	if (reg->pf == NFPROTO_NETDEV && reg->hooknum == NF_NETDEV_INGRESS)
		net_dec_ingress_queue();
#endif
#ifdef HAVE_JUMP_LABEL
	static_key_slow_dec(&nf_hooks_needed[reg->pf][reg->hooknum]);
#endif
	synchronize_net();
	nf_queue_nf_hook_drop(net, p);
	/* other cpu might still process nfqueue verdict that used reg */
	synchronize_net();
	kfree(p);
}
EXPORT_SYMBOL(nf_unregister_net_hook);

int nf_register_net_hooks(struct net *net, const struct nf_hook_ops *reg,
			  unsigned int n)
{
	unsigned int i;
	int err = 0;

	for (i = 0; i < n; i++) {
		err = nf_register_net_hook(net, &reg[i]);
		if (err)
			goto err;
	}
	return err;

err:
	if (i > 0)
		nf_unregister_net_hooks(net, reg, i);
	return err;
}
EXPORT_SYMBOL(nf_register_net_hooks);

void nf_unregister_net_hooks(struct net *net, const struct nf_hook_ops *reg,
			     unsigned int n)
{
	while (n-- > 0)
		nf_unregister_net_hook(net, &reg[n]);
}
EXPORT_SYMBOL(nf_unregister_net_hooks);

static LIST_HEAD(nf_hook_list);

static int _nf_register_hook(struct nf_hook_ops *reg)
{
	struct net *net, *last;
	int ret;

	for_each_net(net) {
		ret = nf_register_net_hook(net, reg);
		if (ret && ret != -ENOENT)
			goto rollback;
	}
	list_add_tail(&reg->list, &nf_hook_list);

	return 0;
rollback:
	last = net;
	for_each_net(net) {
		if (net == last)
			break;
		nf_unregister_net_hook(net, reg);
	}
	return ret;
}

int nf_register_hook(struct nf_hook_ops *reg)
{
	int ret;

	rtnl_lock();
	ret = _nf_register_hook(reg);
	rtnl_unlock();

	return ret;
}
EXPORT_SYMBOL(nf_register_hook);

static void _nf_unregister_hook(struct nf_hook_ops *reg)
{
	struct net *net;

	list_del(&reg->list);
	for_each_net(net)
		nf_unregister_net_hook(net, reg);
}

void nf_unregister_hook(struct nf_hook_ops *reg)
{
	rtnl_lock();
	_nf_unregister_hook(reg);
	rtnl_unlock();
}
EXPORT_SYMBOL(nf_unregister_hook);

int nf_register_hooks(struct nf_hook_ops *reg, unsigned int n)
{
	unsigned int i;
	int err = 0;

	for (i = 0; i < n; i++) {
		err = nf_register_hook(&reg[i]);
		if (err)
			goto err;
	}
	return err;

err:
	if (i > 0)
		nf_unregister_hooks(reg, i);
	return err;
}
EXPORT_SYMBOL(nf_register_hooks);

/* Caller MUST take rtnl_lock() */
int _nf_register_hooks(struct nf_hook_ops *reg, unsigned int n)
{
	unsigned int i;
	int err = 0;

	for (i = 0; i < n; i++) {
		err = _nf_register_hook(&reg[i]);
		if (err)
			goto err;
	}
	return err;

err:
	if (i > 0)
		_nf_unregister_hooks(reg, i);
	return err;
}
EXPORT_SYMBOL(_nf_register_hooks);

void nf_unregister_hooks(struct nf_hook_ops *reg, unsigned int n)
{
	while (n-- > 0)
		nf_unregister_hook(&reg[n]);
}
EXPORT_SYMBOL(nf_unregister_hooks);

/* Caller MUST take rtnl_lock */
void _nf_unregister_hooks(struct nf_hook_ops *reg, unsigned int n)
{
	while (n-- > 0)
		_nf_unregister_hook(&reg[n]);
}
EXPORT_SYMBOL(_nf_unregister_hooks);

unsigned int nf_iterate(struct sk_buff *skb,
			struct nf_hook_state *state,
			struct nf_hook_entry **entryp)
{
	unsigned int verdict;

	/*
	 * The caller must not block between calls to this
	 * function because of risk of continuing from deleted element.
	 */
	while (*entryp) {
		if (state->thresh > (*entryp)->ops.priority) {
			*entryp = rcu_dereference((*entryp)->next);
			continue;
		}

