[linux-2.6-block.git] / net / bridge / br_forward.c

/*
 *	Forwarding decision
 *	Linux ethernet bridge
 *
 *	Authors:
 *	Lennert Buytenhek		<buytenh@gnu.org>
 *
 *	This program is free software; you can redistribute it and/or
 *	modify it under the terms of the GNU General Public License
 *	as published by the Free Software Foundation; either version
 *	2 of the License, or (at your option) any later version.
 */

#include <linux/err.h>
#include <linux/slab.h>
#include <linux/kernel.h>
#include <linux/netdevice.h>
#include <linux/netpoll.h>
#include <linux/skbuff.h>
#include <linux/if_vlan.h>
#include <linux/netfilter_bridge.h>
#include "br_private.h"

static int deliver_clone(const struct net_bridge_port *prev,
			 struct sk_buff *skb,
			 void (*__packet_hook)(const struct net_bridge_port *p,
					       struct sk_buff *skb));

/* Don't forward packets to originating port or forwarding diasabled */
static inline int should_deliver(const struct net_bridge_port *p,
				 const struct sk_buff *skb)
{
	return (((p->flags & BR_HAIRPIN_MODE) || skb->dev != p->dev) &&
		p->state == BR_STATE_FORWARDING);
}

static inline unsigned int packet_length(const struct sk_buff *skb)
{
	return skb->len - (skb->protocol == htons(ETH_P_8021Q) ? VLAN_HLEN : 0);
}

int br_dev_queue_push_xmit(struct sk_buff *skb)
{
	/* ip_fragment doesn't copy the MAC header */
	if (nf_bridge_maybe_copy_header(skb) ||
	    (packet_length(skb) > skb->dev->mtu && !skb_is_gso(skb))) {
		kfree_skb(skb);
	} else {
		skb_push(skb, ETH_HLEN);
		dev_queue_xmit(skb);
	}

	return 0;
}

int br_forward_finish(struct sk_buff *skb)
{
	return NF_HOOK(NFPROTO_BRIDGE, NF_BR_POST_ROUTING, skb, NULL, skb->dev,
		       br_dev_queue_push_xmit);

}

static void __br_deliver(const struct net_bridge_port *to, struct sk_buff *skb)
{
	skb->dev = to->dev;

	if (unlikely(netpoll_tx_running(to->dev))) {
		if (packet_length(skb) > skb->dev->mtu && !skb_is_gso(skb))
			kfree_skb(skb);
		else {
			skb_push(skb, ETH_HLEN);
			br_netpoll_send_skb(to, skb);
		}
		return;
	}

	NF_HOOK(NFPROTO_BRIDGE, NF_BR_LOCAL_OUT, skb, NULL, skb->dev,
		br_forward_finish);
}

static void __br_forward(const struct net_bridge_port *to, struct sk_buff *skb)
{
	struct net_device *indev;

	if (skb_warn_if_lro(skb)) {
		kfree_skb(skb);
		return;
	}

	indev = skb->dev;
	skb->dev = to->dev;
	skb_forward_csum(skb);

	NF_HOOK(NFPROTO_BRIDGE, NF_BR_FORWARD, skb, indev, skb->dev,
		br_forward_finish);
}

/* called with rcu_read_lock */
void br_deliver(const struct net_bridge_port *to, struct sk_buff *skb)
{
	if (to && should_deliver(to, skb)) {
		__br_deliver(to, skb);
		return;
	}

	kfree_skb(skb);
}

/* called with rcu_read_lock */
void br_forward(const struct net_bridge_port *to, struct sk_buff *skb, struct sk_buff *skb0)
{
	if (should_deliver(to, skb)) {
		if (skb0)
			deliver_clone(to, skb, __br_forward);
		else
			__br_forward(to, skb);
		return;
	}

	if (!skb0)
		kfree_skb(skb);
}

static int deliver_clone(const struct net_bridge_port *prev,
			 struct sk_buff *skb,
			 void (*__packet_hook)(const struct net_bridge_port *p,
					       struct sk_buff *skb))
{
	struct net_device *dev = BR_INPUT_SKB_CB(skb)->brdev;

	skb = skb_clone(skb, GFP_ATOMIC);
	if (!skb) {
		dev->stats.tx_dropped++;
		return -ENOMEM;
	}

