[net-next-2.6.git] / net / netfilter / nf_queue.c

#include <linux/kernel.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/proc_fs.h>
#include <linux/skbuff.h>
#include <linux/netfilter.h>
#include <linux/seq_file.h>
#include <linux/rcupdate.h>
#include <net/protocol.h>

#include "nf_internals.h"

/*
 * A queue handler may be registered for each protocol.  Each is protected by
 * long term mutex.  The handler must provide an an outfn() to accept packets
 * for queueing and must reinject all packets it receives, no matter what.
 */
static struct nf_queue_handler *queue_handler[NPROTO];

static DEFINE_MUTEX(queue_handler_mutex);

/* return EBUSY when somebody else is registered, return EEXIST if the
 * same handler is registered, return 0 in case of success. */
int nf_register_queue_handler(int pf, struct nf_queue_handler *qh)
{
	int ret;

	if (pf >= NPROTO)
		return -EINVAL;

	mutex_lock(&queue_handler_mutex);
	if (queue_handler[pf] == qh)
		ret = -EEXIST;
	else if (queue_handler[pf])
		ret = -EBUSY;
	else {
		rcu_assign_pointer(queue_handler[pf], qh);
		ret = 0;
	}
	mutex_unlock(&queue_handler_mutex);

	return ret;
}
EXPORT_SYMBOL(nf_register_queue_handler);

/* The caller must flush their queue before this */
int nf_unregister_queue_handler(int pf, struct nf_queue_handler *qh)
{
	if (pf >= NPROTO)
		return -EINVAL;

	mutex_lock(&queue_handler_mutex);
	if (queue_handler[pf] != qh) {
		mutex_unlock(&queue_handler_mutex);
		return -EINVAL;
	}

	rcu_assign_pointer(queue_handler[pf], NULL);
	mutex_unlock(&queue_handler_mutex);

	synchronize_rcu();

	return 0;
}
EXPORT_SYMBOL(nf_unregister_queue_handler);

void nf_unregister_queue_handlers(struct nf_queue_handler *qh)
{
	int pf;

	mutex_lock(&queue_handler_mutex);
	for (pf = 0; pf < NPROTO; pf++)  {
		if (queue_handler[pf] == qh)
			rcu_assign_pointer(queue_handler[pf], NULL);
	}
	mutex_unlock(&queue_handler_mutex);

	synchronize_rcu();
}
EXPORT_SYMBOL_GPL(nf_unregister_queue_handlers);

/*
 * Any packet that leaves via this function must come back
 * through nf_reinject().
 */
static int __nf_queue(struct sk_buff *skb,
		      struct list_head *elem,
		      int pf, unsigned int hook,
		      struct net_device *indev,
		      struct net_device *outdev,
		      int (*okfn)(struct sk_buff *),
		      unsigned int queuenum)
{
	int status;
	struct nf_info *info;
#ifdef CONFIG_BRIDGE_NETFILTER
	struct net_device *physindev = NULL;
	struct net_device *physoutdev = NULL;
#endif
	struct nf_afinfo *afinfo;
	struct nf_queue_handler *qh;

	/* QUEUE == DROP if noone is waiting, to be safe. */
	rcu_read_lock();

	qh = rcu_dereference(queue_handler[pf]);
	if (!qh) {
		rcu_read_unlock();
		kfree_skb(skb);
		return 1;
	}

	afinfo = nf_get_afinfo(pf);
	if (!afinfo) {
		rcu_read_unlock();
		kfree_skb(skb);
		return 1;
	}

	info = kmalloc(sizeof(*info) + afinfo->route_key_size, GFP_ATOMIC);
	if (!info) {
		if (net_ratelimit())
			printk(KERN_ERR "OOM queueing packet %p\n",
			       skb);
		rcu_read_unlock();
		kfree_skb(skb);
		return 1;
	}

	*info = (struct nf_info) {
		(struct nf_hook_ops *)elem, pf, hook, indev, outdev, okfn };

	/* If it's going away, ignore hook. */
	if (!try_module_get(info->elem->owner)) {
		rcu_read_unlock();
		kfree(info);
		return 0;
	}

	/* Bump dev refs so they don't vanish while packet is out */
	if (indev) dev_hold(indev);
	if (outdev) dev_hold(outdev);

