[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)
{
	__be32 _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 nf_nat_module_is_loaded = 0;
EXPORT_SYMBOL_GPL(nf_nat_module_is_loaded);

static u_int32_t ipv4_get_features(const struct nf_conntrack_tuple *tuple)
{
	if (nf_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,
	},
};

#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
static int log_invalid_proto_min = 0;
static int log_invalid_proto_max = 255;

static ctl_table ip_ct_sysctl_table[] = {
	{
		.ctl_name	= NET_IPV4_NF_CONNTRACK_MAX,
		.procname	= "ip_conntrack_max",
		.data		= &nf_conntrack_max,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec,
	},
	{
		.ctl_name	= NET_IPV4_NF_CONNTRACK_COUNT,
		.procname	= "ip_conntrack_count",
		.data		= &nf_conntrack_count,
		.maxlen		= sizeof(int),
		.mode		= 0444,
		.proc_handler	= &proc_dointvec,
	},
	{
		.ctl_name	= NET_IPV4_NF_CONNTRACK_BUCKETS,
		.procname	= "ip_conntrack_buckets",
		.data		= &nf_conntrack_htable_size,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0444,
		.proc_handler	= &proc_dointvec,
	},
	{
		.ctl_name	= NET_IPV4_NF_CONNTRACK_CHECKSUM,
		.procname	= "ip_conntrack_checksum",
		.data		= &nf_conntrack_checksum,
		.maxlen		= sizeof(int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec,
	},
	{
		.ctl_name	= NET_IPV4_NF_CONNTRACK_LOG_INVALID,
		.procname	= "ip_conntrack_log_invalid",
		.data		= &nf_ct_log_invalid,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec_minmax,
		.strategy	= &sysctl_intvec,
		.extra1		= &log_invalid_proto_min,
		.extra2		= &log_invalid_proto_max,
	},
	{
		.ctl_name	= 0
	}
};
#endif /* CONFIG_SYSCTL && CONFIG_NF_CONNTRACK_PROC_COMPAT */

/* 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 = *(__be32 *)NFA_DATA(tb[CTA_IP_V4_SRC-1]);
	t->dst.u3.ip = *(__be32 *)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
#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	.ctl_table_path  = nf_net_ipv4_netfilter_sysctl_path,
	.ctl_table	 = ip_ct_sysctl_table,
#endif
	.me		 = THIS_MODULE,
};

MODULE_ALIAS("nf_conntrack-" __stringify(AF_INET));
MODULE_ALIAS("ip_conntrack");
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;
	}
#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	ret = nf_conntrack_ipv4_compat_init();
	if (ret < 0)
		goto cleanup_hooks;
#endif
	return ret;
#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
 cleanup_hooks:
 	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
#endif
 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();
#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	nf_conntrack_ipv4_compat_fini();
#endif
	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);
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	{
bff9a89b	44	__be32 _addrs[2], *ap;
9fb9cbb1 YK	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
5b1158e9	114	int nf_nat_module_is_loaded = 0;
13b18339 PM	115	EXPORT_SYMBOL_GPL(nf_nat_module_is_loaded);
13b18339 PM	116
9fb9cbb1 YK	117	static u_int32_t ipv4_get_features(const struct nf_conntrack_tuple *tuple)
9fb9cbb1 YK	118	{
5b1158e9	119	if (nf_nat_module_is_loaded)
9fb9cbb1 YK	120	return NF_CT_F_NAT;
	121
	122	return NF_CT_F_BASIC;
	123	}
	124
	125	static unsigned int ipv4_confirm(unsigned int hooknum,
	126	struct sk_buff **pskb,
	127	const struct net_device *in,
	128	const struct net_device *out,
	129	int (okfn)(struct sk_buff ))
	130	{
	131	/* We've seen it coming out the other side: confirm it */
	132	return nf_conntrack_confirm(pskb);
	133	}
	134
	135	static unsigned int ipv4_conntrack_help(unsigned int hooknum,
	136	struct sk_buff **pskb,
	137	const struct net_device *in,
	138	const struct net_device *out,
	139	int (okfn)(struct sk_buff ))
	140	{
	141	struct nf_conn *ct;
	142	enum ip_conntrack_info ctinfo;
dc808fe2	143	struct nf_conn_help *help;
9fb9cbb1 YK	144
	145	/* This is where we call the helper: as the packet goes out. */
	146	ct = nf_ct_get(*pskb, &ctinfo);
6442f1cf	147	if (!ct \|\| ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY)
dc808fe2 HW	148	return NF_ACCEPT;
	149
	150	help = nfct_help(ct);
	151	if (!help \|\| !help->helper)
	152	return NF_ACCEPT;
	153
	154	return help->helper->help(pskb,
	155	(pskb)->nh.raw - (pskb)->data
	156	+ (pskb)->nh.iph->ihl4,
	157	ct, ctinfo);
9fb9cbb1 YK	158	}
	159
	160	static unsigned int ipv4_conntrack_defrag(unsigned int hooknum,
	161	struct sk_buff **pskb,
	162	const struct net_device *in,
	163	const struct net_device *out,
	164	int (okfn)(struct sk_buff ))
	165	{
	166	#if !defined(CONFIG_IP_NF_NAT) && !defined(CONFIG_IP_NF_NAT_MODULE)
	167	/* Previously seen (loopback)? Ignore. Do this before
	168	fragment check. */
	169	if ((*pskb)->nfct)
	170	return NF_ACCEPT;
	171	#endif
	172
	173	/* Gather fragments. */
	174	if ((*pskb)->nh.iph->frag_off & htons(IP_MF\|IP_OFFSET)) {
	175	pskb = nf_ct_ipv4_gather_frags(pskb,
	176	hooknum == NF_IP_PRE_ROUTING ?