	__packet_hook(prev, skb);
	return 0;
}

static struct net_bridge_port *maybe_deliver(
	struct net_bridge_port *prev, struct net_bridge_port *p,
	struct sk_buff *skb,
	void (*__packet_hook)(const struct net_bridge_port *p,
			      struct sk_buff *skb))
{
	int err;

	if (!should_deliver(p, skb))
		return prev;

	if (!prev)
		goto out;

	err = deliver_clone(prev, skb, __packet_hook);
	if (err)
		return ERR_PTR(err);

out:
	return p;
}

/* called under bridge lock */
static void br_flood(struct net_bridge *br, struct sk_buff *skb,
		     struct sk_buff *skb0,
		     void (*__packet_hook)(const struct net_bridge_port *p,
					   struct sk_buff *skb))
{
	struct net_bridge_port *p;
	struct net_bridge_port *prev;

	prev = NULL;

	list_for_each_entry_rcu(p, &br->port_list, list) {
		prev = maybe_deliver(prev, p, skb, __packet_hook);
		if (IS_ERR(prev))
			goto out;
	}

	if (!prev)
		goto out;

	if (skb0)
		deliver_clone(prev, skb, __packet_hook);
	else
		__packet_hook(prev, skb);
	return;

out:
	if (!skb0)
		kfree_skb(skb);
}


/* called with rcu_read_lock */
void br_flood_deliver(struct net_bridge *br, struct sk_buff *skb)
{
	br_flood(br, skb, NULL, __br_deliver);
}

/* called under bridge lock */
void br_flood_forward(struct net_bridge *br, struct sk_buff *skb,
		      struct sk_buff *skb2)
{
	br_flood(br, skb, skb2, __br_forward);
}

#ifdef CONFIG_BRIDGE_IGMP_SNOOPING
/* called with rcu_read_lock */
static void br_multicast_flood(struct net_bridge_mdb_entry *mdst,
			       struct sk_buff *skb, struct sk_buff *skb0,
			       void (*__packet_hook)(
					const struct net_bridge_port *p,
					struct sk_buff *skb))
{
	struct net_device *dev = BR_INPUT_SKB_CB(skb)->brdev;
	struct net_bridge *br = netdev_priv(dev);
	struct net_bridge_port *prev = NULL;
	struct net_bridge_port_group *p;
	struct hlist_node *rp;

	rp = rcu_dereference(hlist_first_rcu(&br->router_list));
	p = mdst ? rcu_dereference(mdst->ports) : NULL;
	while (p || rp) {
		struct net_bridge_port *port, *lport, *rport;

		lport = p ? p->port : NULL;
		rport = rp ? hlist_entry(rp, struct net_bridge_port, rlist) :
			     NULL;

		port = (unsigned long)lport > (unsigned long)rport ?
		       lport : rport;

		prev = maybe_deliver(prev, port, skb, __packet_hook);
		if (IS_ERR(prev))
			goto out;

		if ((unsigned long)lport >= (unsigned long)port)
			p = rcu_dereference(p->next);
		if ((unsigned long)rport >= (unsigned long)port)
			rp = rcu_dereference(hlist_next_rcu(rp));
	}

	if (!prev)
		goto out;

	if (skb0)
		deliver_clone(prev, skb, __packet_hook);
	else
		__packet_hook(prev, skb);
	return;

out:
	if (!skb0)
		kfree_skb(skb);
}

/* called with rcu_read_lock */
void br_multicast_deliver(struct net_bridge_mdb_entry *mdst,
			  struct sk_buff *skb)
{
	br_multicast_flood(mdst, skb, NULL, __br_deliver);
}