#ifdef CONFIG_BRIDGE_NETFILTER
	if (skb->nf_bridge) {
		physindev = skb->nf_bridge->physindev;
		if (physindev) dev_hold(physindev);
		physoutdev = skb->nf_bridge->physoutdev;
		if (physoutdev) dev_hold(physoutdev);
	}
#endif
	afinfo->saveroute(skb, info);
	status = qh->outfn(skb, info, queuenum, qh->data);

	rcu_read_unlock();

	if (status < 0) {
		/* James M doesn't say fuck enough. */
		if (indev) dev_put(indev);
		if (outdev) dev_put(outdev);
#ifdef CONFIG_BRIDGE_NETFILTER
		if (physindev) dev_put(physindev);
		if (physoutdev) dev_put(physoutdev);
#endif
		module_put(info->elem->owner);
		kfree(info);
		kfree_skb(skb);

		return 1;
	}

	return 1;
}

int nf_queue(struct sk_buff *skb,
	     struct list_head *elem,
	     int pf, unsigned int hook,
	     struct net_device *indev,
	     struct net_device *outdev,
	     int (*okfn)(struct sk_buff *),
	     unsigned int queuenum)
{
	struct sk_buff *segs;

	if (!skb_is_gso(skb))
		return __nf_queue(skb, elem, pf, hook, indev, outdev, okfn,
				  queuenum);

	switch (pf) {
	case AF_INET:
		skb->protocol = htons(ETH_P_IP);
		break;
	case AF_INET6:
		skb->protocol = htons(ETH_P_IPV6);
		break;
	}

	segs = skb_gso_segment(skb, 0);
	kfree_skb(skb);
	if (unlikely(IS_ERR(segs)))
		return 1;

	do {
		struct sk_buff *nskb = segs->next;

		segs->next = NULL;
		if (!__nf_queue(segs, elem, pf, hook, indev, outdev, okfn,
				queuenum))
			kfree_skb(segs);
		segs = nskb;
	} while (segs);
	return 1;
}

void nf_reinject(struct sk_buff *skb, struct nf_info *info,
		 unsigned int verdict)
{
	struct list_head *elem = &info->elem->list;
	struct list_head *i;
	struct nf_afinfo *afinfo;

	rcu_read_lock();

	/* Release those devices we held, or Alexey will kill me. */
	if (info->indev) dev_put(info->indev);
	if (info->outdev) dev_put(info->outdev);
#ifdef CONFIG_BRIDGE_NETFILTER
	if (skb->nf_bridge) {
		if (skb->nf_bridge->physindev)
			dev_put(skb->nf_bridge->physindev);
		if (skb->nf_bridge->physoutdev)
			dev_put(skb->nf_bridge->physoutdev);
	}
#endif

	/* Drop reference to owner of hook which queued us. */
	module_put(info->elem->owner);

	list_for_each_rcu(i, &nf_hooks[info->pf][info->hook]) {
		if (i == elem)
			break;
	}

	if (i == &nf_hooks[info->pf][info->hook]) {
		/* The module which sent it to userspace is gone. */
		NFDEBUG("%s: module disappeared, dropping packet.\n",
			__FUNCTION__);
		verdict = NF_DROP;
	}

	/* Continue traversal iff userspace said ok... */
	if (verdict == NF_REPEAT) {
		elem = elem->prev;
		verdict = NF_ACCEPT;
	}

	if (verdict == NF_ACCEPT) {
		afinfo = nf_get_afinfo(info->pf);
		if (!afinfo || afinfo->reroute(skb, info) < 0)
			verdict = NF_DROP;
	}

	if (verdict == NF_ACCEPT) {
	next_hook:
		verdict = nf_iterate(&nf_hooks[info->pf][info->hook],
				     skb, info->hook,
				     info->indev, info->outdev, &elem,
				     info->okfn, INT_MIN);
	}

	switch (verdict & NF_VERDICT_MASK) {
	case NF_ACCEPT:
	case NF_STOP:
		info->okfn(skb);
	case NF_STOLEN:
		break;
	case NF_QUEUE:
		if (!__nf_queue(skb, elem, info->pf, info->hook,
				info->indev, info->outdev, info->okfn,
				verdict >> NF_VERDICT_BITS))
			goto next_hook;
		break;
	default:
		kfree_skb(skb);
	}
	rcu_read_unlock();
	kfree(info);
	return;
}
EXPORT_SYMBOL(nf_reinject);

