[deliverable/linux.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.
 */
#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/inetdevice.h>
#include <linux/proc_fs.h>
#include <linux/mutex.h>
#include <linux/slab.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);

int nf_register_afinfo(const struct nf_afinfo *afinfo)
{
	int err;

	err = mutex_lock_interruptible(&afinfo_mutex);
	if (err < 0)
		return err;
	rcu_assign_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_assign_pointer(nf_afinfo[afinfo->family], NULL);
	mutex_unlock(&afinfo_mutex);
	synchronize_rcu();
}
EXPORT_SYMBOL_GPL(nf_unregister_afinfo);

struct list_head nf_hooks[NFPROTO_NUMPROTO][NF_MAX_HOOKS] __read_mostly;
EXPORT_SYMBOL(nf_hooks);
static DEFINE_MUTEX(nf_hook_mutex);

int nf_register_hook(struct nf_hook_ops *reg)
{
	struct nf_hook_ops *elem;
	int err;

	err = mutex_lock_interruptible(&nf_hook_mutex);
	if (err < 0)
		return err;
	list_for_each_entry(elem, &nf_hooks[reg->pf][reg->hooknum], list) {
		if (reg->priority < elem->priority)
			break;
	}
	list_add_rcu(&reg->list, elem->list.prev);
	mutex_unlock(&nf_hook_mutex);
	return 0;
}
EXPORT_SYMBOL(nf_register_hook);

void nf_unregister_hook(struct nf_hook_ops *reg)
{
	mutex_lock(&nf_hook_mutex);
	list_del_rcu(&reg->list);
	mutex_unlock(&nf_hook_mutex);

	synchronize_net();
}
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);

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 list_head *head,
			struct sk_buff *skb,
			unsigned int hook,
			const struct net_device *indev,
			const struct net_device *outdev,
			struct list_head **i,
			int (*okfn)(struct sk_buff *),
			int hook_thresh)
{
	unsigned int verdict;

	/*
	 * The caller must not block between calls to this
	 * function because of risk of continuing from deleted element.
	 */
	list_for_each_continue_rcu(*i, head) {
		struct nf_hook_ops *elem = (struct nf_hook_ops *)*i;

		if (hook_thresh > elem->priority)
			continue;

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


/* Returns 1 if okfn() needs to be executed by the caller,
 * -EPERM for NF_DROP, 0 otherwise. */
int nf_hook_slow(u_int8_t pf, unsigned int hook, struct sk_buff *skb,
		 struct net_device *indev,
		 struct net_device *outdev,
		 int (*okfn)(struct sk_buff *),
		 int hook_thresh)
{
	struct list_head *elem;
	unsigned int verdict;
	int ret = 0;

	/* We may already have this, but read-locks nest anyway */
	rcu_read_lock();

	elem = &nf_hooks[pf][hook];
next_hook:
	verdict = nf_iterate(&nf_hooks[pf][hook], skb, hook, indev,
			     outdev, &elem, okfn, hook_thresh);
	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) {
		ret = nf_queue(skb, elem, pf, hook, indev, outdev, okfn,
			       verdict >> NF_VERDICT_QBITS);
		if (ret < 0) {
			if (ret == -ECANCELED)
				goto next_hook;
			if (ret == -ESRCH &&
			   (verdict & NF_VERDICT_FLAG_QUEUE_BYPASS))
				goto next_hook;
			kfree_skb(skb);
		}
		ret = 0;
	}
	rcu_read_unlock();
	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);

#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE)
/* 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 *, struct sk_buff *) __rcu __read_mostly;
EXPORT_SYMBOL(ip_ct_attach);

void nf_ct_attach(struct sk_buff *new, struct sk_buff *skb)
{
	void (*attach)(struct sk_buff *, 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);
#endif /* CONFIG_NF_CONNTRACK */

#ifdef CONFIG_PROC_FS
struct proc_dir_entry *proc_net_netfilter;
EXPORT_SYMBOL(proc_net_netfilter);
#endif

void __init netfilter_init(void)
{
	int i, h;
	for (i = 0; i < ARRAY_SIZE(nf_hooks); i++) {
		for (h = 0; h < NF_MAX_HOOKS; h++)
			INIT_LIST_HEAD(&nf_hooks[i][h]);
	}

#ifdef CONFIG_PROC_FS
	proc_net_netfilter = proc_mkdir("netfilter", init_net.proc_net);
	if (!proc_net_netfilter)
		panic("cannot create netfilter proc entry");
#endif

	if (netfilter_queue_init() < 0)
		panic("cannot initialize nf_queue");
	if (netfilter_log_init() < 0)
		panic("cannot initialize nf_log");
}

