[net-next-2.6.git] / net / ipv6 / netfilter / nf_conntrack_l3proto_ipv6.c

/*
 * Copyright (C)2004 USAGI/WIDE Project
 *
 * 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.
 *
 * Author:
 *	Yasuyuki Kozakai @USAGI <yasuyuki.kozakai@toshiba.co.jp>
 *
 * 16 Dec 2003: Yasuyuki Kozakai @USAGI <yasuyuki.kozakai@toshiba.co.jp>
 *	- support Layer 3 protocol independent connection tracking.
 *	  Based on the original ip_conntrack code which	had the following
 *	  copyright information:
 *		(C) 1999-2001 Paul `Rusty' Russell
 *		(C) 2002-2004 Netfilter Core Team <coreteam@netfilter.org>
 *
 * 23 Mar 2004: Yasuyuki Kozakai @USAGI <yasuyuki.kozakai@toshiba.co.jp>
 *	- add get_features() to support various size of conntrack
 *	  structures.
 */

#include <linux/types.h>
#include <linux/ipv6.h>
#include <linux/in6.h>
#include <linux/netfilter.h>
#include <linux/module.h>
#include <linux/skbuff.h>
#include <linux/icmp.h>
#include <linux/sysctl.h>
#include <net/ipv6.h>

#include <linux/netfilter_ipv6.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>

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

static int ipv6_pkt_to_tuple(const struct sk_buff *skb, unsigned int nhoff,
			     struct nf_conntrack_tuple *tuple)
{
	u_int32_t _addrs[8], *ap;

	ap = skb_header_pointer(skb, nhoff + offsetof(struct ipv6hdr, saddr),
				sizeof(_addrs), _addrs);
	if (ap == NULL)
		return 0;

	memcpy(tuple->src.u3.ip6, ap, sizeof(tuple->src.u3.ip6));
	memcpy(tuple->dst.u3.ip6, ap + 4, sizeof(tuple->dst.u3.ip6));

	return 1;
}

static int ipv6_invert_tuple(struct nf_conntrack_tuple *tuple,
			     const struct nf_conntrack_tuple *orig)
{
	memcpy(tuple->src.u3.ip6, orig->dst.u3.ip6, sizeof(tuple->src.u3.ip6));
	memcpy(tuple->dst.u3.ip6, orig->src.u3.ip6, sizeof(tuple->dst.u3.ip6));

	return 1;
}

static int ipv6_print_tuple(struct seq_file *s,
			    const struct nf_conntrack_tuple *tuple)
{
	return seq_printf(s, "src=" NIP6_FMT " dst=" NIP6_FMT " ",
			  NIP6(*((struct in6_addr *)tuple->src.u3.ip6)),
			  NIP6(*((struct in6_addr *)tuple->dst.u3.ip6)));
}

static int ipv6_print_conntrack(struct seq_file *s,
				const struct nf_conn *conntrack)
{
	return 0;
}

/*
 * Based on ipv6_skip_exthdr() in net/ipv6/exthdr.c
 *
 * This function parses (probably truncated) exthdr set "hdr"
 * of length "len". "nexthdrp" initially points to some place,
 * where type of the first header can be found.
 *
 * It skips all well-known exthdrs, and returns pointer to the start
 * of unparsable area i.e. the first header with unknown type.
 * if success, *nexthdr is updated by type/protocol of this header.
 *
 * NOTES: - it may return pointer pointing beyond end of packet,
 *          if the last recognized header is truncated in the middle.
 *        - if packet is truncated, so that all parsed headers are skipped,
 *          it returns -1.
 *        - if packet is fragmented, return pointer of the fragment header.
 *        - ESP is unparsable for now and considered like
 *          normal payload protocol.
 *        - Note also special handling of AUTH header. Thanks to IPsec wizards.
 */

int nf_ct_ipv6_skip_exthdr(struct sk_buff *skb, int start, u8 *nexthdrp,
			   int len)
{
	u8 nexthdr = *nexthdrp;

	while (ipv6_ext_hdr(nexthdr)) {
		struct ipv6_opt_hdr hdr;
		int hdrlen;

		if (len < (int)sizeof(struct ipv6_opt_hdr))
			return -1;
		if (nexthdr == NEXTHDR_NONE)
			break;
		if (nexthdr == NEXTHDR_FRAGMENT)
			break;
		if (skb_copy_bits(skb, start, &hdr, sizeof(hdr)))
			BUG();
		if (nexthdr == NEXTHDR_AUTH)
			hdrlen = (hdr.hdrlen+2)<<2;
		else
			hdrlen = ipv6_optlen(&hdr);

		nexthdr = hdr.nexthdr;
		len -= hdrlen;
		start += hdrlen;
	}

	*nexthdrp = nexthdr;
	return start;
}

static int
ipv6_prepare(struct sk_buff **pskb, unsigned int hooknum, unsigned int *dataoff,
	     u_int8_t *protonum)
{
	unsigned int extoff;
	unsigned char pnum;
	int protoff;

	extoff = (u8*)((*pskb)->nh.ipv6h + 1) - (*pskb)->data;
	pnum = (*pskb)->nh.ipv6h->nexthdr;

	protoff = nf_ct_ipv6_skip_exthdr(*pskb, extoff, &pnum,
					 (*pskb)->len - extoff);