/* called with rcu_read_lock */
void br_multicast_forward(struct net_bridge_mdb_entry *mdst,
			  struct sk_buff *skb, struct sk_buff *skb2)
{
	br_multicast_flood(mdst, skb, skb2, __br_forward);
}
#endif
Commit	Line	Data
1da177e4 LT	1	/*
	2	* Forwarding decision
	3	* Linux ethernet bridge
	4	*
	5	* Authors:
	6	* Lennert Buytenhek <buytenh@gnu.org>
	7	*
1da177e4 LT	8	* This program is free software; you can redistribute it and/or
	9	* modify it under the terms of the GNU General Public License
	10	* as published by the Free Software Foundation; either version
	11	* 2 of the License, or (at your option) any later version.
	12	*/
	13
025d89c2	14	#include <linux/err.h>
5a0e3ad6	15	#include <linux/slab.h>
1da177e4 LT	16	#include <linux/kernel.h>
1da177e4 LT	17	#include <linux/netdevice.h>
c06ee961	18	#include <linux/netpoll.h>
1da177e4	19	#include <linux/skbuff.h>
85ca719e	20	#include <linux/if_vlan.h>
1da177e4 LT	21	#include <linux/netfilter_bridge.h>
	22	#include "br_private.h"
	23
87faf3cc DM	24	static int deliver_clone(const struct net_bridge_port *prev,
87faf3cc DM	25	struct sk_buff *skb,
7f7708f0 MB	26	void (__packet_hook)(const struct net_bridge_port p,
	27	struct sk_buff *skb));
	28
9ef513be	29	/* Don't forward packets to originating port or forwarding diasabled */
9d6f229f	30	static inline int should_deliver(const struct net_bridge_port *p,
1da177e4 LT	31	const struct sk_buff *skb)
1da177e4 LT	32	{
3982d3d2 FA	33	return (((p->flags & BR_HAIRPIN_MODE) \|\| skb->dev != p->dev) &&
3982d3d2 FA	34	p->state == BR_STATE_FORWARDING);
1da177e4 LT	35	}
1da177e4 LT	36
95c96174	37	static inline unsigned int packet_length(const struct sk_buff *skb)
85ca719e SH	38	{
	39	return skb->len - (skb->protocol == htons(ETH_P_8021Q) ? VLAN_HLEN : 0);
	40	}
	41
1da177e4 LT	42	int br_dev_queue_push_xmit(struct sk_buff *skb)
1da177e4 LT	43	{
f88de8de CG	44	/* ip_fragment doesn't copy the MAC header */
	45	if (nf_bridge_maybe_copy_header(skb) \|\|
	46	(packet_length(skb) > skb->dev->mtu && !skb_is_gso(skb))) {
1da177e4	47	kfree_skb(skb);
f88de8de CG	48	} else {
	49	skb_push(skb, ETH_HLEN);
	50	dev_queue_xmit(skb);
1da177e4 LT	51	}
	52
	53	return 0;
	54	}
	55
	56	int br_forward_finish(struct sk_buff *skb)
	57	{
713aefa3	58	return NF_HOOK(NFPROTO_BRIDGE, NF_BR_POST_ROUTING, skb, NULL, skb->dev,
9ef513be	59	br_dev_queue_push_xmit);
1da177e4	60
1da177e4 LT	61	}
	62
	63	static void __br_deliver(const struct net_bridge_port to, struct sk_buff skb)
	64	{
	65	skb->dev = to->dev;
91d2c34a HX	66
	67	if (unlikely(netpoll_tx_running(to->dev))) {
	68	if (packet_length(skb) > skb->dev->mtu && !skb_is_gso(skb))
	69	kfree_skb(skb);
	70	else {
	71	skb_push(skb, ETH_HLEN);
	72	br_netpoll_send_skb(to, skb);
	73	}
	74	return;
	75	}
	76