		/* Optimization: we don't need to hold module
		   reference here, since function can't sleep. --RR */
repeat:
		verdict = (*entryp)->ops.hook((*entryp)->ops.priv, skb, state);
		if (verdict != NF_ACCEPT) {
#ifdef CONFIG_NETFILTER_DEBUG
			if (unlikely((verdict & NF_VERDICT_MASK)
							> NF_MAX_VERDICT)) {
				NFDEBUG("Evil return from %p(%u).\n",
					(*entryp)->ops.hook, state->hook);
				*entryp = rcu_dereference((*entryp)->next);
				continue;
			}
#endif
			if (verdict != NF_REPEAT)
				return verdict;
			goto repeat;
		}
		*entryp = rcu_dereference((*entryp)->next);
	}
	return NF_ACCEPT;
}


/* Returns 1 if okfn() needs to be executed by the caller,
 * -EPERM for NF_DROP, 0 otherwise.  Caller must hold rcu_read_lock. */
int nf_hook_slow(struct sk_buff *skb, struct nf_hook_state *state)
{
	struct nf_hook_entry *entry;
	unsigned int verdict;
	int ret = 0;

	entry = rcu_dereference(state->hook_entries);
next_hook:
	verdict = nf_iterate(skb, state, &entry);
	if (verdict == NF_ACCEPT || verdict == NF_STOP) {
		ret = 1;
	} else if ((verdict & NF_VERDICT_MASK) == NF_DROP) {
		kfree_skb(skb);
		ret = NF_DROP_GETERR(verdict);
		if (ret == 0)
			ret = -EPERM;
	} else if ((verdict & NF_VERDICT_MASK) == NF_QUEUE) {
		int err;

		RCU_INIT_POINTER(state->hook_entries, entry);
		err = nf_queue(skb, state, verdict >> NF_VERDICT_QBITS);
		if (err < 0) {
			if (err == -ESRCH &&
			   (verdict & NF_VERDICT_FLAG_QUEUE_BYPASS))
				goto next_hook;
			kfree_skb(skb);
		}
	}
	return ret;
}
EXPORT_SYMBOL(nf_hook_slow);


int skb_make_writable(struct sk_buff *skb, unsigned int writable_len)
{
	if (writable_len > skb->len)
		return 0;

	/* Not exclusive use of packet?  Must copy. */
	if (!skb_cloned(skb)) {
		if (writable_len <= skb_headlen(skb))
			return 1;
	} else if (skb_clone_writable(skb, writable_len))
		return 1;

	if (writable_len <= skb_headlen(skb))
		writable_len = 0;
	else
		writable_len -= skb_headlen(skb);

	return !!__pskb_pull_tail(skb, writable_len);
}
EXPORT_SYMBOL(skb_make_writable);

/* This needs to be compiled in any case to avoid dependencies between the
 * nfnetlink_queue code and nf_conntrack.
 */
struct nfnl_ct_hook __rcu *nfnl_ct_hook __read_mostly;
EXPORT_SYMBOL_GPL(nfnl_ct_hook);

#if IS_ENABLED(CONFIG_NF_CONNTRACK)
/* This does not belong here, but locally generated errors need it if connection
   tracking in use: without this, connection may not be in hash table, and hence
   manufactured ICMP or RST packets will not be associated with it. */
void (*ip_ct_attach)(struct sk_buff *, const struct sk_buff *)
		__rcu __read_mostly;
EXPORT_SYMBOL(ip_ct_attach);

void nf_ct_attach(struct sk_buff *new, const struct sk_buff *skb)
{
	void (*attach)(struct sk_buff *, const struct sk_buff *);

	if (skb->nfct) {
		rcu_read_lock();
		attach = rcu_dereference(ip_ct_attach);
		if (attach)
			attach(new, skb);
		rcu_read_unlock();
	}
}
EXPORT_SYMBOL(nf_ct_attach);

void (*nf_ct_destroy)(struct nf_conntrack *) __rcu __read_mostly;
EXPORT_SYMBOL(nf_ct_destroy);

void nf_conntrack_destroy(struct nf_conntrack *nfct)
{
	void (*destroy)(struct nf_conntrack *);

	rcu_read_lock();
	destroy = rcu_dereference(nf_ct_destroy);
	BUG_ON(destroy == NULL);
	destroy(nfct);
	rcu_read_unlock();
}
EXPORT_SYMBOL(nf_conntrack_destroy);

/* Built-in default zone used e.g. by modules. */
const struct nf_conntrack_zone nf_ct_zone_dflt = {
	.id	= NF_CT_DEFAULT_ZONE_ID,
	.dir	= NF_CT_DEFAULT_ZONE_DIR,
};
EXPORT_SYMBOL_GPL(nf_ct_zone_dflt);
#endif /* CONFIG_NF_CONNTRACK */