	/*
	 * (protoff == (*pskb)->len) mean that the packet doesn't have no data
	 * except of IPv6 & ext headers. but it's tracked anyway. - YK
	 */
	if ((protoff < 0) || (protoff > (*pskb)->len)) {
		DEBUGP("ip6_conntrack_core: can't find proto in pkt\n");
		NF_CT_STAT_INC(error);
		NF_CT_STAT_INC(invalid);
		return -NF_ACCEPT;
	}

	*dataoff = protoff;
	*protonum = pnum;
	return NF_ACCEPT;
}

static u_int32_t ipv6_get_features(const struct nf_conntrack_tuple *tuple)
{
	return NF_CT_F_BASIC;
}

static unsigned int ipv6_confirm(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;
	struct nf_conn_help *help;
	enum ip_conntrack_info ctinfo;
	unsigned int ret, protoff;
	unsigned int extoff = (u8*)((*pskb)->nh.ipv6h + 1)
			      - (*pskb)->data;
	unsigned char pnum = (*pskb)->nh.ipv6h->nexthdr;


	/* 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)
		goto out;

	help = nfct_help(ct);
	if (!help || !help->helper)
		goto out;

	protoff = nf_ct_ipv6_skip_exthdr(*pskb, extoff, &pnum,
					 (*pskb)->len - extoff);
	if (protoff < 0 || protoff > (*pskb)->len ||
	    pnum == NEXTHDR_FRAGMENT) {
		DEBUGP("proto header not found\n");
		return NF_ACCEPT;
	}

	ret = help->helper->help(pskb, protoff, ct, ctinfo);
	if (ret != NF_ACCEPT)
		return ret;
out:
	/* We've seen it coming out the other side: confirm it */
	return nf_conntrack_confirm(pskb);
}

static unsigned int ipv6_defrag(unsigned int hooknum,
				struct sk_buff **pskb,
				const struct net_device *in,
				const struct net_device *out,
				int (*okfn)(struct sk_buff *))
{
	struct sk_buff *reasm;

	/* Previously seen (loopback)?  */
	if ((*pskb)->nfct)
		return NF_ACCEPT;

	reasm = nf_ct_frag6_gather(*pskb);

	/* queued */
	if (reasm == NULL)
		return NF_STOLEN;

	/* error occured or not fragmented */
	if (reasm == *pskb)
		return NF_ACCEPT;

	nf_ct_frag6_output(hooknum, reasm, (struct net_device *)in,
			   (struct net_device *)out, okfn);

	return NF_STOLEN;
}

static unsigned int ipv6_conntrack_in(unsigned int hooknum,
				      struct sk_buff **pskb,
				      const struct net_device *in,
				      const struct net_device *out,
				      int (*okfn)(struct sk_buff *))
{
	struct sk_buff *reasm = (*pskb)->nfct_reasm;

	/* This packet is fragmented and has reassembled packet. */
	if (reasm) {
		/* Reassembled packet isn't parsed yet ? */
		if (!reasm->nfct) {
			unsigned int ret;

			ret = nf_conntrack_in(PF_INET6, hooknum, &reasm);
			if (ret != NF_ACCEPT)
				return ret;
		}
		nf_conntrack_get(reasm->nfct);
		(*pskb)->nfct = reasm->nfct;
		return NF_ACCEPT;
	}

	return nf_conntrack_in(PF_INET6, hooknum, pskb);
}

static unsigned int ipv6_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 ipv6hdr)) {
		if (net_ratelimit())
			printk("ipv6_conntrack_local: packet too short\n");
		return NF_ACCEPT;
	}
	return ipv6_conntrack_in(hooknum, pskb, in, out, okfn);
}

static struct nf_hook_ops ipv6_conntrack_ops[] = {
	{
		.hook		= ipv6_defrag,
		.owner		= THIS_MODULE,
		.pf		= PF_INET6,
		.hooknum	= NF_IP6_PRE_ROUTING,
		.priority	= NF_IP6_PRI_CONNTRACK_DEFRAG,
	},
	{
		.hook		= ipv6_conntrack_in,
		.owner		= THIS_MODULE,
		.pf		= PF_INET6,
		.hooknum	= NF_IP6_PRE_ROUTING,
		.priority	= NF_IP6_PRI_CONNTRACK,
	},
	{
		.hook		= ipv6_conntrack_local,
		.owner		= THIS_MODULE,
		.pf		= PF_INET6,
		.hooknum	= NF_IP6_LOCAL_OUT,
		.priority	= NF_IP6_PRI_CONNTRACK,
	},
	{
		.hook		= ipv6_defrag,
		.owner		= THIS_MODULE,
		.pf		= PF_INET6,
		.hooknum	= NF_IP6_LOCAL_OUT,
		.priority	= NF_IP6_PRI_CONNTRACK_DEFRAG,
	},
	{
		.hook		= ipv6_confirm,
		.owner		= THIS_MODULE,
		.pf		= PF_INET6,
		.hooknum	= NF_IP6_POST_ROUTING,
		.priority	= NF_IP6_PRI_LAST,
	},
	{
		.hook		= ipv6_confirm,
		.owner		= THIS_MODULE,
		.pf		= PF_INET6,
		.hooknum	= NF_IP6_LOCAL_IN,
		.priority	= NF_IP6_PRI_LAST-1,
	},
};