713aefa3 JE	77	NF_HOOK(NFPROTO_BRIDGE, NF_BR_LOCAL_OUT, skb, NULL, skb->dev,
713aefa3 JE	78	br_forward_finish);
1da177e4 LT	79	}
	80
	81	static void __br_forward(const struct net_bridge_port to, struct sk_buff skb)
	82	{
	83	struct net_device *indev;
	84
4906f998 HX	85	if (skb_warn_if_lro(skb)) {
	86	kfree_skb(skb);
	87	return;
	88	}
	89
1da177e4 LT	90	indev = skb->dev;
1da177e4 LT	91	skb->dev = to->dev;
35fc92a9	92	skb_forward_csum(skb);
1da177e4	93
713aefa3 JE	94	NF_HOOK(NFPROTO_BRIDGE, NF_BR_FORWARD, skb, indev, skb->dev,
713aefa3 JE	95	br_forward_finish);
1da177e4 LT	96	}
	97
	98	/* called with rcu_read_lock */
	99	void br_deliver(const struct net_bridge_port to, struct sk_buff skb)
	100	{
43598813	101	if (to && should_deliver(to, skb)) {
1da177e4 LT	102	__br_deliver(to, skb);
	103	return;
	104	}
	105
	106	kfree_skb(skb);
	107	}
	108
	109	/* called with rcu_read_lock */
7f7708f0	110	void br_forward(const struct net_bridge_port to, struct sk_buff skb, struct sk_buff *skb0)
1da177e4	111	{
4906f998	112	if (should_deliver(to, skb)) {
7f7708f0 MB	113	if (skb0)
	114	deliver_clone(to, skb, __br_forward);
	115	else
	116	__br_forward(to, skb);
1da177e4 LT	117	return;
	118	}
	119
7f7708f0 MB	120	if (!skb0)
7f7708f0 MB	121	kfree_skb(skb);
1da177e4 LT	122	}
1da177e4 LT	123
87faf3cc DM	124	static int deliver_clone(const struct net_bridge_port *prev,
87faf3cc DM	125	struct sk_buff *skb,
025d89c2 HX	126	void (__packet_hook)(const struct net_bridge_port p,
	127	struct sk_buff *skb))
	128	{
fed396a5 HX	129	struct net_device *dev = BR_INPUT_SKB_CB(skb)->brdev;
fed396a5 HX	130
025d89c2 HX	131	skb = skb_clone(skb, GFP_ATOMIC);
025d89c2 HX	132	if (!skb) {
025d89c2 HX	133	dev->stats.tx_dropped++;
	134	return -ENOMEM;
	135	}
	136
	137	__packet_hook(prev, skb);
	138	return 0;
	139	}
	140
	141	static struct net_bridge_port *maybe_deliver(
	142	struct net_bridge_port prev, struct net_bridge_port p,
	143	struct sk_buff *skb,
	144	void (__packet_hook)(const struct net_bridge_port p,
	145	struct sk_buff *skb))
	146	{
	147	int err;
	148
	149	if (!should_deliver(p, skb))
	150	return prev;
	151
	152	if (!prev)
	153	goto out;
	154
	155	err = deliver_clone(prev, skb, __packet_hook);
	156	if (err)
	157	return ERR_PTR(err);
	158
	159	out:
	160	return p;
	161	}
	162
1da177e4	163	/* called under bridge lock */
e081e1e3	164	static void br_flood(struct net_bridge br, struct sk_buff skb,
b33084be HX	165	struct sk_buff *skb0,
	166	void (__packet_hook)(const struct net_bridge_port p,
	167	struct sk_buff *skb))
1da177e4 LT	168	{
	169	struct net_bridge_port *p;
	170	struct net_bridge_port *prev;
	171
1da177e4 LT	172	prev = NULL;
	173
	174	list_for_each_entry_rcu(p, &br->port_list, list) {
025d89c2 HX	175	prev = maybe_deliver(prev, p, skb, __packet_hook);
	176	if (IS_ERR(prev))