#ifdef CONFIG_PROC_FS
static void *seq_start(struct seq_file *seq, loff_t *pos)
{
	if (*pos >= NPROTO)
		return NULL;

	return pos;
}

static void *seq_next(struct seq_file *s, void *v, loff_t *pos)
{
	(*pos)++;

	if (*pos >= NPROTO)
		return NULL;

	return pos;
}

static void seq_stop(struct seq_file *s, void *v)
{

}

static int seq_show(struct seq_file *s, void *v)
{
	int ret;
	loff_t *pos = v;
	struct nf_queue_handler *qh;

	rcu_read_lock();
	qh = rcu_dereference(queue_handler[*pos]);
	if (!qh)
		ret = seq_printf(s, "%2lld NONE\n", *pos);
	else
		ret = seq_printf(s, "%2lld %s\n", *pos, qh->name);
	rcu_read_unlock();

	return ret;
}

static const struct seq_operations nfqueue_seq_ops = {
	.start	= seq_start,
	.next	= seq_next,
	.stop	= seq_stop,
	.show	= seq_show,
};

static int nfqueue_open(struct inode *inode, struct file *file)
{
	return seq_open(file, &nfqueue_seq_ops);
}

static const struct file_operations nfqueue_file_ops = {
	.owner	 = THIS_MODULE,
	.open	 = nfqueue_open,
	.read	 = seq_read,
	.llseek	 = seq_lseek,
	.release = seq_release,
};
#endif /* PROC_FS */


int __init netfilter_queue_init(void)
{
#ifdef CONFIG_PROC_FS
	struct proc_dir_entry *pde;