#ifdef CONFIG_SYSCTL

/* From nf_conntrack_proto_icmpv6.c */
extern unsigned int nf_ct_icmpv6_timeout;

static struct ctl_table_header *nf_ct_ipv6_sysctl_header;

static ctl_table nf_ct_sysctl_table[] = {
	{
		.ctl_name	= NET_NF_CONNTRACK_ICMPV6_TIMEOUT,
		.procname	= "nf_conntrack_icmpv6_timeout",
		.data		= &nf_ct_icmpv6_timeout,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec_jiffies,
	},
	{
		.ctl_name	= NET_NF_CONNTRACK_FRAG6_TIMEOUT,
		.procname	= "nf_conntrack_frag6_timeout",
		.data		= &nf_ct_frag6_timeout,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec_jiffies,
	},
	{
		.ctl_name	= NET_NF_CONNTRACK_FRAG6_LOW_THRESH,
		.procname	= "nf_conntrack_frag6_low_thresh",
		.data		= &nf_ct_frag6_low_thresh,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec,
	},
	{
		.ctl_name	= NET_NF_CONNTRACK_FRAG6_HIGH_THRESH,
		.procname	= "nf_conntrack_frag6_high_thresh",
		.data		= &nf_ct_frag6_high_thresh,
		.maxlen		= sizeof(unsigned int),
		.mode		= 0644,
		.proc_handler	= &proc_dointvec,
	},
        { .ctl_name = 0 }
};

static ctl_table nf_ct_netfilter_table[] = {
	{
		.ctl_name	= NET_NETFILTER,
		.procname	= "netfilter",
		.mode		= 0555,
		.child		= nf_ct_sysctl_table,
	},
	{ .ctl_name = 0 }
};

static ctl_table nf_ct_net_table[] = {
	{
		.ctl_name	= CTL_NET,
		.procname	= "net",
		.mode		= 0555,
		.child		= nf_ct_netfilter_table,
	},
	{ .ctl_name = 0 }
};
#endif

#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 ipv6_tuple_to_nfattr(struct sk_buff *skb,
				const struct nf_conntrack_tuple *tuple)
{
	NFA_PUT(skb, CTA_IP_V6_SRC, sizeof(u_int32_t) * 4,
		&tuple->src.u3.ip6);
	NFA_PUT(skb, CTA_IP_V6_DST, sizeof(u_int32_t) * 4,
		&tuple->dst.u3.ip6);
	return 0;

nfattr_failure:
	return -1;
}

static const size_t cta_min_ip[CTA_IP_MAX] = {
	[CTA_IP_V6_SRC-1]       = sizeof(u_int32_t)*4,
	[CTA_IP_V6_DST-1]       = sizeof(u_int32_t)*4,
};

static int ipv6_nfattr_to_tuple(struct nfattr *tb[],
				struct nf_conntrack_tuple *t)
{
	if (!tb[CTA_IP_V6_SRC-1] || !tb[CTA_IP_V6_DST-1])
		return -EINVAL;

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

	memcpy(&t->src.u3.ip6, NFA_DATA(tb[CTA_IP_V6_SRC-1]), 
	       sizeof(u_int32_t) * 4);
	memcpy(&t->dst.u3.ip6, NFA_DATA(tb[CTA_IP_V6_DST-1]),
	       sizeof(u_int32_t) * 4);

	return 0;
}
#endif

struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv6 = {
	.l3proto		= PF_INET6,
	.name			= "ipv6",
	.pkt_to_tuple		= ipv6_pkt_to_tuple,
	.invert_tuple		= ipv6_invert_tuple,
	.print_tuple		= ipv6_print_tuple,
	.print_conntrack	= ipv6_print_conntrack,
	.prepare		= ipv6_prepare,
#if defined(CONFIG_NF_CT_NETLINK) || \
    defined(CONFIG_NF_CT_NETLINK_MODULE)
	.tuple_to_nfattr	= ipv6_tuple_to_nfattr,
	.nfattr_to_tuple	= ipv6_nfattr_to_tuple,
#endif
	.get_features		= ipv6_get_features,
	.me			= THIS_MODULE,
};

MODULE_ALIAS("nf_conntrack-" __stringify(AF_INET6));
MODULE_LICENSE("GPL");
MODULE_AUTHOR("Yasuyuki KOZAKAI @USAGI <yasuyuki.kozakai@toshiba.co.jp>");

static int __init nf_conntrack_l3proto_ipv6_init(void)
{
	int ret = 0;

	need_conntrack();

	ret = nf_ct_frag6_init();
	if (ret < 0) {
		printk("nf_conntrack_ipv6: can't initialize frag6.\n");
		return ret;
	}
	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_tcp6);
	if (ret < 0) {
		printk("nf_conntrack_ipv6: can't register tcp.\n");
		goto cleanup_frag6;
	}

	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_udp6);
	if (ret < 0) {
		printk("nf_conntrack_ipv6: can't register udp.\n");
		goto cleanup_tcp;
	}

