[net-next-2.6.git] / net / ipv4 / netfilter / nf_conntrack_l3proto_ipv4.c

/* (C) 1999-2001 Paul `Rusty' Russell
 * (C) 2002-2004 Netfilter Core Team <coreteam@netfilter.org>
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation.
 *
 * 16 Dec 2003: Yasuyuki Kozakai @USAGI <yasuyuki.kozakai@toshiba.co.jp>
 *	- move L3 protocol dependent part to this file.
 * 23 Mar 2004: Yasuyuki Kozakai @USAGI <yasuyuki.kozakai@toshiba.co.jp>
 *	- add get_features() to support various size of conntrack
 *	  structures.
 *
 * Derived from net/ipv4/netfilter/ip_conntrack_standalone.c
 */

#include <linux/types.h>
#include <linux/ip.h>
#include <linux/netfilter.h>
#include <linux/module.h>
#include <linux/skbuff.h>
#include <linux/icmp.h>
#include <linux/sysctl.h>
#include <net/route.h>
#include <net/ip.h>

#include <linux/netfilter_ipv4.h>
#include <net/netfilter/nf_conntrack.h>
#include <net/netfilter/nf_conntrack_helper.h>
#include <net/netfilter/nf_conntrack_l4proto.h>
#include <net/netfilter/nf_conntrack_l3proto.h>
#include <net/netfilter/nf_conntrack_core.h>
#include <net/netfilter/ipv4/nf_conntrack_ipv4.h>

#if 0
#define DEBUGP printk
#else
#define DEBUGP(format, args...)
#endif

static int ipv4_pkt_to_tuple(const struct sk_buff *skb, unsigned int nhoff,
			     struct nf_conntrack_tuple *tuple)
{
	u_int32_t _addrs[2], *ap;
	ap = skb_header_pointer(skb, nhoff + offsetof(struct iphdr, saddr),
				sizeof(u_int32_t) * 2, _addrs);
	if (ap == NULL)
		return 0;

	tuple->src.u3.ip = ap[0];
	tuple->dst.u3.ip = ap[1];

	return 1;
}

static int ipv4_invert_tuple(struct nf_conntrack_tuple *tuple,
			   const struct nf_conntrack_tuple *orig)
{
	tuple->src.u3.ip = orig->dst.u3.ip;
	tuple->dst.u3.ip = orig->src.u3.ip;

	return 1;
}

static int ipv4_print_tuple(struct seq_file *s,
			    const struct nf_conntrack_tuple *tuple)
{
	return seq_printf(s, "src=%u.%u.%u.%u dst=%u.%u.%u.%u ",
		          NIPQUAD(tuple->src.u3.ip),
			  NIPQUAD(tuple->dst.u3.ip));
}

static int ipv4_print_conntrack(struct seq_file *s,
				const struct nf_conn *conntrack)
{
	return 0;
}

/* Returns new sk_buff, or NULL */
static struct sk_buff *
nf_ct_ipv4_gather_frags(struct sk_buff *skb, u_int32_t user)
{
	skb_orphan(skb);

        local_bh_disable();
        skb = ip_defrag(skb, user);
        local_bh_enable();

        if (skb)
		ip_send_check(skb->nh.iph);

        return skb;
}

static int
ipv4_prepare(struct sk_buff **pskb, unsigned int hooknum, unsigned int *dataoff,
	     u_int8_t *protonum)
{
	/* Never happen */
	if ((*pskb)->nh.iph->frag_off & htons(IP_OFFSET)) {
		if (net_ratelimit()) {
			printk(KERN_ERR "ipv4_prepare: Frag of proto %u (hook=%u)\n",
			(*pskb)->nh.iph->protocol, hooknum);
		}
		return -NF_DROP;
	}

	*dataoff = (*pskb)->nh.raw - (*pskb)->data + (*pskb)->nh.iph->ihl*4;
	*protonum = (*pskb)->nh.iph->protocol;

	return NF_ACCEPT;
}

int nat_module_is_loaded = 0;
static u_int32_t ipv4_get_features(const struct nf_conntrack_tuple *tuple)
{
	if (nat_module_is_loaded)
		return NF_CT_F_NAT;

	return NF_CT_F_BASIC;
}

static unsigned int ipv4_confirm(unsigned int hooknum,
				 struct sk_buff **pskb,
				 const struct net_device *in,
				 const struct net_device *out,
				 int (*okfn)(struct sk_buff *))
{
	/* We've seen it coming out the other side: confirm it */
	return nf_conntrack_confirm(pskb);
}

static unsigned int ipv4_conntrack_help(unsigned int hooknum,
				      struct sk_buff **pskb,
				      const struct net_device *in,
				      const struct net_device *out,
				      int (*okfn)(struct sk_buff *))
{
	struct nf_conn *ct;
	enum ip_conntrack_info ctinfo;
	struct nf_conn_help *help;