	177	IP_DEFRAG_CONNTRACK_IN :
	178	IP_DEFRAG_CONNTRACK_OUT);
	179	if (!*pskb)
	180	return NF_STOLEN;
	181	}
	182	return NF_ACCEPT;
	183	}
	184
9fb9cbb1 YK	185	static unsigned int ipv4_conntrack_in(unsigned int hooknum,
	186	struct sk_buff **pskb,
	187	const struct net_device *in,
	188	const struct net_device *out,
	189	int (okfn)(struct sk_buff ))
	190	{
	191	return nf_conntrack_in(PF_INET, hooknum, pskb);
	192	}
	193
	194	static unsigned int ipv4_conntrack_local(unsigned int hooknum,
	195	struct sk_buff **pskb,
	196	const struct net_device *in,
	197	const struct net_device *out,
	198	int (okfn)(struct sk_buff ))
	199	{
	200	/* root is playing with raw sockets. */
	201	if ((*pskb)->len < sizeof(struct iphdr)
	202	\|\| (pskb)->nh.iph->ihl 4 < sizeof(struct iphdr)) {
	203	if (net_ratelimit())
	204	printk("ipt_hook: happy cracking.\n");
	205	return NF_ACCEPT;
	206	}
	207	return nf_conntrack_in(PF_INET, hooknum, pskb);
	208	}
	209
	210	/* Connection tracking may drop packets, but never alters them, so
	211	make it the first hook. */
964ddaa1 PM	212	static struct nf_hook_ops ipv4_conntrack_ops[] = {
	213	{
	214	.hook = ipv4_conntrack_defrag,
	215	.owner = THIS_MODULE,
	216	.pf = PF_INET,
	217	.hooknum = NF_IP_PRE_ROUTING,
	218	.priority = NF_IP_PRI_CONNTRACK_DEFRAG,
	219	},
	220	{
	221	.hook = ipv4_conntrack_in,
	222	.owner = THIS_MODULE,
	223	.pf = PF_INET,
	224	.hooknum = NF_IP_PRE_ROUTING,
	225	.priority = NF_IP_PRI_CONNTRACK,
	226	},
	227	{
	228	.hook = ipv4_conntrack_defrag,
	229	.owner = THIS_MODULE,
	230	.pf = PF_INET,
	231	.hooknum = NF_IP_LOCAL_OUT,
	232	.priority = NF_IP_PRI_CONNTRACK_DEFRAG,
	233	},
	234	{
	235	.hook = ipv4_conntrack_local,
	236	.owner = THIS_MODULE,
	237	.pf = PF_INET,
	238	.hooknum = NF_IP_LOCAL_OUT,
	239	.priority = NF_IP_PRI_CONNTRACK,
	240	},
	241	{
	242	.hook = ipv4_conntrack_help,
	243	.owner = THIS_MODULE,
	244	.pf = PF_INET,
	245	.hooknum = NF_IP_POST_ROUTING,
	246	.priority = NF_IP_PRI_CONNTRACK_HELPER,
	247	},
	248	{
	249	.hook = ipv4_conntrack_help,
	250	.owner = THIS_MODULE,
	251	.pf = PF_INET,
	252	.hooknum = NF_IP_LOCAL_IN,
	253	.priority = NF_IP_PRI_CONNTRACK_HELPER,
	254	},
	255	{
	256	.hook = ipv4_confirm,
	257	.owner = THIS_MODULE,
	258	.pf = PF_INET,
	259	.hooknum = NF_IP_POST_ROUTING,
	260	.priority = NF_IP_PRI_CONNTRACK_CONFIRM,
	261	},
	262	{
	263	.hook = ipv4_confirm,
	264	.owner = THIS_MODULE,
	265	.pf = PF_INET,
	266	.hooknum = NF_IP_LOCAL_IN,
	267	.priority = NF_IP_PRI_CONNTRACK_CONFIRM,
	268	},
9fb9cbb1 YK	269	};
9fb9cbb1 YK	270
a999e683 PM	271	#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	272	static int log_invalid_proto_min = 0;
	273	static int log_invalid_proto_max = 255;
	274
	275	static ctl_table ip_ct_sysctl_table[] = {
	276	{
	277	.ctl_name = NET_IPV4_NF_CONNTRACK_MAX,
	278	.procname = "ip_conntrack_max",
	279	.data = &nf_conntrack_max,