	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_icmpv6);
	if (ret < 0) {
		printk("nf_conntrack_ipv6: can't register icmpv6.\n");
		goto cleanup_udp;
	}

	ret = nf_conntrack_l3proto_register(&nf_conntrack_l3proto_ipv6);
	if (ret < 0) {
		printk("nf_conntrack_ipv6: can't register ipv6\n");
		goto cleanup_icmpv6;
	}

	ret = nf_register_hooks(ipv6_conntrack_ops,
				ARRAY_SIZE(ipv6_conntrack_ops));
	if (ret < 0) {
		printk("nf_conntrack_ipv6: can't register pre-routing defrag "
		       "hook.\n");
		goto cleanup_ipv6;
	}
#ifdef CONFIG_SYSCTL
	nf_ct_ipv6_sysctl_header = register_sysctl_table(nf_ct_net_table, 0);
	if (nf_ct_ipv6_sysctl_header == NULL) {
		printk("nf_conntrack: can't register to sysctl.\n");
		ret = -ENOMEM;
		goto cleanup_hooks;
	}
#endif
	return ret;

#ifdef CONFIG_SYSCTL
 cleanup_hooks:
	nf_unregister_hooks(ipv6_conntrack_ops, ARRAY_SIZE(ipv6_conntrack_ops));
#endif
 cleanup_ipv6:
	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv6);
 cleanup_icmpv6:
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmpv6);
 cleanup_udp:
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp6);
 cleanup_tcp:
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp6);
 cleanup_frag6:
	nf_ct_frag6_cleanup();
	return ret;
}

static void __exit nf_conntrack_l3proto_ipv6_fini(void)
{
	synchronize_net();
#ifdef CONFIG_SYSCTL
 	unregister_sysctl_table(nf_ct_ipv6_sysctl_header);
#endif
	nf_unregister_hooks(ipv6_conntrack_ops, ARRAY_SIZE(ipv6_conntrack_ops));
	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv6);
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmpv6);
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp6);
	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp6);
	nf_ct_frag6_cleanup();
}