#ifdef CONFIG_NF_NAT_NEEDED
void (*nf_nat_decode_session_hook)(struct sk_buff *, struct flowi *);
EXPORT_SYMBOL(nf_nat_decode_session_hook);
#endif

static int nf_register_hook_list(struct net *net)
{
	struct nf_hook_ops *elem;
	int ret;

	rtnl_lock();
	list_for_each_entry(elem, &nf_hook_list, list) {
		ret = nf_register_net_hook(net, elem);
		if (ret && ret != -ENOENT)
			goto out_undo;
	}
	rtnl_unlock();
	return 0;

out_undo:
	list_for_each_entry_continue_reverse(elem, &nf_hook_list, list)
		nf_unregister_net_hook(net, elem);
	rtnl_unlock();
	return ret;
}

static void nf_unregister_hook_list(struct net *net)
{
	struct nf_hook_ops *elem;

	rtnl_lock();
	list_for_each_entry(elem, &nf_hook_list, list)
		nf_unregister_net_hook(net, elem);
	rtnl_unlock();
}

static int __net_init netfilter_net_init(struct net *net)
{
	int i, h, ret;

	for (i = 0; i < ARRAY_SIZE(net->nf.hooks); i++) {
		for (h = 0; h < NF_MAX_HOOKS; h++)
			RCU_INIT_POINTER(net->nf.hooks[i][h], NULL);
	}

#ifdef CONFIG_PROC_FS
	net->nf.proc_netfilter = proc_net_mkdir(net, "netfilter",
						net->proc_net);
	if (!net->nf.proc_netfilter) {
		if (!net_eq(net, &init_net))
			pr_err("cannot create netfilter proc entry");

		return -ENOMEM;
	}
#endif
	ret = nf_register_hook_list(net);
	if (ret)
		remove_proc_entry("netfilter", net->proc_net);

	return ret;
}

static void __net_exit netfilter_net_exit(struct net *net)
{
	nf_unregister_hook_list(net);
	remove_proc_entry("netfilter", net->proc_net);
}

static struct pernet_operations netfilter_net_ops = {
	.init = netfilter_net_init,
	.exit = netfilter_net_exit,
};

int __init netfilter_init(void)
{
	int ret;

	ret = register_pernet_subsys(&netfilter_net_ops);
	if (ret < 0)
		goto err;

	ret = netfilter_log_init();
	if (ret < 0)
		goto err_pernet;