#ifdef CONFIG_SYSCTL
struct ctl_path nf_net_netfilter_sysctl_path[] = {
	{ .procname = "net", },
	{ .procname = "netfilter", },
	{ }
};
EXPORT_SYMBOL_GPL(nf_net_netfilter_sysctl_path);
#endif /* CONFIG_SYSCTL */
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.
f6ebe77f	8	*/
f6ebe77f HW	9	#include <linux/kernel.h>
	10	#include <linux/netfilter.h>
	11	#include <net/protocol.h>
	12	#include <linux/init.h>
	13	#include <linux/skbuff.h>
	14	#include <linux/wait.h>
	15	#include <linux/module.h>
	16	#include <linux/interrupt.h>
	17	#include <linux/if.h>
	18	#include <linux/netdevice.h>
	19	#include <linux/inetdevice.h>
	20	#include <linux/proc_fs.h>
d486dd1f	21	#include <linux/mutex.h>
5a0e3ad6	22	#include <linux/slab.h>
457c4cbc	23	#include <net/net_namespace.h>
f6ebe77f HW	24	#include <net/sock.h>
	25
	26	#include "nf_internals.h"
	27
d486dd1f	28	static DEFINE_MUTEX(afinfo_mutex);
bce8032e	29
0906a372	30	const struct nf_afinfo __rcu *nf_afinfo[NFPROTO_NUMPROTO] __read_mostly;
bce8032e PM	31	EXPORT_SYMBOL(nf_afinfo);
bce8032e PM	32
1e796fda	33	int nf_register_afinfo(const struct nf_afinfo *afinfo)
bce8032e	34	{
d486dd1f PM	35	int err;
	36
	37	err = mutex_lock_interruptible(&afinfo_mutex);
	38	if (err < 0)
	39	return err;
bce8032e	40	rcu_assign_pointer(nf_afinfo[afinfo->family], afinfo);
d486dd1f	41	mutex_unlock(&afinfo_mutex);
bce8032e PM	42	return 0;
	43	}
	44	EXPORT_SYMBOL_GPL(nf_register_afinfo);
	45
1e796fda	46	void nf_unregister_afinfo(const struct nf_afinfo *afinfo)
bce8032e	47	{
d486dd1f	48	mutex_lock(&afinfo_mutex);
bce8032e	49	rcu_assign_pointer(nf_afinfo[afinfo->family], NULL);
d486dd1f	50	mutex_unlock(&afinfo_mutex);
bce8032e PM	51	synchronize_rcu();
	52	}
	53	EXPORT_SYMBOL_GPL(nf_unregister_afinfo);
	54
7e9c6eeb	55	struct list_head nf_hooks[NFPROTO_NUMPROTO][NF_MAX_HOOKS] __read_mostly;
f6ebe77f	56	EXPORT_SYMBOL(nf_hooks);
fd706d69	57	static DEFINE_MUTEX(nf_hook_mutex);
f6ebe77f HW	58
	59	int nf_register_hook(struct nf_hook_ops *reg)
	60	{
4c610979	61	struct nf_hook_ops *elem;
fd706d69	62	int err;
f6ebe77f	63
fd706d69 PM	64	err = mutex_lock_interruptible(&nf_hook_mutex);
	65	if (err < 0)
	66	return err;
4c610979 LZ	67	list_for_each_entry(elem, &nf_hooks[reg->pf][reg->hooknum], list) {
4c610979 LZ	68	if (reg->priority < elem->priority)
f6ebe77f HW	69	break;
f6ebe77f HW	70	}
4c610979	71	list_add_rcu(&reg->list, elem->list.prev);
fd706d69	72	mutex_unlock(&nf_hook_mutex);
f6ebe77f HW	73	return 0;
	74	}
	75	EXPORT_SYMBOL(nf_register_hook);
	76
	77	void nf_unregister_hook(struct nf_hook_ops *reg)
	78	{
fd706d69	79	mutex_lock(&nf_hook_mutex);
f6ebe77f	80	list_del_rcu(&reg->list);
fd706d69	81	mutex_unlock(&nf_hook_mutex);
f6ebe77f HW	82
	83	synchronize_net();
	84	}
	85	EXPORT_SYMBOL(nf_unregister_hook);
	86
972d1cb1 PM	87	int nf_register_hooks(struct nf_hook_ops *reg, unsigned int n)
	88	{
	89	unsigned int i;
	90	int err = 0;
	91
	92	for (i = 0; i < n; i++) {
	93	err = nf_register_hook(&reg[i]);
	94	if (err)
	95	goto err;
	96	}
	97	return err;
	98
	99	err:
	100	if (i > 0)