module_init(nf_conntrack_l3proto_ipv6_init);
module_exit(nf_conntrack_l3proto_ipv6_fini);
Commit	Line	Data
9fb9cbb1 YK	1	/*
	2	* Copyright (C)2004 USAGI/WIDE Project
	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	* Author:
	9	* Yasuyuki Kozakai @USAGI <yasuyuki.kozakai@toshiba.co.jp>
	10	*
	11	* 16 Dec 2003: Yasuyuki Kozakai @USAGI <yasuyuki.kozakai@toshiba.co.jp>
	12	* - support Layer 3 protocol independent connection tracking.
	13	* Based on the original ip_conntrack code which had the following
	14	* copyright information:
	15	* (C) 1999-2001 Paul `Rusty' Russell
	16	* (C) 2002-2004 Netfilter Core Team <coreteam@netfilter.org>
	17	*
	18	* 23 Mar 2004: Yasuyuki Kozakai @USAGI <yasuyuki.kozakai@toshiba.co.jp>
	19	* - add get_features() to support various size of conntrack
	20	* structures.
	21	*/
	22
9fb9cbb1 YK	23	#include <linux/types.h>
	24	#include <linux/ipv6.h>
	25	#include <linux/in6.h>
	26	#include <linux/netfilter.h>
	27	#include <linux/module.h>
	28	#include <linux/skbuff.h>
	29	#include <linux/icmp.h>
	30	#include <linux/sysctl.h>
	31	#include <net/ipv6.h>
	32
	33	#include <linux/netfilter_ipv6.h>
	34	#include <net/netfilter/nf_conntrack.h>
	35	#include <net/netfilter/nf_conntrack_helper.h>
605dcad6	36	#include <net/netfilter/nf_conntrack_l4proto.h>
9fb9cbb1 YK	37	#include <net/netfilter/nf_conntrack_l3proto.h>
	38	#include <net/netfilter/nf_conntrack_core.h>
	39
	40	#if 0
	41	#define DEBUGP printk
	42	#else
	43	#define DEBUGP(format, args...)
	44	#endif
	45
9fb9cbb1 YK	46	static int ipv6_pkt_to_tuple(const struct sk_buff *skb, unsigned int nhoff,
	47	struct nf_conntrack_tuple *tuple)
	48	{
	49	u_int32_t _addrs[8], *ap;
	50
	51	ap = skb_header_pointer(skb, nhoff + offsetof(struct ipv6hdr, saddr),
	52	sizeof(_addrs), _addrs);
	53	if (ap == NULL)
	54	return 0;
	55
	56	memcpy(tuple->src.u3.ip6, ap, sizeof(tuple->src.u3.ip6));
	57	memcpy(tuple->dst.u3.ip6, ap + 4, sizeof(tuple->dst.u3.ip6));
	58
	59	return 1;
	60	}
	61
	62	static int ipv6_invert_tuple(struct nf_conntrack_tuple *tuple,
	63	const struct nf_conntrack_tuple *orig)
	64	{
	65	memcpy(tuple->src.u3.ip6, orig->dst.u3.ip6, sizeof(tuple->src.u3.ip6));
	66	memcpy(tuple->dst.u3.ip6, orig->src.u3.ip6, sizeof(tuple->dst.u3.ip6));
	67
	68	return 1;
	69	}
	70
	71	static int ipv6_print_tuple(struct seq_file *s,
	72	const struct nf_conntrack_tuple *tuple)
	73	{
46b86a2d	74	return seq_printf(s, "src=" NIP6_FMT " dst=" NIP6_FMT " ",
9fb9cbb1 YK	75	NIP6(((struct in6_addr )tuple->src.u3.ip6)),
	76	NIP6(((struct in6_addr )tuple->dst.u3.ip6)));
	77	}
	78
	79	static int ipv6_print_conntrack(struct seq_file *s,
	80	const struct nf_conn *conntrack)
	81	{
	82	return 0;
	83	}
	84
	85	/*
	86	* Based on ipv6_skip_exthdr() in net/ipv6/exthdr.c
	87	*
	88	* This function parses (probably truncated) exthdr set "hdr"
	89	* of length "len". "nexthdrp" initially points to some place,
	90	* where type of the first header can be found.
	91	*
	92	* It skips all well-known exthdrs, and returns pointer to the start
	93	* of unparsable area i.e. the first header with unknown type.
	94	* if success, *nexthdr is updated by type/protocol of this header.
	95	*
	96	* NOTES: - it may return pointer pointing beyond end of packet,
	97	* if the last recognized header is truncated in the middle.
	98	* - if packet is truncated, so that all parsed headers are skipped,
	99	* it returns -1.
	100	* - if packet is fragmented, return pointer of the fragment header.
	101	* - ESP is unparsable for now and considered like
	102	* normal payload protocol.
	103	* - Note also special handling of AUTH header. Thanks to IPsec wizards.
	104	*/
	105
	106	int nf_ct_ipv6_skip_exthdr(struct sk_buff skb, int start, u8 nexthdrp,
	107	int len)
	108	{
	109	u8 nexthdr = *nexthdrp;
	110
	111	while (ipv6_ext_hdr(nexthdr)) {
	112	struct ipv6_opt_hdr hdr;
	113	int hdrlen;
	114
	115	if (len < (int)sizeof(struct ipv6_opt_hdr))
	116	return -1;
	117	if (nexthdr == NEXTHDR_NONE)
	118	break;
	119	if (nexthdr == NEXTHDR_FRAGMENT)
	120	break;
	121	if (skb_copy_bits(skb, start, &hdr, sizeof(hdr)))
	122	BUG();
	123	if (nexthdr == NEXTHDR_AUTH)