	return 0;
err_pernet:
	unregister_pernet_subsys(&netfilter_net_ops);
err:
	return ret;
}
Commit	Line	Data
601e68e1	1	/* netfilter.c: look after the filters for various protocols.
f6ebe77f HW	2	* Heavily influenced by the old firewall.c by David Bonn and Alan Cox.
	3	*
	4	* Thanks to Rob `CmdrTaco' Malda for not influencing this code in any
	5	* way.
	6	*
	7	* Rusty Russell (C)2000 -- This code is GPL.
f229f6ce	8	* Patrick McHardy (c) 2006-2012
f6ebe77f	9	*/
f6ebe77f HW	10	#include <linux/kernel.h>
	11	#include <linux/netfilter.h>
	12	#include <net/protocol.h>
	13	#include <linux/init.h>
	14	#include <linux/skbuff.h>
	15	#include <linux/wait.h>
	16	#include <linux/module.h>
	17	#include <linux/interrupt.h>
	18	#include <linux/if.h>
	19	#include <linux/netdevice.h>
56768644	20	#include <linux/netfilter_ipv6.h>
f6ebe77f HW	21	#include <linux/inetdevice.h>
f6ebe77f HW	22	#include <linux/proc_fs.h>
d486dd1f	23	#include <linux/mutex.h>
5a0e3ad6	24	#include <linux/slab.h>
e3b37f11	25	#include <linux/rcupdate.h>
457c4cbc	26	#include <net/net_namespace.h>
f6ebe77f HW	27	#include <net/sock.h>
	28
	29	#include "nf_internals.h"
	30
d486dd1f	31	static DEFINE_MUTEX(afinfo_mutex);
bce8032e	32
0906a372	33	const struct nf_afinfo __rcu *nf_afinfo[NFPROTO_NUMPROTO] __read_mostly;
bce8032e	34	EXPORT_SYMBOL(nf_afinfo);
2a7851bf FW	35	const struct nf_ipv6_ops __rcu *nf_ipv6_ops __read_mostly;
2a7851bf FW	36	EXPORT_SYMBOL_GPL(nf_ipv6_ops);
bce8032e	37
e7c8899f FW	38	DEFINE_PER_CPU(bool, nf_skb_duplicated);
	39	EXPORT_SYMBOL_GPL(nf_skb_duplicated);
	40
1e796fda	41	int nf_register_afinfo(const struct nf_afinfo *afinfo)
bce8032e	42	{
7926dbfa	43	mutex_lock(&afinfo_mutex);
a9b3cd7f	44	RCU_INIT_POINTER(nf_afinfo[afinfo->family], afinfo);
d486dd1f	45	mutex_unlock(&afinfo_mutex);
bce8032e PM	46	return 0;
	47	}
	48	EXPORT_SYMBOL_GPL(nf_register_afinfo);
	49
1e796fda	50	void nf_unregister_afinfo(const struct nf_afinfo *afinfo)
bce8032e	51	{
d486dd1f	52	mutex_lock(&afinfo_mutex);
a9b3cd7f	53	RCU_INIT_POINTER(nf_afinfo[afinfo->family], NULL);
d486dd1f	54	mutex_unlock(&afinfo_mutex);
bce8032e PM	55	synchronize_rcu();
	56	}
	57	EXPORT_SYMBOL_GPL(nf_unregister_afinfo);
	58
d1c85c2e	59	#ifdef HAVE_JUMP_LABEL
c5905afb	60	struct static_key nf_hooks_needed[NFPROTO_NUMPROTO][NF_MAX_HOOKS];
a2d7ec58 ED	61	EXPORT_SYMBOL(nf_hooks_needed);
	62	#endif
	63
fd706d69	64	static DEFINE_MUTEX(nf_hook_mutex);
e3b37f11 AC	65	#define nf_entry_dereference(e) \
e3b37f11 AC	66	rcu_dereference_protected(e, lockdep_is_held(&nf_hook_mutex))
f6ebe77f	67
bd3769bf	68	static struct nf_hook_entry __rcu *nf_hook_entry_head(struct net net, const struct nf_hook_ops *reg)
f6ebe77f	69	{
0edcf282	70	if (reg->pf != NFPROTO_NETDEV)
bd3769bf LT	71	return net->nf.hooks[reg->pf]+reg->hooknum;
bd3769bf LT	72
e687ad60	73	#ifdef CONFIG_NETFILTER_INGRESS
bd3769bf	74	if (reg->hooknum == NF_NETDEV_INGRESS) {
085db2c0	75	if (reg->dev && dev_net(reg->dev) == net)
bd3769bf	76	return &reg->dev->nf_hooks_ingress;
e687ad60	77	}
7816ec56	78	#endif
bd3769bf	79	return NULL;
e3b37f11	80	}
7181ebaf	81
085db2c0	82	int nf_register_net_hook(struct net net, const struct nf_hook_ops reg)
0edcf282	83	{
bd3769bf LT	84	struct nf_hook_entry __rcu **pp;
bd3769bf LT	85	struct nf_hook_entry entry, p;
085db2c0	86
7816ec56 AC	87	if (reg->pf == NFPROTO_NETDEV) {
	88	#ifndef CONFIG_NETFILTER_INGRESS