	101	nf_unregister_hooks(reg, i);
	102	return err;
	103	}
	104	EXPORT_SYMBOL(nf_register_hooks);
	105
	106	void nf_unregister_hooks(struct nf_hook_ops *reg, unsigned int n)
	107	{
f68c5301 CG	108	while (n-- > 0)
f68c5301 CG	109	nf_unregister_hook(&reg[n]);
972d1cb1 PM	110	}
	111	EXPORT_SYMBOL(nf_unregister_hooks);
	112
f6ebe77f	113	unsigned int nf_iterate(struct list_head *head,
3db05fea	114	struct sk_buff *skb,
76108cea	115	unsigned int hook,
f6ebe77f HW	116	const struct net_device *indev,
	117	const struct net_device *outdev,
	118	struct list_head **i,
	119	int (okfn)(struct sk_buff ),
	120	int hook_thresh)
	121	{
	122	unsigned int verdict;
	123
	124	/*
	125	* The caller must not block between calls to this
	126	* function because of risk of continuing from deleted element.
	127	*/
	128	list_for_each_continue_rcu(*i, head) {
	129	struct nf_hook_ops elem = (struct nf_hook_ops )*i;
	130
	131	if (hook_thresh > elem->priority)
	132	continue;
	133
	134	/* Optimization: we don't need to hold module
601e68e1	135	reference here, since function can't sleep. --RR */
de9963f0	136	repeat:
f6ebe77f HW	137	verdict = elem->hook(hook, skb, indev, outdev, okfn);
	138	if (verdict != NF_ACCEPT) {
	139	#ifdef CONFIG_NETFILTER_DEBUG
	140	if (unlikely((verdict & NF_VERDICT_MASK)
	141	> NF_MAX_VERDICT)) {
	142	NFDEBUG("Evil return from %p(%u).\n",
601e68e1	143	elem->hook, hook);
f6ebe77f HW	144	continue;
	145	}
	146	#endif
	147	if (verdict != NF_REPEAT)
	148	return verdict;
de9963f0	149	goto repeat;
f6ebe77f HW	150	}
	151	}
	152	return NF_ACCEPT;
	153	}
	154
	155
	156	/* Returns 1 if okfn() needs to be executed by the caller,
	157	* -EPERM for NF_DROP, 0 otherwise. */
76108cea	158	int nf_hook_slow(u_int8_t pf, unsigned int hook, struct sk_buff *skb,
f6ebe77f HW	159	struct net_device *indev,
	160	struct net_device *outdev,
	161	int (okfn)(struct sk_buff ),
	162	int hook_thresh)
	163	{
	164	struct list_head *elem;
	165	unsigned int verdict;
	166	int ret = 0;
	167
	168	/* We may already have this, but read-locks nest anyway */
	169	rcu_read_lock();
	170
	171	elem = &nf_hooks[pf][hook];
	172	next_hook:
3db05fea	173	verdict = nf_iterate(&nf_hooks[pf][hook], skb, hook, indev,
f6ebe77f HW	174	outdev, &elem, okfn, hook_thresh);
	175	if (verdict == NF_ACCEPT \|\| verdict == NF_STOP) {
	176	ret = 1;
da683650	177	} else if ((verdict & NF_VERDICT_MASK) == NF_DROP) {
3db05fea	178	kfree_skb(skb);
f615df76	179	ret = NF_DROP_GETERR(verdict);
da683650 EP	180	if (ret == 0)
da683650 EP	181	ret = -EPERM;
f9c63990	182	} else if ((verdict & NF_VERDICT_MASK) == NF_QUEUE) {
f1585086	183	ret = nf_queue(skb, elem, pf, hook, indev, outdev, okfn,
f615df76	184	verdict >> NF_VERDICT_QBITS);
06cdb634 FW	185	if (ret < 0) {
	186	if (ret == -ECANCELED)
	187	goto next_hook;
94b27cc3 FW	188	if (ret == -ESRCH &&
	189	(verdict & NF_VERDICT_FLAG_QUEUE_BYPASS))
	190	goto next_hook;
06cdb634 FW	191	kfree_skb(skb);
06cdb634 FW	192	}
f1585086	193	ret = 0;
f6ebe77f	194	}
f6ebe77f HW	195	rcu_read_unlock();
	196	return ret;
	197	}