	124	hdrlen = (hdr.hdrlen+2)<<2;
	125	else
	126	hdrlen = ipv6_optlen(&hdr);
	127
	128	nexthdr = hdr.nexthdr;
	129	len -= hdrlen;
	130	start += hdrlen;
	131	}
	132
	133	*nexthdrp = nexthdr;
	134	return start;
	135	}
	136
	137	static int
	138	ipv6_prepare(struct sk_buff *pskb, unsigned int hooknum, unsigned int dataoff,
139	u_int8_t *protonum)
140	{
141	unsigned int extoff;
142	unsigned char pnum;
143	int protoff;
144
145	extoff = (u8)((pskb)->nh.ipv6h + 1) - (*pskb)->data;
146	pnum = (*pskb)->nh.ipv6h->nexthdr;
147
148	protoff = nf_ct_ipv6_skip_exthdr(*pskb, extoff, &pnum,
149	(*pskb)->len - extoff);
150
151	/*
152	* (protoff == (*pskb)->len) mean that the packet doesn't have no data
153	* except of IPv6 & ext headers. but it's tracked anyway. - YK
154	*/
155	if ((protoff < 0) \|\| (protoff > (*pskb)->len)) {
156	DEBUGP("ip6_conntrack_core: can't find proto in pkt\n");
157	NF_CT_STAT_INC(error);
158	NF_CT_STAT_INC(invalid);
159	return -NF_ACCEPT;
160	}
161
162	*dataoff = protoff;
163	*protonum = pnum;
164	return NF_ACCEPT;
165	}
166
167	static u_int32_t ipv6_get_features(const struct nf_conntrack_tuple *tuple)
168	{
169	return NF_CT_F_BASIC;
170	}
171
172	static unsigned int ipv6_confirm(unsigned int hooknum,
173	struct sk_buff **pskb,
174	const struct net_device *in,
175	const struct net_device *out,
176	int (okfn)(struct sk_buff ))
177	{
178	struct nf_conn *ct;
dc808fe2	179	struct nf_conn_help *help;
9fb9cbb1	180	enum ip_conntrack_info ctinfo;
dc808fe2 HW	181	unsigned int ret, protoff;
	182	unsigned int extoff = (u8)((pskb)->nh.ipv6h + 1)
	183	- (*pskb)->data;
	184	unsigned char pnum = (*pskb)->nh.ipv6h->nexthdr;
	185
9fb9cbb1 YK	186
	187	/* This is where we call the helper: as the packet goes out. */
	188	ct = nf_ct_get(*pskb, &ctinfo);
6442f1cf	189	if (!ct \|\| ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY)
dc808fe2 HW	190	goto out;
	191
	192	help = nfct_help(ct);
	193	if (!help \|\| !help->helper)
	194	goto out;
	195
	196	protoff = nf_ct_ipv6_skip_exthdr(*pskb, extoff, &pnum,
	197	(*pskb)->len - extoff);
	198	if (protoff < 0 \|\| protoff > (*pskb)->len \|\|
	199	pnum == NEXTHDR_FRAGMENT) {
	200	DEBUGP("proto header not found\n");
	201	return NF_ACCEPT;
9fb9cbb1 YK	202	}
9fb9cbb1 YK	203
dc808fe2 HW	204	ret = help->helper->help(pskb, protoff, ct, ctinfo);
	205	if (ret != NF_ACCEPT)
	206	return ret;
	207	out:
9fb9cbb1	208	/* We've seen it coming out the other side: confirm it */
9fb9cbb1 YK	209	return nf_conntrack_confirm(pskb);
	210	}
	211
9fb9cbb1 YK	212	static unsigned int ipv6_defrag(unsigned int hooknum,
	213	struct sk_buff **pskb,
	214	const struct net_device *in,
	215	const struct net_device *out,
	216	int (okfn)(struct sk_buff ))
	217	{
	218	struct sk_buff *reasm;
	219
	220	/* Previously seen (loopback)? */
	221	if ((*pskb)->nfct)
	222	return NF_ACCEPT;
	223
	224	reasm = nf_ct_frag6_gather(*pskb);
	225
	226	/* queued */
	227	if (reasm == NULL)
	228	return NF_STOLEN;
	229
	230	/* error occured or not fragmented */
	231	if (reasm == *pskb)
	232	return NF_ACCEPT;
	233
	234	nf_ct_frag6_output(hooknum, reasm, (struct net_device *)in,
	235	(struct net_device *)out, okfn);
	236
	237	return NF_STOLEN;
	238	}
	239
	240	static unsigned int ipv6_conntrack_in(unsigned int hooknum,
	241	struct sk_buff **pskb,
	242	const struct net_device *in,
	243	const struct net_device *out,
	244	int (okfn)(struct sk_buff ))
	245	{
	246	struct sk_buff reasm = (pskb)->nfct_reasm;
	247
	248	/* This packet is fragmented and has reassembled packet. */
	249	if (reasm) {
	250	/* Reassembled packet isn't parsed yet ? */
	251	if (!reasm->nfct) {
	252	unsigned int ret;
	253
	254	ret = nf_conntrack_in(PF_INET6, hooknum, &reasm);
	255	if (ret != NF_ACCEPT)
	256	return ret;
	257	}
	258	nf_conntrack_get(reasm->nfct);
	259	(*pskb)->nfct = reasm->nfct;
	260	return NF_ACCEPT;
	261	}
	262
	263	return nf_conntrack_in(PF_INET6, hooknum, pskb);
	264	}
	265
	266	static unsigned int ipv6_conntrack_local(unsigned int hooknum,
	267	struct sk_buff **pskb,
	268	const struct net_device *in,
	269	const struct net_device *out,
	270	int (okfn)(struct sk_buff ))