	/* This is where we call the helper: as the packet goes out. */
	ct = nf_ct_get(*pskb, &ctinfo);
	if (!ct || ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY)
		return NF_ACCEPT;

	help = nfct_help(ct);
	if (!help || !help->helper)
		return NF_ACCEPT;

	return help->helper->help(pskb,
			       (*pskb)->nh.raw - (*pskb)->data
					       + (*pskb)->nh.iph->ihl*4,
			       ct, ctinfo);
}

static unsigned int ipv4_conntrack_defrag(unsigned int hooknum,
					  struct sk_buff **pskb,
					  const struct net_device *in,
					  const struct net_device *out,
					  int (*okfn)(struct sk_buff *))
{
#if !defined(CONFIG_IP_NF_NAT) && !defined(CONFIG_IP_NF_NAT_MODULE)
	/* Previously seen (loopback)?  Ignore.  Do this before
	   fragment check. */
	if ((*pskb)->nfct)
		return NF_ACCEPT;
#endif

	/* Gather fragments. */
	if ((*pskb)->nh.iph->frag_off & htons(IP_MF|IP_OFFSET)) {
		*pskb = nf_ct_ipv4_gather_frags(*pskb,
						hooknum == NF_IP_PRE_ROUTING ?
						IP_DEFRAG_CONNTRACK_IN :
						IP_DEFRAG_CONNTRACK_OUT);
		if (!*pskb)
			return NF_STOLEN;
	}
	return NF_ACCEPT;
}

static unsigned int ipv4_conntrack_in(unsigned int hooknum,
				      struct sk_buff **pskb,
				      const struct net_device *in,
				      const struct net_device *out,
				      int (*okfn)(struct sk_buff *))
{
	return nf_conntrack_in(PF_INET, hooknum, pskb);
}

static unsigned int ipv4_conntrack_local(unsigned int hooknum,
				         struct sk_buff **pskb,
				         const struct net_device *in,
				         const struct net_device *out,
				         int (*okfn)(struct sk_buff *))
{
	/* root is playing with raw sockets. */
	if ((*pskb)->len < sizeof(struct iphdr)
	    || (*pskb)->nh.iph->ihl * 4 < sizeof(struct iphdr)) {
		if (net_ratelimit())
			printk("ipt_hook: happy cracking.\n");
		return NF_ACCEPT;
	}
	return nf_conntrack_in(PF_INET, hooknum, pskb);
}

/* Connection tracking may drop packets, but never alters them, so
   make it the first hook. */
static struct nf_hook_ops ipv4_conntrack_ops[] = {
	{
		.hook		= ipv4_conntrack_defrag,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_PRE_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK_DEFRAG,
	},
	{
		.hook		= ipv4_conntrack_in,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_PRE_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK,
	},
	{
		.hook           = ipv4_conntrack_defrag,
		.owner          = THIS_MODULE,
		.pf             = PF_INET,
		.hooknum        = NF_IP_LOCAL_OUT,
		.priority       = NF_IP_PRI_CONNTRACK_DEFRAG,
	},
	{
		.hook		= ipv4_conntrack_local,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_LOCAL_OUT,
		.priority	= NF_IP_PRI_CONNTRACK,
	},
	{
		.hook		= ipv4_conntrack_help,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_POST_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK_HELPER,
	},
	{
		.hook		= ipv4_conntrack_help,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_LOCAL_IN,
		.priority	= NF_IP_PRI_CONNTRACK_HELPER,
	},
	{
		.hook		= ipv4_confirm,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_POST_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK_CONFIRM,
	},
	{
		.hook		= ipv4_confirm,
		.owner		= THIS_MODULE,
		.pf		= PF_INET,
		.hooknum	= NF_IP_LOCAL_IN,
		.priority	= NF_IP_PRI_CONNTRACK_CONFIRM,
	},
};

/* Fast function for those who don't want to parse /proc (and I don't
   blame them). */
/* Reversing the socket's dst/src point of view gives us the reply
   mapping. */
static int
getorigdst(struct sock *sk, int optval, void __user *user, int *len)
{
	struct inet_sock *inet = inet_sk(sk);
	struct nf_conntrack_tuple_hash *h;
	struct nf_conntrack_tuple tuple;
	
