[deliverable/linux.git] / net / ipv6 / netfilter / ip6t_SYNPROXY.c

/*
 * Copyright (c) 2013 Patrick McHardy <kaber@trash.net>
 *
 * 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.
 */

#include <linux/module.h>
#include <linux/skbuff.h>
#include <net/ip6_checksum.h>
#include <net/ip6_route.h>
#include <net/tcp.h>

#include <linux/netfilter_ipv6/ip6_tables.h>
#include <linux/netfilter/x_tables.h>
#include <linux/netfilter/xt_SYNPROXY.h>
#include <net/netfilter/nf_conntrack.h>
#include <net/netfilter/nf_conntrack_seqadj.h>
#include <net/netfilter/nf_conntrack_synproxy.h>

static struct ipv6hdr *
synproxy_build_ip(struct sk_buff *skb, const struct in6_addr *saddr,
				       const struct in6_addr *daddr)
{
	struct ipv6hdr *iph;

	skb_reset_network_header(skb);
	iph = (struct ipv6hdr *)skb_put(skb, sizeof(*iph));
	ip6_flow_hdr(iph, 0, 0);
	iph->hop_limit	= 64;	//XXX
	iph->nexthdr	= IPPROTO_TCP;
	iph->saddr	= *saddr;
	iph->daddr	= *daddr;

	return iph;
}

static void
synproxy_send_tcp(const struct sk_buff *skb, struct sk_buff *nskb,
		  struct nf_conntrack *nfct, enum ip_conntrack_info ctinfo,
		  struct ipv6hdr *niph, struct tcphdr *nth,
		  unsigned int tcp_hdr_size)
{
	struct net *net = nf_ct_net((struct nf_conn *)nfct);
	struct dst_entry *dst;
	struct flowi6 fl6;

	nth->check = ~tcp_v6_check(tcp_hdr_size, &niph->saddr, &niph->daddr, 0);
	nskb->ip_summed   = CHECKSUM_PARTIAL;
	nskb->csum_start  = (unsigned char *)nth - nskb->head;
	nskb->csum_offset = offsetof(struct tcphdr, check);

	memset(&fl6, 0, sizeof(fl6));
	fl6.flowi6_proto = IPPROTO_TCP;
	fl6.saddr = niph->saddr;
	fl6.daddr = niph->daddr;
	fl6.fl6_sport = nth->source;
	fl6.fl6_dport = nth->dest;
	security_skb_classify_flow((struct sk_buff *)skb, flowi6_to_flowi(&fl6));
	dst = ip6_route_output(net, NULL, &fl6);
	if (dst == NULL || dst->error) {
		dst_release(dst);
		goto free_nskb;
	}
	dst = xfrm_lookup(net, dst, flowi6_to_flowi(&fl6), NULL, 0);
	if (IS_ERR(dst))
		goto free_nskb;

	skb_dst_set(nskb, dst);

	if (nfct) {
		nskb->nfct = nfct;
		nskb->nfctinfo = ctinfo;
		nf_conntrack_get(nfct);
	}

	ip6_local_out(nskb);
	return;

free_nskb:
	kfree_skb(nskb);
}

static void
synproxy_send_client_synack(const struct sk_buff *skb, const struct tcphdr *th,
			    const struct synproxy_options *opts)
{
	struct sk_buff *nskb;
	struct ipv6hdr *iph, *niph;
	struct tcphdr *nth;
	unsigned int tcp_hdr_size;
	u16 mss = opts->mss;

	iph = ipv6_hdr(skb);

	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
			 GFP_ATOMIC);
	if (nskb == NULL)
		return;
	skb_reserve(nskb, MAX_TCP_HEADER);

	niph = synproxy_build_ip(nskb, &iph->daddr, &iph->saddr);

	skb_reset_transport_header(nskb);
	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
	nth->source	= th->dest;
	nth->dest	= th->source;
	nth->seq	= htonl(__cookie_v6_init_sequence(iph, th, &mss));
	nth->ack_seq	= htonl(ntohl(th->seq) + 1);
	tcp_flag_word(nth) = TCP_FLAG_SYN | TCP_FLAG_ACK;
	if (opts->options & XT_SYNPROXY_OPT_ECN)
		tcp_flag_word(nth) |= TCP_FLAG_ECE;
	nth->doff	= tcp_hdr_size / 4;
	nth->window	= 0;
	nth->check	= 0;
	nth->urg_ptr	= 0;