	198	EXPORT_SYMBOL(nf_hook_slow);
	199
	200
37d41879	201	int skb_make_writable(struct sk_buff *skb, unsigned int writable_len)
f6ebe77f	202	{
37d41879	203	if (writable_len > skb->len)
f6ebe77f HW	204	return 0;
	205
	206	/* Not exclusive use of packet? Must copy. */
37d41879 HX	207	if (!skb_cloned(skb)) {
	208	if (writable_len <= skb_headlen(skb))
	209	return 1;
	210	} else if (skb_clone_writable(skb, writable_len))
	211	return 1;
	212
	213	if (writable_len <= skb_headlen(skb))
	214	writable_len = 0;
	215	else
	216	writable_len -= skb_headlen(skb);
	217
	218	return !!__pskb_pull_tail(skb, writable_len);
f6ebe77f HW	219	}
	220	EXPORT_SYMBOL(skb_make_writable);
	221
5f79e0f9	222	#if defined(CONFIG_NF_CONNTRACK) \|\| defined(CONFIG_NF_CONNTRACK_MODULE)
f6ebe77f HW	223	/* This does not belong here, but locally generated errors need it if connection
	224	tracking in use: without this, connection may not be in hash table, and hence
	225	manufactured ICMP or RST packets will not be associated with it. */
0e60ebe0	226	void (ip_ct_attach)(struct sk_buff , struct sk_buff *) __rcu __read_mostly;
f6ebe77f HW	227	EXPORT_SYMBOL(ip_ct_attach);
	228
	229	void nf_ct_attach(struct sk_buff new, struct sk_buff skb)
	230	{
	231	void (attach)(struct sk_buff , struct sk_buff *);
	232
c3a47ab3 PM	233	if (skb->nfct) {
	234	rcu_read_lock();
	235	attach = rcu_dereference(ip_ct_attach);
	236	if (attach)
	237	attach(new, skb);
	238	rcu_read_unlock();
f6ebe77f HW	239	}
	240	}
	241	EXPORT_SYMBOL(nf_ct_attach);
de6e05c4	242
0e60ebe0	243	void (nf_ct_destroy)(struct nf_conntrack ) __rcu __read_mostly;
de6e05c4 YK	244	EXPORT_SYMBOL(nf_ct_destroy);
	245
	246	void nf_conntrack_destroy(struct nf_conntrack *nfct)
	247	{
	248	void (destroy)(struct nf_conntrack );
	249
	250	rcu_read_lock();
	251	destroy = rcu_dereference(nf_ct_destroy);
	252	BUG_ON(destroy == NULL);
	253	destroy(nfct);
	254	rcu_read_unlock();
	255	}
	256	EXPORT_SYMBOL(nf_conntrack_destroy);
	257	#endif /* CONFIG_NF_CONNTRACK */
f6ebe77f HW	258
	259	#ifdef CONFIG_PROC_FS
	260	struct proc_dir_entry *proc_net_netfilter;
	261	EXPORT_SYMBOL(proc_net_netfilter);
	262	#endif
	263
	264	void __init netfilter_init(void)
	265	{
	266	int i, h;
7e9c6eeb	267	for (i = 0; i < ARRAY_SIZE(nf_hooks); i++) {
f6ebe77f HW	268	for (h = 0; h < NF_MAX_HOOKS; h++)
	269	INIT_LIST_HEAD(&nf_hooks[i][h]);
	270	}
	271
	272	#ifdef CONFIG_PROC_FS
457c4cbc	273	proc_net_netfilter = proc_mkdir("netfilter", init_net.proc_net);
f6ebe77f HW	274	if (!proc_net_netfilter)
	275	panic("cannot create netfilter proc entry");
	276	#endif
	277
	278	if (netfilter_queue_init() < 0)
	279	panic("cannot initialize nf_queue");
	280	if (netfilter_log_init() < 0)
	281	panic("cannot initialize nf_log");
	282	}
4f536522 PM	283
	284	#ifdef CONFIG_SYSCTL
	285	struct ctl_path nf_net_netfilter_sysctl_path[] = {
f8572d8f EB	286	{ .procname = "net", },
f8572d8f EB	287	{ .procname = "netfilter", },
4f536522 PM	288	{ }
	289	};
	290	EXPORT_SYMBOL_GPL(nf_net_netfilter_sysctl_path);
	291	#endif /* CONFIG_SYSCTL */