	89	if (reg->hooknum == NF_NETDEV_INGRESS)
	90	return -EOPNOTSUPP;
	91	#endif
	92	if (reg->hooknum != NF_NETDEV_INGRESS \|\|
	93	!reg->dev \|\| dev_net(reg->dev) != net)
	94	return -EINVAL;
	95	}
d4bb5caa	96
bd3769bf LT	97	pp = nf_hook_entry_head(net, reg);
	98	if (!pp)
	99	return -EINVAL;
	100
7181ebaf PNA	101	entry = kmalloc(sizeof(*entry), GFP_KERNEL);
7181ebaf PNA	102	if (!entry)
085db2c0	103	return -ENOMEM;
0edcf282	104
7181ebaf PNA	105	entry->orig_ops = reg;
7181ebaf PNA	106	entry->ops = *reg;
e3b37f11 AC	107	entry->next = NULL;
	108
	109	mutex_lock(&nf_hook_mutex);
085db2c0	110
bd3769bf LT	111	/* Find the spot in the list */
	112	while ((p = nf_entry_dereference(*pp)) != NULL) {
	113	if (reg->priority < p->orig_ops->priority)
	114	break;
	115	pp = &p->next;
f6ebe77f	116	}
bd3769bf LT	117	rcu_assign_pointer(entry->next, p);
bd3769bf LT	118	rcu_assign_pointer(*pp, entry);
e3b37f11	119
fd706d69	120	mutex_unlock(&nf_hook_mutex);
4c091156 EB	121	#ifdef CONFIG_NETFILTER_INGRESS
	122	if (reg->pf == NFPROTO_NETDEV && reg->hooknum == NF_NETDEV_INGRESS)
	123	net_inc_ingress_queue();
	124	#endif
d1c85c2e	125	#ifdef HAVE_JUMP_LABEL
c5905afb	126	static_key_slow_inc(&nf_hooks_needed[reg->pf][reg->hooknum]);
a2d7ec58	127	#endif
f6ebe77f HW	128	return 0;
f6ebe77f HW	129	}
085db2c0	130	EXPORT_SYMBOL(nf_register_net_hook);
f6ebe77f	131
085db2c0	132	void nf_unregister_net_hook(struct net net, const struct nf_hook_ops reg)
f6ebe77f	133	{
bd3769bf LT	134	struct nf_hook_entry __rcu **pp;
	135	struct nf_hook_entry *p;
	136
	137	pp = nf_hook_entry_head(net, reg);
	138	if (WARN_ON_ONCE(!pp))
	139	return;
085db2c0	140
fd706d69	141	mutex_lock(&nf_hook_mutex);
bd3769bf LT	142	while ((p = nf_entry_dereference(*pp)) != NULL) {
	143	if (p->orig_ops == reg) {
	144	rcu_assign_pointer(*pp, p->next);
	145	break;
085db2c0	146	}
bd3769bf	147	pp = &p->next;
085db2c0	148	}
fd706d69	149	mutex_unlock(&nf_hook_mutex);
bd3769bf	150	if (!p) {
085db2c0 EB	151	WARN(1, "nf_unregister_net_hook: hook not found!\n");
	152	return;
	153	}
e687ad60	154	#ifdef CONFIG_NETFILTER_INGRESS
4c091156 EB	155	if (reg->pf == NFPROTO_NETDEV && reg->hooknum == NF_NETDEV_INGRESS)
4c091156 EB	156	net_dec_ingress_queue();
e687ad60	157	#endif
d1c85c2e	158	#ifdef HAVE_JUMP_LABEL
c5905afb	159	static_key_slow_dec(&nf_hooks_needed[reg->pf][reg->hooknum]);
a2d7ec58	160	#endif
f6ebe77f	161	synchronize_net();
bd3769bf	162	nf_queue_nf_hook_drop(net, p);
514ed62e FW	163	/* other cpu might still process nfqueue verdict that used reg */
514ed62e FW	164	synchronize_net();
bd3769bf	165	kfree(p);
085db2c0 EB	166	}
	167	EXPORT_SYMBOL(nf_unregister_net_hook);
	168
	169	int nf_register_net_hooks(struct net net, const struct nf_hook_ops reg,
	170	unsigned int n)
	171	{
	172	unsigned int i;
	173	int err = 0;
	174
	175	for (i = 0; i < n; i++) {
	176	err = nf_register_net_hook(net, &reg[i]);
	177	if (err)
	178	goto err;
	179	}
	180	return err;
	181
	182	err:
	183	if (i > 0)
	184	nf_unregister_net_hooks(net, reg, i);
	185	return err;
	186	}
	187	EXPORT_SYMBOL(nf_register_net_hooks);
	188
	189	void nf_unregister_net_hooks(struct net net, const struct nf_hook_ops reg,
	190	unsigned int n)
	191	{
	192	while (n-- > 0)
	193	nf_unregister_net_hook(net, &reg[n]);
	194	}
	195	EXPORT_SYMBOL(nf_unregister_net_hooks);
	196
	197	static LIST_HEAD(nf_hook_list);
	198