	synproxy_build_options(nth, opts);

	synproxy_send_tcp(skb, nskb, skb->nfct, IP_CT_ESTABLISHED_REPLY,
			  niph, nth, tcp_hdr_size);
}

static void
synproxy_send_server_syn(const struct synproxy_net *snet,
			 const struct sk_buff *skb, const struct tcphdr *th,
			 const struct synproxy_options *opts, u32 recv_seq)
{
	struct sk_buff *nskb;
	struct ipv6hdr *iph, *niph;
	struct tcphdr *nth;
	unsigned int tcp_hdr_size;

	iph = ipv6_hdr(skb);

	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
			 GFP_ATOMIC);
	if (nskb == NULL)
		return;
	skb_reserve(nskb, MAX_TCP_HEADER);

	niph = synproxy_build_ip(nskb, &iph->saddr, &iph->daddr);

	skb_reset_transport_header(nskb);
	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
	nth->source	= th->source;
	nth->dest	= th->dest;
	nth->seq	= htonl(recv_seq - 1);
	/* ack_seq is used to relay our ISN to the synproxy hook to initialize
	 * sequence number translation once a connection tracking entry exists.
	 */
	nth->ack_seq	= htonl(ntohl(th->ack_seq) - 1);
	tcp_flag_word(nth) = TCP_FLAG_SYN;
	if (opts->options & XT_SYNPROXY_OPT_ECN)
		tcp_flag_word(nth) |= TCP_FLAG_ECE | TCP_FLAG_CWR;
	nth->doff	= tcp_hdr_size / 4;
	nth->window	= th->window;
	nth->check	= 0;
	nth->urg_ptr	= 0;

	synproxy_build_options(nth, opts);

	synproxy_send_tcp(skb, nskb, &snet->tmpl->ct_general, IP_CT_NEW,
			  niph, nth, tcp_hdr_size);
}

static void
synproxy_send_server_ack(const struct synproxy_net *snet,
			 const struct ip_ct_tcp *state,
			 const struct sk_buff *skb, const struct tcphdr *th,
			 const struct synproxy_options *opts)
{
	struct sk_buff *nskb;
	struct ipv6hdr *iph, *niph;
	struct tcphdr *nth;
	unsigned int tcp_hdr_size;

	iph = ipv6_hdr(skb);

	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
			 GFP_ATOMIC);
	if (nskb == NULL)
		return;
	skb_reserve(nskb, MAX_TCP_HEADER);

	niph = synproxy_build_ip(nskb, &iph->daddr, &iph->saddr);

	skb_reset_transport_header(nskb);
	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
	nth->source	= th->dest;
	nth->dest	= th->source;
	nth->seq	= htonl(ntohl(th->ack_seq));
	nth->ack_seq	= htonl(ntohl(th->seq) + 1);
	tcp_flag_word(nth) = TCP_FLAG_ACK;
	nth->doff	= tcp_hdr_size / 4;
	nth->window	= htons(state->seen[IP_CT_DIR_ORIGINAL].td_maxwin);
	nth->check	= 0;
	nth->urg_ptr	= 0;

	synproxy_build_options(nth, opts);

	synproxy_send_tcp(skb, nskb, NULL, 0, niph, nth, tcp_hdr_size);
}

static void
synproxy_send_client_ack(const struct synproxy_net *snet,
			 const struct sk_buff *skb, const struct tcphdr *th,
			 const struct synproxy_options *opts)
{
	struct sk_buff *nskb;
	struct ipv6hdr *iph, *niph;
	struct tcphdr *nth;
	unsigned int tcp_hdr_size;

	iph = ipv6_hdr(skb);