	NF_CT_TUPLE_U_BLANK(&tuple);
	tuple.src.u3.ip = inet->rcv_saddr;
	tuple.src.u.tcp.port = inet->sport;
	tuple.dst.u3.ip = inet->daddr;
	tuple.dst.u.tcp.port = inet->dport;
	tuple.src.l3num = PF_INET;
	tuple.dst.protonum = IPPROTO_TCP;

	/* We only do TCP at the moment: is there a better way? */
	if (strcmp(sk->sk_prot->name, "TCP")) {
		DEBUGP("SO_ORIGINAL_DST: Not a TCP socket\n");
		return -ENOPROTOOPT;
	}

	if ((unsigned int) *len < sizeof(struct sockaddr_in)) {
		DEBUGP("SO_ORIGINAL_DST: len %u not %u\n",
		       *len, sizeof(struct sockaddr_in));
		return -EINVAL;
	}

	h = nf_conntrack_find_get(&tuple, NULL);
	if (h) {
		struct sockaddr_in sin;
		struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h);

		sin.sin_family = AF_INET;
		sin.sin_port = ct->tuplehash[IP_CT_DIR_ORIGINAL]
			.tuple.dst.u.tcp.port;
		sin.sin_addr.s_addr = ct->tuplehash[IP_CT_DIR_ORIGINAL]
			.tuple.dst.u3.ip;
		memset(sin.sin_zero, 0, sizeof(sin.sin_zero));

		DEBUGP("SO_ORIGINAL_DST: %u.%u.%u.%u %u\n",
		       NIPQUAD(sin.sin_addr.s_addr), ntohs(sin.sin_port));
		nf_ct_put(ct);
		if (copy_to_user(user, &sin, sizeof(sin)) != 0)
			return -EFAULT;
		else
			return 0;
	}
	DEBUGP("SO_ORIGINAL_DST: Can't find %u.%u.%u.%u/%u-%u.%u.%u.%u/%u.\n",
	       NIPQUAD(tuple.src.u3.ip), ntohs(tuple.src.u.tcp.port),
	       NIPQUAD(tuple.dst.u3.ip), ntohs(tuple.dst.u.tcp.port));
	return -ENOENT;
}

#if defined(CONFIG_NF_CT_NETLINK) || \
    defined(CONFIG_NF_CT_NETLINK_MODULE)

#include <linux/netfilter/nfnetlink.h>
#include <linux/netfilter/nfnetlink_conntrack.h>

static int ipv4_tuple_to_nfattr(struct sk_buff *skb,
				const struct nf_conntrack_tuple *tuple)
{
	NFA_PUT(skb, CTA_IP_V4_SRC, sizeof(u_int32_t),
		&tuple->src.u3.ip);
	NFA_PUT(skb, CTA_IP_V4_DST, sizeof(u_int32_t),
		&tuple->dst.u3.ip);
	return 0;

nfattr_failure:
	return -1;
}

static const size_t cta_min_ip[CTA_IP_MAX] = {
	[CTA_IP_V4_SRC-1]       = sizeof(u_int32_t),
	[CTA_IP_V4_DST-1]       = sizeof(u_int32_t),
};

static int ipv4_nfattr_to_tuple(struct nfattr *tb[],
				struct nf_conntrack_tuple *t)
{
	if (!tb[CTA_IP_V4_SRC-1] || !tb[CTA_IP_V4_DST-1])
		return -EINVAL;

	if (nfattr_bad_size(tb, CTA_IP_MAX, cta_min_ip))
		return -EINVAL;

	t->src.u3.ip =
		*(u_int32_t *)NFA_DATA(tb[CTA_IP_V4_SRC-1]);
	t->dst.u3.ip =
		*(u_int32_t *)NFA_DATA(tb[CTA_IP_V4_DST-1]);

	return 0;
}
#endif

static struct nf_sockopt_ops so_getorigdst = {
	.pf		= PF_INET,
	.get_optmin	= SO_ORIGINAL_DST,
	.get_optmax	= SO_ORIGINAL_DST+1,
	.get		= &getorigdst,
};

struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv4 = {
	.l3proto	 = PF_INET,
	.name		 = "ipv4",
	.pkt_to_tuple	 = ipv4_pkt_to_tuple,
	.invert_tuple	 = ipv4_invert_tuple,
	.print_tuple	 = ipv4_print_tuple,
	.print_conntrack = ipv4_print_conntrack,
	.prepare	 = ipv4_prepare,
	.get_features	 = ipv4_get_features,
#if defined(CONFIG_NF_CT_NETLINK) || \
    defined(CONFIG_NF_CT_NETLINK_MODULE)
	.tuple_to_nfattr = ipv4_tuple_to_nfattr,
	.nfattr_to_tuple = ipv4_nfattr_to_tuple,
#endif
	.me		 = THIS_MODULE,
};