	pde = create_proc_entry("nf_queue", S_IRUGO, proc_net_netfilter);
	if (!pde)
		return -1;
	pde->proc_fops = &nfqueue_file_ops;
#endif
	return 0;
}
Commit	Line	Data
f6ebe77f HW	1	#include <linux/kernel.h>
	2	#include <linux/init.h>
	3	#include <linux/module.h>
	4	#include <linux/proc_fs.h>
	5	#include <linux/skbuff.h>
	6	#include <linux/netfilter.h>
bbd86b9f	7	#include <linux/seq_file.h>
7a11b984	8	#include <linux/rcupdate.h>
f6ebe77f HW	9	#include <net/protocol.h>
	10
	11	#include "nf_internals.h"
	12
601e68e1	13	/*
f6ebe77f HW	14	* A queue handler may be registered for each protocol. Each is protected by
	15	* long term mutex. The handler must provide an an outfn() to accept packets
	16	* for queueing and must reinject all packets it receives, no matter what.
	17	*/
bbd86b9f	18	static struct nf_queue_handler *queue_handler[NPROTO];
f6ebe77f	19
585426fd	20	static DEFINE_MUTEX(queue_handler_mutex);
f6ebe77f	21
d72367b6 HW	22	/* return EBUSY when somebody else is registered, return EEXIST if the
d72367b6 HW	23	* same handler is registered, return 0 in case of success. */
bbd86b9f	24	int nf_register_queue_handler(int pf, struct nf_queue_handler *qh)
601e68e1	25	{
f6ebe77f HW	26	int ret;
	27
	28	if (pf >= NPROTO)
	29	return -EINVAL;
	30
585426fd	31	mutex_lock(&queue_handler_mutex);
d72367b6 HW	32	if (queue_handler[pf] == qh)
	33	ret = -EEXIST;
	34	else if (queue_handler[pf])
f6ebe77f HW	35	ret = -EBUSY;
f6ebe77f HW	36	else {
585426fd	37	rcu_assign_pointer(queue_handler[pf], qh);
f6ebe77f HW	38	ret = 0;
f6ebe77f HW	39	}
585426fd	40	mutex_unlock(&queue_handler_mutex);
f6ebe77f HW	41
	42	return ret;
	43	}
	44	EXPORT_SYMBOL(nf_register_queue_handler);
	45
	46	/* The caller must flush their queue before this */
ce7663d8	47	int nf_unregister_queue_handler(int pf, struct nf_queue_handler *qh)
f6ebe77f HW	48	{
	49	if (pf >= NPROTO)
	50	return -EINVAL;
	51
585426fd	52	mutex_lock(&queue_handler_mutex);
ce7663d8	53	if (queue_handler[pf] != qh) {
585426fd	54	mutex_unlock(&queue_handler_mutex);
ce7663d8 YK	55	return -EINVAL;
	56	}
	57
585426fd YK	58	rcu_assign_pointer(queue_handler[pf], NULL);
	59	mutex_unlock(&queue_handler_mutex);
	60
	61	synchronize_rcu();
601e68e1	62
f6ebe77f HW	63	return 0;
	64	}
	65	EXPORT_SYMBOL(nf_unregister_queue_handler);
	66
bbd86b9f	67	void nf_unregister_queue_handlers(struct nf_queue_handler *qh)
f6ebe77f HW	68	{
	69	int pf;
	70
585426fd	71	mutex_lock(&queue_handler_mutex);
f6ebe77f	72	for (pf = 0; pf < NPROTO; pf++) {
bbd86b9f	73	if (queue_handler[pf] == qh)
585426fd	74	rcu_assign_pointer(queue_handler[pf], NULL);
f6ebe77f	75	}
585426fd YK	76	mutex_unlock(&queue_handler_mutex);
	77
	78	synchronize_rcu();
f6ebe77f HW	79	}
	80	EXPORT_SYMBOL_GPL(nf_unregister_queue_handlers);
	81
601e68e1 YH	82	/*
601e68e1 YH	83	* Any packet that leaves via this function must come back
f6ebe77f HW	84	* through nf_reinject().
f6ebe77f HW	85	*/
394f545d PM	86	static int __nf_queue(struct sk_buff *skb,
	87	struct list_head *elem,
	88	int pf, unsigned int hook,
	89	struct net_device *indev,
	90	struct net_device *outdev,
	91	int (okfn)(struct sk_buff ),
	92	unsigned int queuenum)
f6ebe77f HW	93	{
	94	int status;
	95	struct nf_info *info;
	96	#ifdef CONFIG_BRIDGE_NETFILTER
	97	struct net_device *physindev = NULL;
	98	struct net_device *physoutdev = NULL;
	99	#endif
bce8032e	100	struct nf_afinfo *afinfo;
585426fd	101	struct nf_queue_handler *qh;
f6ebe77f HW	102
f6ebe77f HW	103	/* QUEUE == DROP if noone is waiting, to be safe. */
585426fd YK	104	rcu_read_lock();