	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
			 GFP_ATOMIC);
	if (nskb == NULL)
		return;
	skb_reserve(nskb, MAX_TCP_HEADER);

	niph = synproxy_build_ip(nskb, &iph->saddr, &iph->daddr);

	skb_reset_transport_header(nskb);
	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
	nth->source	= th->source;
	nth->dest	= th->dest;
	nth->seq	= htonl(ntohl(th->seq) + 1);
	nth->ack_seq	= th->ack_seq;
	tcp_flag_word(nth) = TCP_FLAG_ACK;
	nth->doff	= tcp_hdr_size / 4;
	nth->window	= ntohs(htons(th->window) >> opts->wscale);
	nth->check	= 0;
	nth->urg_ptr	= 0;

	synproxy_build_options(nth, opts);

	synproxy_send_tcp(skb, nskb, NULL, 0, niph, nth, tcp_hdr_size);
}

static bool
synproxy_recv_client_ack(const struct synproxy_net *snet,
			 const struct sk_buff *skb, const struct tcphdr *th,
			 struct synproxy_options *opts, u32 recv_seq)
{
	int mss;

	mss = __cookie_v6_check(ipv6_hdr(skb), th, ntohl(th->ack_seq) - 1);
	if (mss == 0) {
		this_cpu_inc(snet->stats->cookie_invalid);
		return false;
	}

	this_cpu_inc(snet->stats->cookie_valid);
	opts->mss = mss;

	if (opts->options & XT_SYNPROXY_OPT_TIMESTAMP)
		synproxy_check_timestamp_cookie(opts);

	synproxy_send_server_syn(snet, skb, th, opts, recv_seq);
	return true;
}

static unsigned int
synproxy_tg6(struct sk_buff *skb, const struct xt_action_param *par)
{
	const struct xt_synproxy_info *info = par->targinfo;
	struct synproxy_net *snet = synproxy_pernet(dev_net(par->in));
	struct synproxy_options opts = {};
	struct tcphdr *th, _th;

	if (nf_ip6_checksum(skb, par->hooknum, par->thoff, IPPROTO_TCP))
		return NF_DROP;

	th = skb_header_pointer(skb, par->thoff, sizeof(_th), &_th);
	if (th == NULL)
		return NF_DROP;

	if (!synproxy_parse_options(skb, par->thoff, th, &opts))
		return NF_DROP;

	if (th->syn && !(th->ack || th->fin || th->rst)) {
		/* Initial SYN from client */
		this_cpu_inc(snet->stats->syn_received);

		if (th->ece && th->cwr)
			opts.options |= XT_SYNPROXY_OPT_ECN;

		opts.options &= info->options;
		if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
			synproxy_init_timestamp_cookie(info, &opts);
		else
			opts.options &= ~(XT_SYNPROXY_OPT_WSCALE |
					  XT_SYNPROXY_OPT_SACK_PERM |
					  XT_SYNPROXY_OPT_ECN);

		synproxy_send_client_synack(skb, th, &opts);
		return NF_DROP;

	} else if (th->ack && !(th->fin || th->rst || th->syn)) {
		/* ACK from client */
		synproxy_recv_client_ack(snet, skb, th, &opts, ntohl(th->seq));
		return NF_DROP;
	}

	return XT_CONTINUE;
}

static unsigned int ipv6_synproxy_hook(unsigned int hooknum,
				       struct sk_buff *skb,
				       const struct net_device *in,
				       const struct net_device *out,
				       int (*okfn)(struct sk_buff *))
{
	struct synproxy_net *snet = synproxy_pernet(dev_net(in ? : out));
	enum ip_conntrack_info ctinfo;
	struct nf_conn *ct;
	struct nf_conn_synproxy *synproxy;
	struct synproxy_options opts = {};
	const struct ip_ct_tcp *state;
	struct tcphdr *th, _th;
	__be16 frag_off;
	u8 nexthdr;
	int thoff;

	ct = nf_ct_get(skb, &ctinfo);
	if (ct == NULL)
		return NF_ACCEPT;

	synproxy = nfct_synproxy(ct);
	if (synproxy == NULL)
		return NF_ACCEPT;

	if (nf_is_loopback_packet(skb))
		return NF_ACCEPT;

	nexthdr = ipv6_hdr(skb)->nexthdr;
	thoff = ipv6_skip_exthdr(skb, sizeof(struct ipv6hdr), &nexthdr,
				 &frag_off);
	if (thoff < 0)
		return NF_ACCEPT;

	th = skb_header_pointer(skb, thoff, sizeof(_th), &_th);
	if (th == NULL)
		return NF_DROP;

	state = &ct->proto.tcp;
	switch (state->state) {
	case TCP_CONNTRACK_CLOSE:
		if (th->rst && !test_bit(IPS_SEEN_REPLY_BIT, &ct->status)) {
			nf_ct_seqadj_init(ct, ctinfo, synproxy->isn -
						      ntohl(th->seq) + 1);
			break;
		}

		if (!th->syn || th->ack ||
		    CTINFO2DIR(ctinfo) != IP_CT_DIR_ORIGINAL)
			break;