MODULE_ALIAS("nf_conntrack-" __stringify(AF_INET));
MODULE_LICENSE("GPL");

static int __init nf_conntrack_l3proto_ipv4_init(void)
{
	int ret = 0;

	need_conntrack();

	ret = nf_register_sockopt(&so_getorigdst);
	if (ret < 0) {
		printk(KERN_ERR "Unable to register netfilter socket option\n");
		return ret;
	}

	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_tcp4);
	if (ret < 0) {
		printk("nf_conntrack_ipv4: can't register tcp.\n");
		goto cleanup_sockopt;
	}

	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_udp4);
	if (ret < 0) {
		printk("nf_conntrack_ipv4: can't register udp.\n");
		goto cleanup_tcp;
	}

	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_icmp);
	if (ret < 0) {
		printk("nf_conntrack_ipv4: can't register icmp.\n");
		goto cleanup_udp;
	}

	ret = nf_conntrack_l3proto_register(&nf_conntrack_l3proto_ipv4);
	if (ret < 0) {
		printk("nf_conntrack_ipv4: can't register ipv4\n");
		goto cleanup_icmp;
	}

	ret = nf_register_hooks(ipv4_conntrack_ops,
				ARRAY_SIZE(ipv4_conntrack_ops));
	if (ret < 0) {
		printk("nf_conntrack_ipv4: can't register hooks.\n");
		goto cleanup_ipv4;
	}
	return ret;

 cleanup_ipv4:
	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
 cleanup_icmp:
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
 cleanup_udp:
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
 cleanup_tcp:
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
 cleanup_sockopt:
	nf_unregister_sockopt(&so_getorigdst);
	return ret;
}

static void __exit nf_conntrack_l3proto_ipv4_fini(void)
{
	synchronize_net();
	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
	nf_unregister_sockopt(&so_getorigdst);
}

module_init(nf_conntrack_l3proto_ipv4_init);
module_exit(nf_conntrack_l3proto_ipv4_fini);