e8bffe0c	199	static int _nf_register_hook(struct nf_hook_ops *reg)
085db2c0 EB	200	{
	201	struct net net, last;
	202	int ret;
	203
085db2c0 EB	204	for_each_net(net) {
	205	ret = nf_register_net_hook(net, reg);
	206	if (ret && ret != -ENOENT)
	207	goto rollback;
	208	}
	209	list_add_tail(&reg->list, &nf_hook_list);
085db2c0 EB	210
	211	return 0;
	212	rollback:
	213	last = net;
	214	for_each_net(net) {
	215	if (net == last)
	216	break;
	217	nf_unregister_net_hook(net, reg);
	218	}
e8bffe0c MB	219	return ret;
	220	}
	221
	222	int nf_register_hook(struct nf_hook_ops *reg)
	223	{
	224	int ret;
	225
	226	rtnl_lock();
	227	ret = _nf_register_hook(reg);
085db2c0	228	rtnl_unlock();
e8bffe0c	229
085db2c0 EB	230	return ret;
	231	}
	232	EXPORT_SYMBOL(nf_register_hook);
	233
e8bffe0c	234	static void _nf_unregister_hook(struct nf_hook_ops *reg)
085db2c0 EB	235	{
	236	struct net *net;
	237
085db2c0 EB	238	list_del(&reg->list);
	239	for_each_net(net)
	240	nf_unregister_net_hook(net, reg);
e8bffe0c MB	241	}
	242
	243	void nf_unregister_hook(struct nf_hook_ops *reg)
	244	{
	245	rtnl_lock();
	246	_nf_unregister_hook(reg);
085db2c0	247	rtnl_unlock();
f6ebe77f HW	248	}
	249	EXPORT_SYMBOL(nf_unregister_hook);
	250
972d1cb1 PM	251	int nf_register_hooks(struct nf_hook_ops *reg, unsigned int n)
	252	{
	253	unsigned int i;
	254	int err = 0;
	255
	256	for (i = 0; i < n; i++) {
	257	err = nf_register_hook(&reg[i]);
	258	if (err)
	259	goto err;
	260	}
	261	return err;
	262
	263	err:
	264	if (i > 0)
	265	nf_unregister_hooks(reg, i);
	266	return err;
	267	}
	268	EXPORT_SYMBOL(nf_register_hooks);
	269
e8bffe0c MB	270	/* Caller MUST take rtnl_lock() */
	271	int _nf_register_hooks(struct nf_hook_ops *reg, unsigned int n)
	272	{
	273	unsigned int i;
	274	int err = 0;
	275
	276	for (i = 0; i < n; i++) {
	277	err = _nf_register_hook(&reg[i]);
	278	if (err)
	279	goto err;
	280	}
	281	return err;
	282
	283	err:
	284	if (i > 0)
	285	_nf_unregister_hooks(reg, i);
	286	return err;
	287	}
	288	EXPORT_SYMBOL(_nf_register_hooks);
	289
972d1cb1 PM	290	void nf_unregister_hooks(struct nf_hook_ops *reg, unsigned int n)
972d1cb1 PM	291	{
f68c5301 CG	292	while (n-- > 0)
f68c5301 CG	293	nf_unregister_hook(&reg[n]);
972d1cb1 PM	294	}
	295	EXPORT_SYMBOL(nf_unregister_hooks);
	296
e8bffe0c MB	297	/* Caller MUST take rtnl_lock */
	298	void _nf_unregister_hooks(struct nf_hook_ops *reg, unsigned int n)
	299	{
	300	while (n-- > 0)
	301	_nf_unregister_hook(&reg[n]);
	302	}
	303	EXPORT_SYMBOL(_nf_unregister_hooks);
	304
e3b37f11	305	unsigned int nf_iterate(struct sk_buff *skb,
cfdfab31	306	struct nf_hook_state *state,
e3b37f11	307	struct nf_hook_entry **entryp)
f6ebe77f HW	308	{
	309	unsigned int verdict;
	310
	311	/*
	312	* The caller must not block between calls to this
	313	* function because of risk of continuing from deleted element.
	314	*/
e3b37f11 AC	315	while (*entryp) {
	316	if (state->thresh > (*entryp)->ops.priority) {
	317	entryp = rcu_dereference((entryp)->next);
f6ebe77f	318	continue;
e3b37f11	319	}
f6ebe77f HW	320
f6ebe77f HW	321	/* Optimization: we don't need to hold module
601e68e1	322	reference here, since function can't sleep. --RR */
de9963f0	323	repeat:
e3b37f11	324	verdict = (entryp)->ops.hook((entryp)->ops.priv, skb, state);
f6ebe77f HW	325	if (verdict != NF_ACCEPT) {
	326	#ifdef CONFIG_NETFILTER_DEBUG
	327	if (unlikely((verdict & NF_VERDICT_MASK)
	328	> NF_MAX_VERDICT)) {