		/* Reopened connection - reset the sequence number and timestamp
		 * adjustments, they will get initialized once the connection is
		 * reestablished.
		 */
		nf_ct_seqadj_init(ct, ctinfo, 0);
		synproxy->tsoff = 0;
		this_cpu_inc(snet->stats->conn_reopened);

		/* fall through */
	case TCP_CONNTRACK_SYN_SENT:
		if (!synproxy_parse_options(skb, thoff, th, &opts))
			return NF_DROP;

		if (!th->syn && th->ack &&
		    CTINFO2DIR(ctinfo) == IP_CT_DIR_ORIGINAL) {
			/* Keep-Alives are sent with SEG.SEQ = SND.NXT-1,
			 * therefore we need to add 1 to make the SYN sequence
			 * number match the one of first SYN.
			 */
			if (synproxy_recv_client_ack(snet, skb, th, &opts,
						     ntohl(th->seq) + 1))
				this_cpu_inc(snet->stats->cookie_retrans);

			return NF_DROP;
		}

		synproxy->isn = ntohl(th->ack_seq);
		if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
			synproxy->its = opts.tsecr;
		break;
	case TCP_CONNTRACK_SYN_RECV:
		if (!th->syn || !th->ack)
			break;

		if (!synproxy_parse_options(skb, thoff, th, &opts))
			return NF_DROP;

		if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
			synproxy->tsoff = opts.tsval - synproxy->its;

		opts.options &= ~(XT_SYNPROXY_OPT_MSS |
				  XT_SYNPROXY_OPT_WSCALE |
				  XT_SYNPROXY_OPT_SACK_PERM);

		swap(opts.tsval, opts.tsecr);
		synproxy_send_server_ack(snet, state, skb, th, &opts);

		nf_ct_seqadj_init(ct, ctinfo, synproxy->isn - ntohl(th->seq));

		swap(opts.tsval, opts.tsecr);
		synproxy_send_client_ack(snet, skb, th, &opts);

		consume_skb(skb);
		return NF_STOLEN;
	default:
		break;
	}

	synproxy_tstamp_adjust(skb, thoff, th, ct, ctinfo, synproxy);
	return NF_ACCEPT;
}

static int synproxy_tg6_check(const struct xt_tgchk_param *par)
{
	const struct ip6t_entry *e = par->entryinfo;

	if (!(e->ipv6.flags & IP6T_F_PROTO) ||
	    e->ipv6.proto != IPPROTO_TCP ||
	    e->ipv6.invflags & XT_INV_PROTO)
		return -EINVAL;

	return nf_ct_l3proto_try_module_get(par->family);
}

static void synproxy_tg6_destroy(const struct xt_tgdtor_param *par)
{
	nf_ct_l3proto_module_put(par->family);
}

static struct xt_target synproxy_tg6_reg __read_mostly = {
	.name		= "SYNPROXY",
	.family		= NFPROTO_IPV6,
	.target		= synproxy_tg6,
	.targetsize	= sizeof(struct xt_synproxy_info),
	.checkentry	= synproxy_tg6_check,
	.destroy	= synproxy_tg6_destroy,
	.me		= THIS_MODULE,
};

static struct nf_hook_ops ipv6_synproxy_ops[] __read_mostly = {
	{
		.hook		= ipv6_synproxy_hook,
		.owner		= THIS_MODULE,
		.pf		= NFPROTO_IPV6,
		.hooknum	= NF_INET_LOCAL_IN,
		.priority	= NF_IP_PRI_CONNTRACK_CONFIRM - 1,
	},
	{
		.hook		= ipv6_synproxy_hook,
		.owner		= THIS_MODULE,
		.pf		= NFPROTO_IPV6,
		.hooknum	= NF_INET_POST_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK_CONFIRM - 1,
	},
};

static int __init synproxy_tg6_init(void)
{
	int err;