EXPORT_SYMBOL(nf_ct_ipv4_gather_frags);
Commit	Line	Data
9fb9cbb1 YK	1	/* (C) 1999-2001 Paul `Rusty' Russell
	2	* (C) 2002-2004 Netfilter Core Team <coreteam@netfilter.org>
	3	*
	4	* This program is free software; you can redistribute it and/or modify
	5	* it under the terms of the GNU General Public License version 2 as
	6	* published by the Free Software Foundation.
	7	*
	8	* 16 Dec 2003: Yasuyuki Kozakai @USAGI <yasuyuki.kozakai@toshiba.co.jp>
	9	* - move L3 protocol dependent part to this file.
	10	* 23 Mar 2004: Yasuyuki Kozakai @USAGI <yasuyuki.kozakai@toshiba.co.jp>
	11	* - add get_features() to support various size of conntrack
	12	* structures.
	13	*
	14	* Derived from net/ipv4/netfilter/ip_conntrack_standalone.c
	15	*/
	16
9fb9cbb1 YK	17	#include <linux/types.h>
	18	#include <linux/ip.h>
	19	#include <linux/netfilter.h>
	20	#include <linux/module.h>
	21	#include <linux/skbuff.h>
	22	#include <linux/icmp.h>
	23	#include <linux/sysctl.h>
0ae2cfe7	24	#include <net/route.h>
9fb9cbb1 YK	25	#include <net/ip.h>
	26
	27	#include <linux/netfilter_ipv4.h>
	28	#include <net/netfilter/nf_conntrack.h>
	29	#include <net/netfilter/nf_conntrack_helper.h>
605dcad6	30	#include <net/netfilter/nf_conntrack_l4proto.h>
9fb9cbb1 YK	31	#include <net/netfilter/nf_conntrack_l3proto.h>
	32	#include <net/netfilter/nf_conntrack_core.h>
	33	#include <net/netfilter/ipv4/nf_conntrack_ipv4.h>
	34
	35	#if 0
	36	#define DEBUGP printk
	37	#else
	38	#define DEBUGP(format, args...)
	39	#endif
	40
9fb9cbb1 YK	41	static int ipv4_pkt_to_tuple(const struct sk_buff *skb, unsigned int nhoff,
	42	struct nf_conntrack_tuple *tuple)
	43	{
	44	u_int32_t _addrs[2], *ap;
	45	ap = skb_header_pointer(skb, nhoff + offsetof(struct iphdr, saddr),
	46	sizeof(u_int32_t) * 2, _addrs);
	47	if (ap == NULL)
	48	return 0;
	49
	50	tuple->src.u3.ip = ap[0];
	51	tuple->dst.u3.ip = ap[1];
	52
	53	return 1;
	54	}
	55
	56	static int ipv4_invert_tuple(struct nf_conntrack_tuple *tuple,
	57	const struct nf_conntrack_tuple *orig)
	58	{
	59	tuple->src.u3.ip = orig->dst.u3.ip;
	60	tuple->dst.u3.ip = orig->src.u3.ip;
	61
	62	return 1;
	63	}
	64
	65	static int ipv4_print_tuple(struct seq_file *s,
	66	const struct nf_conntrack_tuple *tuple)
	67	{
	68	return seq_printf(s, "src=%u.%u.%u.%u dst=%u.%u.%u.%u ",
	69	NIPQUAD(tuple->src.u3.ip),
	70	NIPQUAD(tuple->dst.u3.ip));
	71	}
	72
	73	static int ipv4_print_conntrack(struct seq_file *s,
	74	const struct nf_conn *conntrack)
	75	{
	76	return 0;
	77	}
	78
	79	/* Returns new sk_buff, or NULL */
	80	static struct sk_buff *
	81	nf_ct_ipv4_gather_frags(struct sk_buff *skb, u_int32_t user)
	82	{
	83	skb_orphan(skb);
	84
	85	local_bh_disable();
	86	skb = ip_defrag(skb, user);
	87	local_bh_enable();
	88
	89	if (skb)
	90	ip_send_check(skb->nh.iph);
	91
	92	return skb;
	93	}
	94
	95	static int
	96	ipv4_prepare(struct sk_buff *pskb, unsigned int hooknum, unsigned int dataoff,
	97	u_int8_t *protonum)
	98	{
	99	/* Never happen */
	100	if ((*pskb)->nh.iph->frag_off & htons(IP_OFFSET)) {
	101	if (net_ratelimit()) {
	102	printk(KERN_ERR "ipv4_prepare: Frag of proto %u (hook=%u)\n",
	103	(*pskb)->nh.iph->protocol, hooknum);
	104	}
105	return -NF_DROP;
106	}
107
108	dataoff = (pskb)->nh.raw - (pskb)->data + (pskb)->nh.iph->ihl*4;
109	protonum = (pskb)->nh.iph->protocol;
110
111	return NF_ACCEPT;
112	}
113
114	int nat_module_is_loaded = 0;
115	static u_int32_t ipv4_get_features(const struct nf_conntrack_tuple *tuple)
116	{
117	if (nat_module_is_loaded)
118	return NF_CT_F_NAT;
119
120	return NF_CT_F_BASIC;
121	}
122
123	static unsigned int ipv4_confirm(unsigned int hooknum,