	105
	106	qh = rcu_dereference(queue_handler[pf]);
	107	if (!qh) {
	108	rcu_read_unlock();
394f545d	109	kfree_skb(skb);
f6ebe77f HW	110	return 1;
	111	}
	112
bce8032e PM	113	afinfo = nf_get_afinfo(pf);
bce8032e PM	114	if (!afinfo) {
585426fd	115	rcu_read_unlock();
394f545d	116	kfree_skb(skb);
bce8032e PM	117	return 1;
	118	}
	119
	120	info = kmalloc(sizeof(*info) + afinfo->route_key_size, GFP_ATOMIC);
f6ebe77f HW	121	if (!info) {
	122	if (net_ratelimit())
	123	printk(KERN_ERR "OOM queueing packet %p\n",
394f545d	124	skb);
585426fd	125	rcu_read_unlock();
394f545d	126	kfree_skb(skb);
f6ebe77f HW	127	return 1;
	128	}
	129
601e68e1	130	*info = (struct nf_info) {
f6ebe77f HW	131	(struct nf_hook_ops *)elem, pf, hook, indev, outdev, okfn };
	132
	133	/* If it's going away, ignore hook. */
	134	if (!try_module_get(info->elem->owner)) {
585426fd	135	rcu_read_unlock();
f6ebe77f HW	136	kfree(info);
	137	return 0;
	138	}
	139
	140	/* Bump dev refs so they don't vanish while packet is out */
	141	if (indev) dev_hold(indev);
	142	if (outdev) dev_hold(outdev);
	143
	144	#ifdef CONFIG_BRIDGE_NETFILTER
394f545d PM	145	if (skb->nf_bridge) {
394f545d PM	146	physindev = skb->nf_bridge->physindev;
f6ebe77f	147	if (physindev) dev_hold(physindev);
394f545d	148	physoutdev = skb->nf_bridge->physoutdev;
f6ebe77f HW	149	if (physoutdev) dev_hold(physoutdev);
	150	}
	151	#endif
394f545d	152	afinfo->saveroute(skb, info);
585426fd	153	status = qh->outfn(skb, info, queuenum, qh->data);
f6ebe77f	154
585426fd	155	rcu_read_unlock();
f6ebe77f HW	156
	157	if (status < 0) {
	158	/* James M doesn't say fuck enough. */
	159	if (indev) dev_put(indev);
	160	if (outdev) dev_put(outdev);
	161	#ifdef CONFIG_BRIDGE_NETFILTER
	162	if (physindev) dev_put(physindev);
	163	if (physoutdev) dev_put(physoutdev);
	164	#endif
	165	module_put(info->elem->owner);
	166	kfree(info);
394f545d	167	kfree_skb(skb);
f6ebe77f HW	168
	169	return 1;
	170	}
	171
	172	return 1;
	173	}
	174
394f545d PM	175	int nf_queue(struct sk_buff *skb,
	176	struct list_head *elem,
	177	int pf, unsigned int hook,
	178	struct net_device *indev,
	179	struct net_device *outdev,
	180	int (okfn)(struct sk_buff ),
	181	unsigned int queuenum)
	182	{
	183	struct sk_buff *segs;
	184
	185	if (!skb_is_gso(skb))
	186	return __nf_queue(skb, elem, pf, hook, indev, outdev, okfn,
	187	queuenum);
	188
	189	switch (pf) {
	190	case AF_INET:
	191	skb->protocol = htons(ETH_P_IP);
	192	break;
	193	case AF_INET6:
	194	skb->protocol = htons(ETH_P_IPV6);
	195	break;
	196	}
	197
	198	segs = skb_gso_segment(skb, 0);
	199	kfree_skb(skb);
	200	if (unlikely(IS_ERR(segs)))
	201	return 1;
	202
	203	do {
	204	struct sk_buff *nskb = segs->next;
	205
	206	segs->next = NULL;
	207	if (!__nf_queue(segs, elem, pf, hook, indev, outdev, okfn,
	208	queuenum))
	209	kfree_skb(segs);
	210	segs = nskb;
	211	} while (segs);
	212	return 1;
	213	}
	214
f6ebe77f HW	215	void nf_reinject(struct sk_buff skb, struct nf_info info,
	216	unsigned int verdict)
	217	{
	218	struct list_head *elem = &info->elem->list;
	219	struct list_head *i;
bce8032e	220	struct nf_afinfo *afinfo;
f6ebe77f HW	221
	222	rcu_read_lock();
	223
	224	/* Release those devices we held, or Alexey will kill me. */
	225	if (info->indev) dev_put(info->indev);
	226	if (info->outdev) dev_put(info->outdev);
	227	#ifdef CONFIG_BRIDGE_NETFILTER
	228	if (skb->nf_bridge) {
	229	if (skb->nf_bridge->physindev)
	230	dev_put(skb->nf_bridge->physindev);
	231	if (skb->nf_bridge->physoutdev)
	232	dev_put(skb->nf_bridge->physoutdev);