	280	.maxlen = sizeof(int),
	281	.mode = 0644,
	282	.proc_handler = &proc_dointvec,
	283	},
	284	{
	285	.ctl_name = NET_IPV4_NF_CONNTRACK_COUNT,
	286	.procname = "ip_conntrack_count",
	287	.data = &nf_conntrack_count,
	288	.maxlen = sizeof(int),
	289	.mode = 0444,
	290	.proc_handler = &proc_dointvec,
	291	},
	292	{
	293	.ctl_name = NET_IPV4_NF_CONNTRACK_BUCKETS,
	294	.procname = "ip_conntrack_buckets",
	295	.data = &nf_conntrack_htable_size,
	296	.maxlen = sizeof(unsigned int),
	297	.mode = 0444,
	298	.proc_handler = &proc_dointvec,
	299	},
	300	{
	301	.ctl_name = NET_IPV4_NF_CONNTRACK_CHECKSUM,
	302	.procname = "ip_conntrack_checksum",
	303	.data = &nf_conntrack_checksum,
	304	.maxlen = sizeof(int),
	305	.mode = 0644,
	306	.proc_handler = &proc_dointvec,
	307	},
	308	{
	309	.ctl_name = NET_IPV4_NF_CONNTRACK_LOG_INVALID,
	310	.procname = "ip_conntrack_log_invalid",
	311	.data = &nf_ct_log_invalid,
	312	.maxlen = sizeof(unsigned int),
	313	.mode = 0644,
	314	.proc_handler = &proc_dointvec_minmax,
	315	.strategy = &sysctl_intvec,
	316	.extra1 = &log_invalid_proto_min,
	317	.extra2 = &log_invalid_proto_max,
	318	},
	319	{
	320	.ctl_name = 0
	321	}
	322	};
	323	#endif /* CONFIG_SYSCTL && CONFIG_NF_CONNTRACK_PROC_COMPAT */
	324
9fb9cbb1 YK	325	/* Fast function for those who don't want to parse /proc (and I don't
	326	blame them). */
	327	/* Reversing the socket's dst/src point of view gives us the reply
	328	mapping. */
	329	static int
	330	getorigdst(struct sock sk, int optval, void __user user, int *len)
	331	{
	332	struct inet_sock *inet = inet_sk(sk);
	333	struct nf_conntrack_tuple_hash *h;
	334	struct nf_conntrack_tuple tuple;
	335
	336	NF_CT_TUPLE_U_BLANK(&tuple);
	337	tuple.src.u3.ip = inet->rcv_saddr;
	338	tuple.src.u.tcp.port = inet->sport;
	339	tuple.dst.u3.ip = inet->daddr;
	340	tuple.dst.u.tcp.port = inet->dport;
	341	tuple.src.l3num = PF_INET;
	342	tuple.dst.protonum = IPPROTO_TCP;
	343
	344	/* We only do TCP at the moment: is there a better way? */
	345	if (strcmp(sk->sk_prot->name, "TCP")) {
	346	DEBUGP("SO_ORIGINAL_DST: Not a TCP socket\n");
	347	return -ENOPROTOOPT;
	348	}
	349
	350	if ((unsigned int) *len < sizeof(struct sockaddr_in)) {
	351	DEBUGP("SO_ORIGINAL_DST: len %u not %u\n",
	352	*len, sizeof(struct sockaddr_in));
	353	return -EINVAL;
	354	}
	355
	356	h = nf_conntrack_find_get(&tuple, NULL);
	357	if (h) {
	358	struct sockaddr_in sin;
	359	struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h);
	360
	361	sin.sin_family = AF_INET;
	362	sin.sin_port = ct->tuplehash[IP_CT_DIR_ORIGINAL]
	363	.tuple.dst.u.tcp.port;
	364	sin.sin_addr.s_addr = ct->tuplehash[IP_CT_DIR_ORIGINAL]
	365	.tuple.dst.u3.ip;
6c813c3f	366	memset(sin.sin_zero, 0, sizeof(sin.sin_zero));
9fb9cbb1 YK	367
	368	DEBUGP("SO_ORIGINAL_DST: %u.%u.%u.%u %u\n",
	369	NIPQUAD(sin.sin_addr.s_addr), ntohs(sin.sin_port));