	177	goto out;
1da177e4 LT	178	}
1da177e4 LT	179
b33084be HX	180	if (!prev)
	181	goto out;
	182
025d89c2 HX	183	if (skb0)
	184	deliver_clone(prev, skb, __packet_hook);
	185	else
	186	__packet_hook(prev, skb);
b33084be	187	return;
1da177e4	188
b33084be HX	189	out:
	190	if (!skb0)
	191	kfree_skb(skb);
1da177e4 LT	192	}
	193
	194
	195	/* called with rcu_read_lock */
e081e1e3	196	void br_flood_deliver(struct net_bridge br, struct sk_buff skb)
1da177e4	197	{
b33084be	198	br_flood(br, skb, NULL, __br_deliver);
1da177e4 LT	199	}
	200
	201	/* called under bridge lock */
b33084be HX	202	void br_flood_forward(struct net_bridge br, struct sk_buff skb,
b33084be HX	203	struct sk_buff *skb2)
1da177e4	204	{
b33084be	205	br_flood(br, skb, skb2, __br_forward);
1da177e4	206	}
5cb5e947 HX	207
	208	#ifdef CONFIG_BRIDGE_IGMP_SNOOPING
	209	/* called with rcu_read_lock */
	210	static void br_multicast_flood(struct net_bridge_mdb_entry *mdst,
	211	struct sk_buff skb, struct sk_buff skb0,
	212	void (*__packet_hook)(
	213	const struct net_bridge_port *p,
	214	struct sk_buff *skb))
	215	{
	216	struct net_device *dev = BR_INPUT_SKB_CB(skb)->brdev;
	217	struct net_bridge *br = netdev_priv(dev);
afe0159d	218	struct net_bridge_port *prev = NULL;
5cb5e947 HX	219	struct net_bridge_port_group *p;
	220	struct hlist_node *rp;
	221
e8051688	222	rp = rcu_dereference(hlist_first_rcu(&br->router_list));
83f6a740	223	p = mdst ? rcu_dereference(mdst->ports) : NULL;
5cb5e947	224	while (p \|\| rp) {
afe0159d	225	struct net_bridge_port port, lport, *rport;
afe0159d	226
5cb5e947 HX	227	lport = p ? p->port : NULL;
	228	rport = rp ? hlist_entry(rp, struct net_bridge_port, rlist) :
	229	NULL;
	230
	231	port = (unsigned long)lport > (unsigned long)rport ?
	232	lport : rport;
	233
	234	prev = maybe_deliver(prev, port, skb, __packet_hook);
	235	if (IS_ERR(prev))
	236	goto out;
	237
	238	if ((unsigned long)lport >= (unsigned long)port)
83f6a740	239	p = rcu_dereference(p->next);
5cb5e947	240	if ((unsigned long)rport >= (unsigned long)port)
e8051688	241	rp = rcu_dereference(hlist_next_rcu(rp));
5cb5e947 HX	242	}
	243
	244	if (!prev)
	245	goto out;
	246
	247	if (skb0)
	248	deliver_clone(prev, skb, __packet_hook);
	249	else
	250	__packet_hook(prev, skb);
	251	return;
	252
	253	out:
	254	if (!skb0)
	255	kfree_skb(skb);
	256	}
	257
	258	/* called with rcu_read_lock */
	259	void br_multicast_deliver(struct net_bridge_mdb_entry *mdst,
	260	struct sk_buff *skb)
	261	{
	262	br_multicast_flood(mdst, skb, NULL, __br_deliver);
	263	}
	264
	265	/* called with rcu_read_lock */
	266	void br_multicast_forward(struct net_bridge_mdb_entry *mdst,
	267	struct sk_buff skb, struct sk_buff skb2)
	268	{
	269	br_multicast_flood(mdst, skb, skb2, __br_forward);
	270	}
	271	#endif