	271	{
	272	/* root is playing with raw sockets. */
	273	if ((*pskb)->len < sizeof(struct ipv6hdr)) {
	274	if (net_ratelimit())
	275	printk("ipv6_conntrack_local: packet too short\n");
276	return NF_ACCEPT;
277	}
278	return ipv6_conntrack_in(hooknum, pskb, in, out, okfn);
279	}
280
964ddaa1 PM	281	static struct nf_hook_ops ipv6_conntrack_ops[] = {
	282	{
	283	.hook = ipv6_defrag,
	284	.owner = THIS_MODULE,
	285	.pf = PF_INET6,
	286	.hooknum = NF_IP6_PRE_ROUTING,
	287	.priority = NF_IP6_PRI_CONNTRACK_DEFRAG,
	288	},
	289	{
	290	.hook = ipv6_conntrack_in,
	291	.owner = THIS_MODULE,
	292	.pf = PF_INET6,
	293	.hooknum = NF_IP6_PRE_ROUTING,
	294	.priority = NF_IP6_PRI_CONNTRACK,
	295	},
	296	{
	297	.hook = ipv6_conntrack_local,
	298	.owner = THIS_MODULE,
	299	.pf = PF_INET6,
	300	.hooknum = NF_IP6_LOCAL_OUT,
	301	.priority = NF_IP6_PRI_CONNTRACK,
	302	},
	303	{
	304	.hook = ipv6_defrag,
	305	.owner = THIS_MODULE,
	306	.pf = PF_INET6,
	307	.hooknum = NF_IP6_LOCAL_OUT,
	308	.priority = NF_IP6_PRI_CONNTRACK_DEFRAG,
	309	},
	310	{
	311	.hook = ipv6_confirm,
	312	.owner = THIS_MODULE,
	313	.pf = PF_INET6,
	314	.hooknum = NF_IP6_POST_ROUTING,
	315	.priority = NF_IP6_PRI_LAST,
	316	},
	317	{
	318	.hook = ipv6_confirm,
	319	.owner = THIS_MODULE,
	320	.pf = PF_INET6,
	321	.hooknum = NF_IP6_LOCAL_IN,
	322	.priority = NF_IP6_PRI_LAST-1,
	323	},
9fb9cbb1 YK	324	};
	325
	326	#ifdef CONFIG_SYSCTL
	327
	328	/* From nf_conntrack_proto_icmpv6.c */
babbdb1a	329	extern unsigned int nf_ct_icmpv6_timeout;
9fb9cbb1	330
9fb9cbb1 YK	331	static struct ctl_table_header *nf_ct_ipv6_sysctl_header;
	332
	333	static ctl_table nf_ct_sysctl_table[] = {
	334	{
	335	.ctl_name = NET_NF_CONNTRACK_ICMPV6_TIMEOUT,
	336	.procname = "nf_conntrack_icmpv6_timeout",
	337	.data = &nf_ct_icmpv6_timeout,
	338	.maxlen = sizeof(unsigned int),
	339	.mode = 0644,
	340	.proc_handler = &proc_dointvec_jiffies,
	341	},
	342	{
	343	.ctl_name = NET_NF_CONNTRACK_FRAG6_TIMEOUT,
	344	.procname = "nf_conntrack_frag6_timeout",
	345	.data = &nf_ct_frag6_timeout,
	346	.maxlen = sizeof(unsigned int),
	347	.mode = 0644,
	348	.proc_handler = &proc_dointvec_jiffies,
	349	},
	350	{
	351	.ctl_name = NET_NF_CONNTRACK_FRAG6_LOW_THRESH,
	352	.procname = "nf_conntrack_frag6_low_thresh",
	353	.data = &nf_ct_frag6_low_thresh,
	354	.maxlen = sizeof(unsigned int),
	355	.mode = 0644,
7686a02c	356	.proc_handler = &proc_dointvec,
9fb9cbb1 YK	357	},
	358	{
	359	.ctl_name = NET_NF_CONNTRACK_FRAG6_HIGH_THRESH,
	360	.procname = "nf_conntrack_frag6_high_thresh",
	361	.data = &nf_ct_frag6_high_thresh,
	362	.maxlen = sizeof(unsigned int),
	363	.mode = 0644,
7686a02c	364	.proc_handler = &proc_dointvec,
9fb9cbb1 YK	365	},
	366	{ .ctl_name = 0 }
	367	};
	368
	369	static ctl_table nf_ct_netfilter_table[] = {
	370	{
	371	.ctl_name = NET_NETFILTER,
	372	.procname = "netfilter",
	373	.mode = 0555,
	374	.child = nf_ct_sysctl_table,
	375	},
	376	{ .ctl_name = 0 }
	377	};
	378
	379	static ctl_table nf_ct_net_table[] = {
	380	{
	381	.ctl_name = CTL_NET,
	382	.procname = "net",
	383	.mode = 0555,
	384	.child = nf_ct_netfilter_table,
	385	},
	386	{ .ctl_name = 0 }
	387	};
	388	#endif
	389
c1d10adb PNA	390	#if defined(CONFIG_NF_CT_NETLINK) \|\| \
	391	defined(CONFIG_NF_CT_NETLINK_MODULE)
	392
	393	#include <linux/netfilter/nfnetlink.h>
	394	#include <linux/netfilter/nfnetlink_conntrack.h>
	395
	396	static int ipv6_tuple_to_nfattr(struct sk_buff *skb,
	397	const struct nf_conntrack_tuple *tuple)
	398	{
	399	NFA_PUT(skb, CTA_IP_V6_SRC, sizeof(u_int32_t) * 4,
	400	&tuple->src.u3.ip6);
	401	NFA_PUT(skb, CTA_IP_V6_DST, sizeof(u_int32_t) * 4,
	402	&tuple->dst.u3.ip6);
	403	return 0;
	404
	405	nfattr_failure:
	406	return -1;
	407	}
	408
	409	static const size_t cta_min_ip[CTA_IP_MAX] = {
	410	[CTA_IP_V6_SRC-1] = sizeof(u_int32_t)*4,
	411	[CTA_IP_V6_DST-1] = sizeof(u_int32_t)*4,
	412	};
	413
	414	static int ipv6_nfattr_to_tuple(struct nfattr *tb[],
	415	struct nf_conntrack_tuple *t)
	416	{
	417	if (!tb[CTA_IP_V6_SRC-1] \|\| !tb[CTA_IP_V6_DST-1])
	418	return -EINVAL;
	419
	420	if (nfattr_bad_size(tb, CTA_IP_MAX, cta_min_ip))
	421	return -EINVAL;