	370	nf_ct_put(ct);
	371	if (copy_to_user(user, &sin, sizeof(sin)) != 0)
	372	return -EFAULT;
	373	else
	374	return 0;
	375	}
	376	DEBUGP("SO_ORIGINAL_DST: Can't find %u.%u.%u.%u/%u-%u.%u.%u.%u/%u.\n",
	377	NIPQUAD(tuple.src.u3.ip), ntohs(tuple.src.u.tcp.port),
	378	NIPQUAD(tuple.dst.u3.ip), ntohs(tuple.dst.u.tcp.port));
	379	return -ENOENT;
	380	}
	381
c1d10adb PNA	382	#if defined(CONFIG_NF_CT_NETLINK) \|\| \
	383	defined(CONFIG_NF_CT_NETLINK_MODULE)
	384
	385	#include <linux/netfilter/nfnetlink.h>
	386	#include <linux/netfilter/nfnetlink_conntrack.h>
	387
	388	static int ipv4_tuple_to_nfattr(struct sk_buff *skb,
	389	const struct nf_conntrack_tuple *tuple)
	390	{
	391	NFA_PUT(skb, CTA_IP_V4_SRC, sizeof(u_int32_t),
	392	&tuple->src.u3.ip);
	393	NFA_PUT(skb, CTA_IP_V4_DST, sizeof(u_int32_t),
	394	&tuple->dst.u3.ip);
	395	return 0;
	396
	397	nfattr_failure:
	398	return -1;
	399	}
	400
	401	static const size_t cta_min_ip[CTA_IP_MAX] = {
	402	[CTA_IP_V4_SRC-1] = sizeof(u_int32_t),
	403	[CTA_IP_V4_DST-1] = sizeof(u_int32_t),
	404	};
	405
	406	static int ipv4_nfattr_to_tuple(struct nfattr *tb[],
	407	struct nf_conntrack_tuple *t)
	408	{
	409	if (!tb[CTA_IP_V4_SRC-1] \|\| !tb[CTA_IP_V4_DST-1])
	410	return -EINVAL;
	411
	412	if (nfattr_bad_size(tb, CTA_IP_MAX, cta_min_ip))
	413	return -EINVAL;
	414
bff9a89b PM	415	t->src.u3.ip = (__be32 )NFA_DATA(tb[CTA_IP_V4_SRC-1]);
bff9a89b PM	416	t->dst.u3.ip = (__be32 )NFA_DATA(tb[CTA_IP_V4_DST-1]);
c1d10adb PNA	417
	418	return 0;
	419	}
	420	#endif
	421
9fb9cbb1 YK	422	static struct nf_sockopt_ops so_getorigdst = {
	423	.pf = PF_INET,
	424	.get_optmin = SO_ORIGINAL_DST,
	425	.get_optmax = SO_ORIGINAL_DST+1,
	426	.get = &getorigdst,
	427	};
	428
	429	struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv4 = {
	430	.l3proto = PF_INET,
	431	.name = "ipv4",
	432	.pkt_to_tuple = ipv4_pkt_to_tuple,
	433	.invert_tuple = ipv4_invert_tuple,
	434	.print_tuple = ipv4_print_tuple,
	435	.print_conntrack = ipv4_print_conntrack,
	436	.prepare = ipv4_prepare,
	437	.get_features = ipv4_get_features,
c1d10adb PNA	438	#if defined(CONFIG_NF_CT_NETLINK) \|\| \
	439	defined(CONFIG_NF_CT_NETLINK_MODULE)
	440	.tuple_to_nfattr = ipv4_tuple_to_nfattr,
	441	.nfattr_to_tuple = ipv4_nfattr_to_tuple,
a999e683 PM	442	#endif
	443	#if defined(CONFIG_SYSCTL) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	444	.ctl_table_path = nf_net_ipv4_netfilter_sysctl_path,
	445	.ctl_table = ip_ct_sysctl_table,
c1d10adb	446	#endif
9fb9cbb1 YK	447	.me = THIS_MODULE,
	448	};
	449
32292a7f	450	MODULE_ALIAS("nf_conntrack-" __stringify(AF_INET));
d2483dde	451	MODULE_ALIAS("ip_conntrack");
32292a7f PM	452	MODULE_LICENSE("GPL");
	453
	454	static int __init nf_conntrack_l3proto_ipv4_init(void)
9fb9cbb1 YK	455	{
	456	int ret = 0;
	457
32292a7f	458	need_conntrack();
9fb9cbb1 YK	459
	460	ret = nf_register_sockopt(&so_getorigdst);