124	struct sk_buff **pskb,
125	const struct net_device *in,
126	const struct net_device *out,
127	int (okfn)(struct sk_buff ))
128	{
129	/* We've seen it coming out the other side: confirm it */
130	return nf_conntrack_confirm(pskb);
131	}
132
133	static unsigned int ipv4_conntrack_help(unsigned int hooknum,
134	struct sk_buff **pskb,
135	const struct net_device *in,
136	const struct net_device *out,
137	int (okfn)(struct sk_buff ))
138	{
139	struct nf_conn *ct;
140	enum ip_conntrack_info ctinfo;
dc808fe2	141	struct nf_conn_help *help;
9fb9cbb1 YK	142
	143	/* This is where we call the helper: as the packet goes out. */
	144	ct = nf_ct_get(*pskb, &ctinfo);
6442f1cf	145	if (!ct \|\| ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY)
dc808fe2 HW	146	return NF_ACCEPT;
	147
	148	help = nfct_help(ct);
	149	if (!help \|\| !help->helper)
	150	return NF_ACCEPT;
	151
	152	return help->helper->help(pskb,
	153	(pskb)->nh.raw - (pskb)->data
	154	+ (pskb)->nh.iph->ihl4,
	155	ct, ctinfo);
9fb9cbb1 YK	156	}
	157
	158	static unsigned int ipv4_conntrack_defrag(unsigned int hooknum,
	159	struct sk_buff **pskb,
	160	const struct net_device *in,
	161	const struct net_device *out,
	162	int (okfn)(struct sk_buff ))
	163	{
	164	#if !defined(CONFIG_IP_NF_NAT) && !defined(CONFIG_IP_NF_NAT_MODULE)
	165	/* Previously seen (loopback)? Ignore. Do this before
	166	fragment check. */
	167	if ((*pskb)->nfct)
	168	return NF_ACCEPT;
	169	#endif
	170
	171	/* Gather fragments. */
	172	if ((*pskb)->nh.iph->frag_off & htons(IP_MF\|IP_OFFSET)) {
	173	pskb = nf_ct_ipv4_gather_frags(pskb,
	174	hooknum == NF_IP_PRE_ROUTING ?
	175	IP_DEFRAG_CONNTRACK_IN :
	176	IP_DEFRAG_CONNTRACK_OUT);
	177	if (!*pskb)
	178	return NF_STOLEN;
	179	}
	180	return NF_ACCEPT;
	181	}
	182
9fb9cbb1 YK	183	static unsigned int ipv4_conntrack_in(unsigned int hooknum,
	184	struct sk_buff **pskb,
	185	const struct net_device *in,
	186	const struct net_device *out,
	187	int (okfn)(struct sk_buff ))
	188	{
	189	return nf_conntrack_in(PF_INET, hooknum, pskb);
	190	}
	191
	192	static unsigned int ipv4_conntrack_local(unsigned int hooknum,
	193	struct sk_buff **pskb,
	194	const struct net_device *in,
	195	const struct net_device *out,
	196	int (okfn)(struct sk_buff ))
	197	{
	198	/* root is playing with raw sockets. */
	199	if ((*pskb)->len < sizeof(struct iphdr)
	200	\|\| (pskb)->nh.iph->ihl 4 < sizeof(struct iphdr)) {
	201	if (net_ratelimit())
	202	printk("ipt_hook: happy cracking.\n");
	203	return NF_ACCEPT;
	204	}
	205	return nf_conntrack_in(PF_INET, hooknum, pskb);
	206	}
	207
	208	/* Connection tracking may drop packets, but never alters them, so
	209	make it the first hook. */
964ddaa1 PM	210	static struct nf_hook_ops ipv4_conntrack_ops[] = {
	211	{
	212	.hook = ipv4_conntrack_defrag,
	213	.owner = THIS_MODULE,
	214	.pf = PF_INET,
	215	.hooknum = NF_IP_PRE_ROUTING,
	216	.priority = NF_IP_PRI_CONNTRACK_DEFRAG,
	217	},
	218	{
	219	.hook = ipv4_conntrack_in,
	220	.owner = THIS_MODULE,
	221	.pf = PF_INET,
	222	.hooknum = NF_IP_PRE_ROUTING,
	223	.priority = NF_IP_PRI_CONNTRACK,
	224	},
	225	{
	226	.hook = ipv4_conntrack_defrag,
	227	.owner = THIS_MODULE,
	228	.pf = PF_INET,
	229	.hooknum = NF_IP_LOCAL_OUT,
	230	.priority = NF_IP_PRI_CONNTRACK_DEFRAG,
	231	},
	232	{
	233	.hook = ipv4_conntrack_local,
	234	.owner = THIS_MODULE,
	235	.pf = PF_INET,
	236	.hooknum = NF_IP_LOCAL_OUT,
	237	.priority = NF_IP_PRI_CONNTRACK,
	238	},
	239	{
	240	.hook = ipv4_conntrack_help,
	241	.owner = THIS_MODULE,
	242	.pf = PF_INET,
	243	.hooknum = NF_IP_POST_ROUTING,
	244	.priority = NF_IP_PRI_CONNTRACK_HELPER,