	err = nf_register_hooks(ipv6_synproxy_ops,
				ARRAY_SIZE(ipv6_synproxy_ops));
	if (err < 0)
		goto err1;

	err = xt_register_target(&synproxy_tg6_reg);
	if (err < 0)
		goto err2;

	return 0;

err2:
	nf_unregister_hooks(ipv6_synproxy_ops, ARRAY_SIZE(ipv6_synproxy_ops));
err1:
	return err;
}

static void __exit synproxy_tg6_exit(void)
{
	xt_unregister_target(&synproxy_tg6_reg);
	nf_unregister_hooks(ipv6_synproxy_ops, ARRAY_SIZE(ipv6_synproxy_ops));
}

module_init(synproxy_tg6_init);
module_exit(synproxy_tg6_exit);

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Patrick McHardy <kaber@trash.net>");
Commit	Line	Data
4ad36228 PM	1	/*
	2	* Copyright (c) 2013 Patrick McHardy <kaber@trash.net>
	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
	9	#include <linux/module.h>
	10	#include <linux/skbuff.h>
	11	#include <net/ip6_checksum.h>
	12	#include <net/ip6_route.h>
	13	#include <net/tcp.h>
	14
	15	#include <linux/netfilter_ipv6/ip6_tables.h>
	16	#include <linux/netfilter/x_tables.h>
	17	#include <linux/netfilter/xt_SYNPROXY.h>
	18	#include <net/netfilter/nf_conntrack.h>
	19	#include <net/netfilter/nf_conntrack_seqadj.h>
	20	#include <net/netfilter/nf_conntrack_synproxy.h>
	21
	22	static struct ipv6hdr *
	23	synproxy_build_ip(struct sk_buff skb, const struct in6_addr saddr,
	24	const struct in6_addr *daddr)
	25	{
	26	struct ipv6hdr *iph;
	27
	28	skb_reset_network_header(skb);
	29	iph = (struct ipv6hdr )skb_put(skb, sizeof(iph));
	30	ip6_flow_hdr(iph, 0, 0);
	31	iph->hop_limit = 64; //XXX
	32	iph->nexthdr = IPPROTO_TCP;
	33	iph->saddr = *saddr;
	34	iph->daddr = *daddr;
	35
	36	return iph;
	37	}
	38
	39	static void
	40	synproxy_send_tcp(const struct sk_buff skb, struct sk_buff nskb,
	41	struct nf_conntrack *nfct, enum ip_conntrack_info ctinfo,
	42	struct ipv6hdr niph, struct tcphdr nth,
	43	unsigned int tcp_hdr_size)
	44	{
	45	struct net net = nf_ct_net((struct nf_conn )nfct);
	46	struct dst_entry *dst;
	47	struct flowi6 fl6;
	48
	49	nth->check = ~tcp_v6_check(tcp_hdr_size, &niph->saddr, &niph->daddr, 0);
	50	nskb->ip_summed = CHECKSUM_PARTIAL;
	51	nskb->csum_start = (unsigned char *)nth - nskb->head;
	52	nskb->csum_offset = offsetof(struct tcphdr, check);
	53
	54	memset(&fl6, 0, sizeof(fl6));
	55	fl6.flowi6_proto = IPPROTO_TCP;
	56	fl6.saddr = niph->saddr;
	57	fl6.daddr = niph->daddr;
	58	fl6.fl6_sport = nth->source;
	59	fl6.fl6_dport = nth->dest;
	60	security_skb_classify_flow((struct sk_buff *)skb, flowi6_to_flowi(&fl6));
	61	dst = ip6_route_output(net, NULL, &fl6);
	62	if (dst == NULL \|\| dst->error) {
	63	dst_release(dst);
	64	goto free_nskb;
65	}
66	dst = xfrm_lookup(net, dst, flowi6_to_flowi(&fl6), NULL, 0);
67	if (IS_ERR(dst))
68	goto free_nskb;
69
70	skb_dst_set(nskb, dst);
71
72	if (nfct) {
73	nskb->nfct = nfct;
74	nskb->nfctinfo = ctinfo;
75	nf_conntrack_get(nfct);
76	}
77
78	ip6_local_out(nskb);
79	return;
80
81	free_nskb:
82	kfree_skb(nskb);
83	}
84
85	static void
86	synproxy_send_client_synack(const struct sk_buff skb, const struct tcphdr th,
87	const struct synproxy_options *opts)
88	{
89	struct sk_buff *nskb;
90	struct ipv6hdr iph, niph;
91	struct tcphdr *nth;
92	unsigned int tcp_hdr_size;
93	u16 mss = opts->mss;
94
95	iph = ipv6_hdr(skb);