	461	if (ret < 0) {
	462	printk(KERN_ERR "Unable to register netfilter socket option\n");
32292a7f	463	return ret;
9fb9cbb1 YK	464	}
9fb9cbb1 YK	465
605dcad6	466	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_tcp4);
9fb9cbb1 YK	467	if (ret < 0) {
	468	printk("nf_conntrack_ipv4: can't register tcp.\n");
	469	goto cleanup_sockopt;
	470	}
	471
605dcad6	472	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_udp4);
9fb9cbb1 YK	473	if (ret < 0) {
	474	printk("nf_conntrack_ipv4: can't register udp.\n");
	475	goto cleanup_tcp;
	476	}
	477
605dcad6	478	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_icmp);
9fb9cbb1 YK	479	if (ret < 0) {
	480	printk("nf_conntrack_ipv4: can't register icmp.\n");
	481	goto cleanup_udp;
	482	}
	483
	484	ret = nf_conntrack_l3proto_register(&nf_conntrack_l3proto_ipv4);
	485	if (ret < 0) {
	486	printk("nf_conntrack_ipv4: can't register ipv4\n");
	487	goto cleanup_icmp;
	488	}
	489
964ddaa1 PM	490	ret = nf_register_hooks(ipv4_conntrack_ops,
964ddaa1 PM	491	ARRAY_SIZE(ipv4_conntrack_ops));
9fb9cbb1	492	if (ret < 0) {
964ddaa1	493	printk("nf_conntrack_ipv4: can't register hooks.\n");
9fb9cbb1 YK	494	goto cleanup_ipv4;
9fb9cbb1 YK	495	}
e4bd8bce PM	496	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	497	ret = nf_conntrack_ipv4_compat_init();
	498	if (ret < 0)
	499	goto cleanup_hooks;
	500	#endif
9fb9cbb1	501	return ret;
e4bd8bce PM	502	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	503	cleanup_hooks:
	504	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
	505	#endif
9fb9cbb1 YK	506	cleanup_ipv4:
	507	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
	508	cleanup_icmp:
605dcad6	509	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
9fb9cbb1	510	cleanup_udp:
605dcad6	511	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
9fb9cbb1	512	cleanup_tcp:
605dcad6	513	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
9fb9cbb1 YK	514	cleanup_sockopt:
9fb9cbb1 YK	515	nf_unregister_sockopt(&so_getorigdst);
9fb9cbb1 YK	516	return ret;
	517	}
	518
65b4b4e8	519	static void __exit nf_conntrack_l3proto_ipv4_fini(void)
9fb9cbb1	520	{
32292a7f	521	synchronize_net();
e4bd8bce PM	522	#if defined(CONFIG_PROC_FS) && defined(CONFIG_NF_CONNTRACK_PROC_COMPAT)
	523	nf_conntrack_ipv4_compat_fini();
	524	#endif
32292a7f PM	525	nf_unregister_hooks(ipv4_conntrack_ops, ARRAY_SIZE(ipv4_conntrack_ops));
32292a7f PM	526	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4);
605dcad6 MJ	527	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmp);
	528	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp4);
	529	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp4);
32292a7f	530	nf_unregister_sockopt(&so_getorigdst);
9fb9cbb1 YK	531	}
9fb9cbb1 YK	532
65b4b4e8 AM	533	module_init(nf_conntrack_l3proto_ipv4_init);
65b4b4e8 AM	534	module_exit(nf_conntrack_l3proto_ipv4_fini);