	329	NFDEBUG("Evil return from %p(%u).\n",
e3b37f11 AC	330	(*entryp)->ops.hook, state->hook);
e3b37f11 AC	331	entryp = rcu_dereference((entryp)->next);
f6ebe77f HW	332	continue;
	333	}
	334	#endif
2a6decfd	335	if (verdict != NF_REPEAT)
f6ebe77f	336	return verdict;
de9963f0	337	goto repeat;
f6ebe77f	338	}
e3b37f11	339	entryp = rcu_dereference((entryp)->next);
f6ebe77f HW	340	}
	341	return NF_ACCEPT;
	342	}
	343
	344
	345	/* Returns 1 if okfn() needs to be executed by the caller,
e2361cb9	346	* -EPERM for NF_DROP, 0 otherwise. Caller must hold rcu_read_lock. */
cfdfab31	347	int nf_hook_slow(struct sk_buff skb, struct nf_hook_state state)
f6ebe77f	348	{
e3b37f11	349	struct nf_hook_entry *entry;
f6ebe77f HW	350	unsigned int verdict;
	351	int ret = 0;
	352
e3b37f11	353	entry = rcu_dereference(state->hook_entries);
f6ebe77f	354	next_hook:
e3b37f11	355	verdict = nf_iterate(skb, state, &entry);
f6ebe77f HW	356	if (verdict == NF_ACCEPT \|\| verdict == NF_STOP) {
f6ebe77f HW	357	ret = 1;
da683650	358	} else if ((verdict & NF_VERDICT_MASK) == NF_DROP) {
3db05fea	359	kfree_skb(skb);
f615df76	360	ret = NF_DROP_GETERR(verdict);
da683650 EP	361	if (ret == 0)
da683650 EP	362	ret = -EPERM;
f9c63990	363	} else if ((verdict & NF_VERDICT_MASK) == NF_QUEUE) {
e3b37f11 AC	364	int err;
	365
	366	RCU_INIT_POINTER(state->hook_entries, entry);
	367	err = nf_queue(skb, state, verdict >> NF_VERDICT_QBITS);
563e1232	368	if (err < 0) {
563e1232	369	if (err == -ESRCH &&
94b27cc3 FW	370	(verdict & NF_VERDICT_FLAG_QUEUE_BYPASS))
94b27cc3 FW	371	goto next_hook;
06cdb634 FW	372	kfree_skb(skb);
06cdb634 FW	373	}
f6ebe77f	374	}
f6ebe77f HW	375	return ret;
	376	}
	377	EXPORT_SYMBOL(nf_hook_slow);
	378
	379
37d41879	380	int skb_make_writable(struct sk_buff *skb, unsigned int writable_len)
f6ebe77f	381	{
37d41879	382	if (writable_len > skb->len)
f6ebe77f HW	383	return 0;
	384
	385	/* Not exclusive use of packet? Must copy. */
37d41879 HX	386	if (!skb_cloned(skb)) {
	387	if (writable_len <= skb_headlen(skb))
	388	return 1;
	389	} else if (skb_clone_writable(skb, writable_len))
	390	return 1;
	391
	392	if (writable_len <= skb_headlen(skb))
	393	writable_len = 0;
	394	else
	395	writable_len -= skb_headlen(skb);
	396
	397	return !!__pskb_pull_tail(skb, writable_len);
f6ebe77f HW	398	}
	399	EXPORT_SYMBOL(skb_make_writable);
	400
b7bd1809 PNA	401	/* This needs to be compiled in any case to avoid dependencies between the
	402	* nfnetlink_queue code and nf_conntrack.
	403	*/
a4b4766c KM	404	struct nfnl_ct_hook __rcu *nfnl_ct_hook __read_mostly;
a4b4766c KM	405	EXPORT_SYMBOL_GPL(nfnl_ct_hook);
b7bd1809	406
c0cd1156	407	#if IS_ENABLED(CONFIG_NF_CONNTRACK)
f6ebe77f HW	408	/* This does not belong here, but locally generated errors need it if connection
	409	tracking in use: without this, connection may not be in hash table, and hence
	410	manufactured ICMP or RST packets will not be associated with it. */
312a0c16 PM	411	void (ip_ct_attach)(struct sk_buff , const struct sk_buff *)
312a0c16 PM	412	__rcu __read_mostly;
f6ebe77f HW	413	EXPORT_SYMBOL(ip_ct_attach);
f6ebe77f HW	414
312a0c16	415	void nf_ct_attach(struct sk_buff new, const struct sk_buff skb)
f6ebe77f	416	{
312a0c16	417	void (attach)(struct sk_buff , const struct sk_buff *);
f6ebe77f	418
c3a47ab3 PM	419	if (skb->nfct) {
	420	rcu_read_lock();
	421	attach = rcu_dereference(ip_ct_attach);