96
97	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
98	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
99	GFP_ATOMIC);
100	if (nskb == NULL)
101	return;
102	skb_reserve(nskb, MAX_TCP_HEADER);
103
104	niph = synproxy_build_ip(nskb, &iph->daddr, &iph->saddr);
105
106	skb_reset_transport_header(nskb);
107	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
108	nth->source = th->dest;
109	nth->dest = th->source;
110	nth->seq = htonl(__cookie_v6_init_sequence(iph, th, &mss));
111	nth->ack_seq = htonl(ntohl(th->seq) + 1);
112	tcp_flag_word(nth) = TCP_FLAG_SYN \| TCP_FLAG_ACK;
113	if (opts->options & XT_SYNPROXY_OPT_ECN)
114	tcp_flag_word(nth) \|= TCP_FLAG_ECE;
115	nth->doff = tcp_hdr_size / 4;
116	nth->window = 0;
117	nth->check = 0;
118	nth->urg_ptr = 0;
119
120	synproxy_build_options(nth, opts);
121
122	synproxy_send_tcp(skb, nskb, skb->nfct, IP_CT_ESTABLISHED_REPLY,
123	niph, nth, tcp_hdr_size);
124	}
125
126	static void
127	synproxy_send_server_syn(const struct synproxy_net *snet,
128	const struct sk_buff skb, const struct tcphdr th,
129	const struct synproxy_options *opts, u32 recv_seq)
130	{
131	struct sk_buff *nskb;
132	struct ipv6hdr iph, niph;
133	struct tcphdr *nth;
134	unsigned int tcp_hdr_size;
135
136	iph = ipv6_hdr(skb);
137
138	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
139	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
140	GFP_ATOMIC);
141	if (nskb == NULL)
142	return;
143	skb_reserve(nskb, MAX_TCP_HEADER);
144
145	niph = synproxy_build_ip(nskb, &iph->saddr, &iph->daddr);
146
147	skb_reset_transport_header(nskb);
148	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
149	nth->source = th->source;
150	nth->dest = th->dest;
151	nth->seq = htonl(recv_seq - 1);
152	/* ack_seq is used to relay our ISN to the synproxy hook to initialize
153	* sequence number translation once a connection tracking entry exists.
154	*/
155	nth->ack_seq = htonl(ntohl(th->ack_seq) - 1);
156	tcp_flag_word(nth) = TCP_FLAG_SYN;
157	if (opts->options & XT_SYNPROXY_OPT_ECN)
158	tcp_flag_word(nth) \|= TCP_FLAG_ECE \| TCP_FLAG_CWR;
159	nth->doff = tcp_hdr_size / 4;
160	nth->window = th->window;
161	nth->check = 0;
162	nth->urg_ptr = 0;
163
164	synproxy_build_options(nth, opts);
165
166	synproxy_send_tcp(skb, nskb, &snet->tmpl->ct_general, IP_CT_NEW,
167	niph, nth, tcp_hdr_size);
168	}
169
170	static void
171	synproxy_send_server_ack(const struct synproxy_net *snet,
172	const struct ip_ct_tcp *state,
173	const struct sk_buff skb, const struct tcphdr th,
174	const struct synproxy_options *opts)
175	{
176	struct sk_buff *nskb;
177	struct ipv6hdr iph, niph;
178	struct tcphdr *nth;
179	unsigned int tcp_hdr_size;
180
181	iph = ipv6_hdr(skb);
182
183	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
184	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
185	GFP_ATOMIC);
186	if (nskb == NULL)
187	return;
188	skb_reserve(nskb, MAX_TCP_HEADER);
189
190	niph = synproxy_build_ip(nskb, &iph->daddr, &iph->saddr);
191
192	skb_reset_transport_header(nskb);
193	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
194	nth->source = th->dest;
195	nth->dest = th->source;
196	nth->seq = htonl(ntohl(th->ack_seq));
197	nth->ack_seq = htonl(ntohl(th->seq) + 1);