	245	},
	246	{
	247	.hook = ipv4_conntrack_help,
	248	.owner = THIS_MODULE,
	249	.pf = PF_INET,
	250	.hooknum = NF_IP_LOCAL_IN,
	251	.priority = NF_IP_PRI_CONNTRACK_HELPER,
	252	},
	253	{
	254	.hook = ipv4_confirm,
	255	.owner = THIS_MODULE,
	256	.pf = PF_INET,
	257	.hooknum = NF_IP_POST_ROUTING,
	258	.priority = NF_IP_PRI_CONNTRACK_CONFIRM,
	259	},
	260	{
	261	.hook = ipv4_confirm,
	262	.owner = THIS_MODULE,
	263	.pf = PF_INET,
	264	.hooknum = NF_IP_LOCAL_IN,
	265	.priority = NF_IP_PRI_CONNTRACK_CONFIRM,
	266	},
9fb9cbb1 YK	267	};
9fb9cbb1 YK	268
9fb9cbb1 YK	269	/* Fast function for those who don't want to parse /proc (and I don't
	270	blame them). */
	271	/* Reversing the socket's dst/src point of view gives us the reply
	272	mapping. */
	273	static int
	274	getorigdst(struct sock sk, int optval, void __user user, int *len)
	275	{
	276	struct inet_sock *inet = inet_sk(sk);
	277	struct nf_conntrack_tuple_hash *h;
	278	struct nf_conntrack_tuple tuple;
	279
	280	NF_CT_TUPLE_U_BLANK(&tuple);
	281	tuple.src.u3.ip = inet->rcv_saddr;
	282	tuple.src.u.tcp.port = inet->sport;
	283	tuple.dst.u3.ip = inet->daddr;
	284	tuple.dst.u.tcp.port = inet->dport;
	285	tuple.src.l3num = PF_INET;
	286	tuple.dst.protonum = IPPROTO_TCP;
	287
	288	/* We only do TCP at the moment: is there a better way? */
	289	if (strcmp(sk->sk_prot->name, "TCP")) {
	290	DEBUGP("SO_ORIGINAL_DST: Not a TCP socket\n");
	291	return -ENOPROTOOPT;
	292	}
	293
	294	if ((unsigned int) *len < sizeof(struct sockaddr_in)) {
	295	DEBUGP("SO_ORIGINAL_DST: len %u not %u\n",
	296	*len, sizeof(struct sockaddr_in));
	297	return -EINVAL;
	298	}
	299
	300	h = nf_conntrack_find_get(&tuple, NULL);
	301	if (h) {
	302	struct sockaddr_in sin;
	303	struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h);
	304
	305	sin.sin_family = AF_INET;
	306	sin.sin_port = ct->tuplehash[IP_CT_DIR_ORIGINAL]
	307	.tuple.dst.u.tcp.port;
	308	sin.sin_addr.s_addr = ct->tuplehash[IP_CT_DIR_ORIGINAL]
	309	.tuple.dst.u3.ip;
6c813c3f	310	memset(sin.sin_zero, 0, sizeof(sin.sin_zero));
9fb9cbb1 YK	311
	312	DEBUGP("SO_ORIGINAL_DST: %u.%u.%u.%u %u\n",
	313	NIPQUAD(sin.sin_addr.s_addr), ntohs(sin.sin_port));
	314	nf_ct_put(ct);
	315	if (copy_to_user(user, &sin, sizeof(sin)) != 0)
	316	return -EFAULT;
	317	else
	318	return 0;
	319	}
	320	DEBUGP("SO_ORIGINAL_DST: Can't find %u.%u.%u.%u/%u-%u.%u.%u.%u/%u.\n",
	321	NIPQUAD(tuple.src.u3.ip), ntohs(tuple.src.u.tcp.port),
	322	NIPQUAD(tuple.dst.u3.ip), ntohs(tuple.dst.u.tcp.port));
	323	return -ENOENT;
	324	}
	325
c1d10adb PNA	326	#if defined(CONFIG_NF_CT_NETLINK) \|\| \
	327	defined(CONFIG_NF_CT_NETLINK_MODULE)
	328
	329	#include <linux/netfilter/nfnetlink.h>
	330	#include <linux/netfilter/nfnetlink_conntrack.h>
	331
	332	static int ipv4_tuple_to_nfattr(struct sk_buff *skb,
	333	const struct nf_conntrack_tuple *tuple)
	334	{
	335	NFA_PUT(skb, CTA_IP_V4_SRC, sizeof(u_int32_t),
	336	&tuple->src.u3.ip);
	337	NFA_PUT(skb, CTA_IP_V4_DST, sizeof(u_int32_t),
	338	&tuple->dst.u3.ip);
	339	return 0;
	340
	341	nfattr_failure:
	342	return -1;
	343	}
	344
	345	static const size_t cta_min_ip[CTA_IP_MAX] = {
	346	[CTA_IP_V4_SRC-1] = sizeof(u_int32_t),
	347	[CTA_IP_V4_DST-1] = sizeof(u_int32_t),
	348	};
	349
	350	static int ipv4_nfattr_to_tuple(struct nfattr *tb[],
	351	struct nf_conntrack_tuple *t)
	352	{
	353	if (!tb[CTA_IP_V4_SRC-1] \|\| !tb[CTA_IP_V4_DST-1])
	354	return -EINVAL;
	355
	356	if (nfattr_bad_size(tb, CTA_IP_MAX, cta_min_ip))
	357	return -EINVAL;
	358
	359	t->src.u3.ip =
	360	(u_int32_t )NFA_DATA(tb[CTA_IP_V4_SRC-1]);
	361	t->dst.u3.ip =
	362	(u_int32_t )NFA_DATA(tb[CTA_IP_V4_DST-1]);