	233	}
	234	#endif
	235
	236	/* Drop reference to owner of hook which queued us. */
	237	module_put(info->elem->owner);
	238
	239	list_for_each_rcu(i, &nf_hooks[info->pf][info->hook]) {
601e68e1 YH	240	if (i == elem)
	241	break;
	242	}
	243
45fe4dc0	244	if (i == &nf_hooks[info->pf][info->hook]) {
f6ebe77f HW	245	/* The module which sent it to userspace is gone. */
	246	NFDEBUG("%s: module disappeared, dropping packet.\n",
	247	__FUNCTION__);
	248	verdict = NF_DROP;
	249	}
	250
	251	/* Continue traversal iff userspace said ok... */
	252	if (verdict == NF_REPEAT) {
	253	elem = elem->prev;
	254	verdict = NF_ACCEPT;
	255	}
	256
7a11b984	257	if (verdict == NF_ACCEPT) {
bce8032e	258	afinfo = nf_get_afinfo(info->pf);
3db05fea	259	if (!afinfo \|\| afinfo->reroute(skb, info) < 0)
7a11b984 PM	260	verdict = NF_DROP;
	261	}
	262
f6ebe77f HW	263	if (verdict == NF_ACCEPT) {
	264	next_hook:
	265	verdict = nf_iterate(&nf_hooks[info->pf][info->hook],
3db05fea	266	skb, info->hook,
f6ebe77f HW	267	info->indev, info->outdev, &elem,
	268	info->okfn, INT_MIN);
	269	}
	270
	271	switch (verdict & NF_VERDICT_MASK) {
	272	case NF_ACCEPT:
3bc38712	273	case NF_STOP:
f6ebe77f	274	info->okfn(skb);
3bc38712	275	case NF_STOLEN:
f6ebe77f	276	break;
f6ebe77f	277	case NF_QUEUE:
394f545d PM	278	if (!__nf_queue(skb, elem, info->pf, info->hook,
	279	info->indev, info->outdev, info->okfn,
	280	verdict >> NF_VERDICT_BITS))
f6ebe77f HW	281	goto next_hook;
f6ebe77f HW	282	break;
3bc38712 PM	283	default:
3bc38712 PM	284	kfree_skb(skb);
f6ebe77f HW	285	}
f6ebe77f HW	286	rcu_read_unlock();
f6ebe77f HW	287	kfree(info);
	288	return;
	289	}
	290	EXPORT_SYMBOL(nf_reinject);
	291
bbd86b9f HW	292	#ifdef CONFIG_PROC_FS
	293	static void seq_start(struct seq_file seq, loff_t *pos)
	294	{
	295	if (*pos >= NPROTO)
	296	return NULL;
	297
	298	return pos;
	299	}
	300
	301	static void seq_next(struct seq_file s, void v, loff_t pos)
	302	{
	303	(*pos)++;
	304
	305	if (*pos >= NPROTO)
	306	return NULL;
	307
	308	return pos;
	309	}
	310
	311	static void seq_stop(struct seq_file s, void v)
	312	{
	313
	314	}
	315
	316	static int seq_show(struct seq_file s, void v)
	317	{
	318	int ret;
	319	loff_t *pos = v;
	320	struct nf_queue_handler *qh;
	321
585426fd YK	322	rcu_read_lock();
585426fd YK	323	qh = rcu_dereference(queue_handler[*pos]);
bbd86b9f HW	324	if (!qh)
	325	ret = seq_printf(s, "%2lld NONE\n", *pos);
	326	else
	327	ret = seq_printf(s, "%2lld %s\n", *pos, qh->name);
585426fd	328	rcu_read_unlock();
bbd86b9f HW	329
	330	return ret;
	331	}
	332
56b3d975	333	static const struct seq_operations nfqueue_seq_ops = {
bbd86b9f HW	334	.start = seq_start,
	335	.next = seq_next,
	336	.stop = seq_stop,
	337	.show = seq_show,
	338	};
	339
	340	static int nfqueue_open(struct inode inode, struct file file)
	341	{
	342	return seq_open(file, &nfqueue_seq_ops);
	343	}
	344
da7071d7	345	static const struct file_operations nfqueue_file_ops = {
bbd86b9f HW	346	.owner = THIS_MODULE,
	347	.open = nfqueue_open,
	348	.read = seq_read,
	349	.llseek = seq_lseek,
	350	.release = seq_release,
	351	};
	352	#endif /* PROC_FS */
	353
	354
f6ebe77f HW	355	int __init netfilter_queue_init(void)
f6ebe77f HW	356	{
bbd86b9f HW	357	#ifdef CONFIG_PROC_FS
bbd86b9f HW	358	struct proc_dir_entry *pde;
f6ebe77f	359
bbd86b9f	360	pde = create_proc_entry("nf_queue", S_IRUGO, proc_net_netfilter);
e02f7d16	361	if (!pde)
bbd86b9f	362	return -1;
bbd86b9f HW	363	pde->proc_fops = &nfqueue_file_ops;
bbd86b9f HW	364	#endif
f6ebe77f HW	365	return 0;
	366	}
	367