	422	if (attach)
	423	attach(new, skb);
	424	rcu_read_unlock();
f6ebe77f HW	425	}
	426	}
	427	EXPORT_SYMBOL(nf_ct_attach);
de6e05c4	428
0e60ebe0	429	void (nf_ct_destroy)(struct nf_conntrack ) __rcu __read_mostly;
de6e05c4 YK	430	EXPORT_SYMBOL(nf_ct_destroy);
	431
	432	void nf_conntrack_destroy(struct nf_conntrack *nfct)
	433	{
	434	void (destroy)(struct nf_conntrack );
	435
	436	rcu_read_lock();
	437	destroy = rcu_dereference(nf_ct_destroy);
	438	BUG_ON(destroy == NULL);
	439	destroy(nfct);
	440	rcu_read_unlock();
	441	}
	442	EXPORT_SYMBOL(nf_conntrack_destroy);
9cb01766	443
62da9865 DB	444	/* Built-in default zone used e.g. by modules. */
	445	const struct nf_conntrack_zone nf_ct_zone_dflt = {
	446	.id = NF_CT_DEFAULT_ZONE_ID,
	447	.dir = NF_CT_DEFAULT_ZONE_DIR,
	448	};
	449	EXPORT_SYMBOL_GPL(nf_ct_zone_dflt);
de6e05c4	450	#endif /* CONFIG_NF_CONNTRACK */
f6ebe77f	451
c7232c99 PM	452	#ifdef CONFIG_NF_NAT_NEEDED
	453	void (nf_nat_decode_session_hook)(struct sk_buff , struct flowi *);
	454	EXPORT_SYMBOL(nf_nat_decode_session_hook);
	455	#endif
	456
085db2c0 EB	457	static int nf_register_hook_list(struct net *net)
	458	{
	459	struct nf_hook_ops *elem;
	460	int ret;
	461
	462	rtnl_lock();
	463	list_for_each_entry(elem, &nf_hook_list, list) {
	464	ret = nf_register_net_hook(net, elem);
	465	if (ret && ret != -ENOENT)
	466	goto out_undo;
	467	}
	468	rtnl_unlock();
	469	return 0;
	470
	471	out_undo:
	472	list_for_each_entry_continue_reverse(elem, &nf_hook_list, list)
	473	nf_unregister_net_hook(net, elem);
	474	rtnl_unlock();
	475	return ret;
	476	}
	477
	478	static void nf_unregister_hook_list(struct net *net)
	479	{
	480	struct nf_hook_ops *elem;
	481
	482	rtnl_lock();
	483	list_for_each_entry(elem, &nf_hook_list, list)
	484	nf_unregister_net_hook(net, elem);
	485	rtnl_unlock();
	486	}
	487
f3c1a44a G	488	static int __net_init netfilter_net_init(struct net *net)
f3c1a44a G	489	{
085db2c0 EB	490	int i, h, ret;
	491
	492	for (i = 0; i < ARRAY_SIZE(net->nf.hooks); i++) {
	493	for (h = 0; h < NF_MAX_HOOKS; h++)
e3b37f11	494	RCU_INIT_POINTER(net->nf.hooks[i][h], NULL);
085db2c0 EB	495	}
085db2c0 EB	496
f3c1a44a G	497	#ifdef CONFIG_PROC_FS
	498	net->nf.proc_netfilter = proc_net_mkdir(net, "netfilter",
	499	net->proc_net);
12202fa7 PNA	500	if (!net->nf.proc_netfilter) {
	501	if (!net_eq(net, &init_net))
	502	pr_err("cannot create netfilter proc entry");
	503
f3c1a44a G	504	return -ENOMEM;
	505	}
	506	#endif
085db2c0 EB	507	ret = nf_register_hook_list(net);
	508	if (ret)
	509	remove_proc_entry("netfilter", net->proc_net);
	510
	511	return ret;
f3c1a44a G	512	}
	513
	514	static void __net_exit netfilter_net_exit(struct net *net)
	515	{
085db2c0	516	nf_unregister_hook_list(net);
f3c1a44a G	517	remove_proc_entry("netfilter", net->proc_net);
	518	}
	519
	520	static struct pernet_operations netfilter_net_ops = {
	521	.init = netfilter_net_init,
	522	.exit = netfilter_net_exit,
	523	};
	524
6d11cfdb	525	int __init netfilter_init(void)
f6ebe77f	526	{
085db2c0	527	int ret;
f6ebe77f	528
6d11cfdb PNA	529	ret = register_pernet_subsys(&netfilter_net_ops);
	530	if (ret < 0)
	531	goto err;
	532
	533	ret = netfilter_log_init();
	534	if (ret < 0)
	535	goto err_pernet;
f6ebe77f	536
6d11cfdb PNA	537	return 0;
	538	err_pernet:
	539	unregister_pernet_subsys(&netfilter_net_ops);
	540	err:
	541	return ret;
f6ebe77f	542	}