	363
	364	return 0;
	365	}
	366	#endif
	367
9fb9cbb1 YK	368	static struct nf_sockopt_ops so_getorigdst = {
	369	.pf = PF_INET,
	370	.get_optmin = SO_ORIGINAL_DST,
	371	.get_optmax = SO_ORIGINAL_DST+1,
	372	.get = &getorigdst,
	373	};
	374
	375	struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv4 = {
	376	.l3proto = PF_INET,
	377	.name = "ipv4",
	378	.pkt_to_tuple = ipv4_pkt_to_tuple,
	379	.invert_tuple = ipv4_invert_tuple,
	380	.print_tuple = ipv4_print_tuple,
	381	.print_conntrack = ipv4_print_conntrack,
	382	.prepare = ipv4_prepare,
	383	.get_features = ipv4_get_features,
c1d10adb PNA	384	#if defined(CONFIG_NF_CT_NETLINK) \|\| \
	385	defined(CONFIG_NF_CT_NETLINK_MODULE)
	386	.tuple_to_nfattr = ipv4_tuple_to_nfattr,
	387	.nfattr_to_tuple = ipv4_nfattr_to_tuple,
	388	#endif
9fb9cbb1 YK	389	.me = THIS_MODULE,
	390	};
	391
32292a7f PM	392	MODULE_ALIAS("nf_conntrack-" __stringify(AF_INET));
	393	MODULE_LICENSE("GPL");
	394
	395	static int __init nf_conntrack_l3proto_ipv4_init(void)
9fb9cbb1 YK	396	{
	397	int ret = 0;
	398
32292a7f	399	need_conntrack();
9fb9cbb1 YK	400
	401	ret = nf_register_sockopt(&so_getorigdst);
	402	if (ret < 0) {
	403	printk(KERN_ERR "Unable to register netfilter socket option\n");
32292a7f	404	return ret;
9fb9cbb1 YK	405	}
9fb9cbb1 YK	406
605dcad6	407	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_tcp4);
9fb9cbb1 YK	408	if (ret < 0) {
	409	printk("nf_conntrack_ipv4: can't register tcp.\n");
	410	goto cleanup_sockopt;
	411	}
	412
605dcad6	413	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_udp4);
9fb9cbb1 YK	414	if (ret < 0) {
	415	printk("nf_conntrack_ipv4: can't register udp.\n");
	416	goto cleanup_tcp;
	417	}
	418
605dcad6	419	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_icmp);
9fb9cbb1 YK	420	if (ret < 0) {
	421	printk("nf_conntrack_ipv4: can't register icmp.\n");
	422	goto cleanup_udp;
	423	}
	424
	425	ret = nf_conntrack_l3proto_register(&nf_conntrack_l3proto_ipv4);
	426	if (ret < 0) {
	427	printk("nf_conntrack_ipv4: can't register ipv4\n");
	428	goto cleanup_icmp;
	429	}
	430
964ddaa1 PM	431	ret = nf_register_hooks(ipv4_conntrack_ops,
964ddaa1 PM	432	ARRAY_SIZE(ipv4_conntrack_ops));
9fb9cbb1	433	if (ret < 0) {
964ddaa1	434	printk("nf_conntrack_ipv4: can't register hooks.\n");
9fb9cbb1 YK	435	goto cleanup_ipv4;
9fb9cbb1 YK	436	}
9fb9cbb1 YK	437	return ret;
9fb9cbb1 YK	438
9fb9cbb1 YK	439	cleanup_ipv4:
	440	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
	441	cleanup_icmp:
605dcad6	442	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
9fb9cbb1	443	cleanup_udp:
605dcad6	444	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
9fb9cbb1	445	cleanup_tcp:
605dcad6	446	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
9fb9cbb1 YK	447	cleanup_sockopt:
9fb9cbb1 YK	448	nf_unregister_sockopt(&so_getorigdst);
9fb9cbb1 YK	449	return ret;
	450	}
	451
65b4b4e8	452	static void __exit nf_conntrack_l3proto_ipv4_fini(void)
9fb9cbb1	453	{
32292a7f	454	synchronize_net();
32292a7f PM	455	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
32292a7f PM	456	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
605dcad6 MJ	457	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
	458	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
	459	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
32292a7f	460	nf_unregister_sockopt(&so_getorigdst);
9fb9cbb1 YK	461	}
9fb9cbb1 YK	462
65b4b4e8 AM	463	module_init(nf_conntrack_l3proto_ipv4_init);
65b4b4e8 AM	464	module_exit(nf_conntrack_l3proto_ipv4_fini);
9fb9cbb1	465
9fb9cbb1	466	EXPORT_SYMBOL(nf_ct_ipv4_gather_frags);