	422
	423	memcpy(&t->src.u3.ip6, NFA_DATA(tb[CTA_IP_V6_SRC-1]),
	424	sizeof(u_int32_t) * 4);
	425	memcpy(&t->dst.u3.ip6, NFA_DATA(tb[CTA_IP_V6_DST-1]),
	426	sizeof(u_int32_t) * 4);
	427
	428	return 0;
	429	}
	430	#endif
	431
9fb9cbb1 YK	432	struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv6 = {
	433	.l3proto = PF_INET6,
	434	.name = "ipv6",
	435	.pkt_to_tuple = ipv6_pkt_to_tuple,
	436	.invert_tuple = ipv6_invert_tuple,
	437	.print_tuple = ipv6_print_tuple,
	438	.print_conntrack = ipv6_print_conntrack,
	439	.prepare = ipv6_prepare,
c1d10adb PNA	440	#if defined(CONFIG_NF_CT_NETLINK) \|\| \
	441	defined(CONFIG_NF_CT_NETLINK_MODULE)
	442	.tuple_to_nfattr = ipv6_tuple_to_nfattr,
	443	.nfattr_to_tuple = ipv6_nfattr_to_tuple,
	444	#endif
9fb9cbb1 YK	445	.get_features = ipv6_get_features,
	446	.me = THIS_MODULE,
	447	};
	448
32292a7f PM	449	MODULE_ALIAS("nf_conntrack-" __stringify(AF_INET6));
	450	MODULE_LICENSE("GPL");
	451	MODULE_AUTHOR("Yasuyuki KOZAKAI @USAGI <yasuyuki.kozakai@toshiba.co.jp>");
	452
	453	static int __init nf_conntrack_l3proto_ipv6_init(void)
9fb9cbb1 YK	454	{
	455	int ret = 0;
	456
32292a7f	457	need_conntrack();
9fb9cbb1 YK	458
	459	ret = nf_ct_frag6_init();
	460	if (ret < 0) {
	461	printk("nf_conntrack_ipv6: can't initialize frag6.\n");
32292a7f	462	return ret;
9fb9cbb1	463	}
605dcad6	464	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_tcp6);
9fb9cbb1 YK	465	if (ret < 0) {
	466	printk("nf_conntrack_ipv6: can't register tcp.\n");
	467	goto cleanup_frag6;
	468	}
	469
605dcad6	470	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_udp6);
9fb9cbb1 YK	471	if (ret < 0) {
	472	printk("nf_conntrack_ipv6: can't register udp.\n");
	473	goto cleanup_tcp;
	474	}
	475
605dcad6	476	ret = nf_conntrack_l4proto_register(&nf_conntrack_l4proto_icmpv6);
9fb9cbb1 YK	477	if (ret < 0) {
	478	printk("nf_conntrack_ipv6: can't register icmpv6.\n");
	479	goto cleanup_udp;
	480	}
	481
	482	ret = nf_conntrack_l3proto_register(&nf_conntrack_l3proto_ipv6);
	483	if (ret < 0) {
	484	printk("nf_conntrack_ipv6: can't register ipv6\n");
	485	goto cleanup_icmpv6;
	486	}
	487
964ddaa1 PM	488	ret = nf_register_hooks(ipv6_conntrack_ops,
964ddaa1 PM	489	ARRAY_SIZE(ipv6_conntrack_ops));
9fb9cbb1 YK	490	if (ret < 0) {
	491	printk("nf_conntrack_ipv6: can't register pre-routing defrag "
	492	"hook.\n");
	493	goto cleanup_ipv6;
	494	}
9fb9cbb1 YK	495	#ifdef CONFIG_SYSCTL
	496	nf_ct_ipv6_sysctl_header = register_sysctl_table(nf_ct_net_table, 0);
	497	if (nf_ct_ipv6_sysctl_header == NULL) {
	498	printk("nf_conntrack: can't register to sysctl.\n");
	499	ret = -ENOMEM;
964ddaa1	500	goto cleanup_hooks;
9fb9cbb1 YK	501	}
	502	#endif
	503	return ret;
	504
9fb9cbb1	505	#ifdef CONFIG_SYSCTL
964ddaa1	506	cleanup_hooks:
964ddaa1	507	nf_unregister_hooks(ipv6_conntrack_ops, ARRAY_SIZE(ipv6_conntrack_ops));
32292a7f	508	#endif
9fb9cbb1 YK	509	cleanup_ipv6:
	510	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv6);
	511	cleanup_icmpv6:
605dcad6	512	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmpv6);
9fb9cbb1	513	cleanup_udp:
605dcad6	514	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp6);
9fb9cbb1	515	cleanup_tcp:
605dcad6	516	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp6);
9fb9cbb1 YK	517	cleanup_frag6:
9fb9cbb1 YK	518	nf_ct_frag6_cleanup();
9fb9cbb1 YK	519	return ret;
	520	}
	521
65b4b4e8	522	static void __exit nf_conntrack_l3proto_ipv6_fini(void)
9fb9cbb1	523	{
32292a7f PM	524	synchronize_net();
	525	#ifdef CONFIG_SYSCTL
	526	unregister_sysctl_table(nf_ct_ipv6_sysctl_header);
	527	#endif
	528	nf_unregister_hooks(ipv6_conntrack_ops, ARRAY_SIZE(ipv6_conntrack_ops));
	529	nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv6);
605dcad6 MJ	530	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_icmpv6);
	531	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_udp6);
	532	nf_conntrack_l4proto_unregister(&nf_conntrack_l4proto_tcp6);
32292a7f	533	nf_ct_frag6_cleanup();
9fb9cbb1 YK	534	}
9fb9cbb1 YK	535
65b4b4e8 AM	536	module_init(nf_conntrack_l3proto_ipv6_init);
65b4b4e8 AM	537	module_exit(nf_conntrack_l3proto_ipv6_fini);