198	tcp_flag_word(nth) = TCP_FLAG_ACK;
199	nth->doff = tcp_hdr_size / 4;
200	nth->window = htons(state->seen[IP_CT_DIR_ORIGINAL].td_maxwin);
201	nth->check = 0;
202	nth->urg_ptr = 0;
203
204	synproxy_build_options(nth, opts);
205
206	synproxy_send_tcp(skb, nskb, NULL, 0, niph, nth, tcp_hdr_size);
207	}
208
209	static void
210	synproxy_send_client_ack(const struct synproxy_net *snet,
211	const struct sk_buff skb, const struct tcphdr th,
212	const struct synproxy_options *opts)
213	{
214	struct sk_buff *nskb;
215	struct ipv6hdr iph, niph;
216	struct tcphdr *nth;
217	unsigned int tcp_hdr_size;
218
219	iph = ipv6_hdr(skb);
220
221	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
222	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
223	GFP_ATOMIC);
224	if (nskb == NULL)
225	return;
226	skb_reserve(nskb, MAX_TCP_HEADER);
227
228	niph = synproxy_build_ip(nskb, &iph->saddr, &iph->daddr);
229
230	skb_reset_transport_header(nskb);
231	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
232	nth->source = th->source;
233	nth->dest = th->dest;
234	nth->seq = htonl(ntohl(th->seq) + 1);
235	nth->ack_seq = th->ack_seq;
236	tcp_flag_word(nth) = TCP_FLAG_ACK;
237	nth->doff = tcp_hdr_size / 4;
238	nth->window = ntohs(htons(th->window) >> opts->wscale);
239	nth->check = 0;
240	nth->urg_ptr = 0;
241
242	synproxy_build_options(nth, opts);
243
244	synproxy_send_tcp(skb, nskb, NULL, 0, niph, nth, tcp_hdr_size);
245	}
246
247	static bool
248	synproxy_recv_client_ack(const struct synproxy_net *snet,
249	const struct sk_buff skb, const struct tcphdr th,
250	struct synproxy_options *opts, u32 recv_seq)
251	{
252	int mss;
253
254	mss = __cookie_v6_check(ipv6_hdr(skb), th, ntohl(th->ack_seq) - 1);
255	if (mss == 0) {
256	this_cpu_inc(snet->stats->cookie_invalid);
257	return false;
258	}
259
260	this_cpu_inc(snet->stats->cookie_valid);
261	opts->mss = mss;
262
263	if (opts->options & XT_SYNPROXY_OPT_TIMESTAMP)
264	synproxy_check_timestamp_cookie(opts);
265
266	synproxy_send_server_syn(snet, skb, th, opts, recv_seq);
267	return true;
268	}
269
270	static unsigned int
271	synproxy_tg6(struct sk_buff skb, const struct xt_action_param par)
272	{
273	const struct xt_synproxy_info *info = par->targinfo;
274	struct synproxy_net *snet = synproxy_pernet(dev_net(par->in));
275	struct synproxy_options opts = {};
276	struct tcphdr *th, _th;
277
278	if (nf_ip6_checksum(skb, par->hooknum, par->thoff, IPPROTO_TCP))
279	return NF_DROP;
280
281	th = skb_header_pointer(skb, par->thoff, sizeof(_th), &_th);
282	if (th == NULL)
283	return NF_DROP;
284
f4a87e7b PM	285	if (!synproxy_parse_options(skb, par->thoff, th, &opts))
f4a87e7b PM	286	return NF_DROP;
4ad36228	287
775ada6d	288	if (th->syn && !(th->ack \|\| th->fin \|\| th->rst)) {
4ad36228 PM	289	/* Initial SYN from client */
	290	this_cpu_inc(snet->stats->syn_received);
	291
	292	if (th->ece && th->cwr)
	293	opts.options \|= XT_SYNPROXY_OPT_ECN;
	294
	295	opts.options &= info->options;
	296	if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
	297	synproxy_init_timestamp_cookie(info, &opts);
	298	else
	299	opts.options &= ~(XT_SYNPROXY_OPT_WSCALE \|
	300	XT_SYNPROXY_OPT_SACK_PERM \|
	301	XT_SYNPROXY_OPT_ECN);
	302
	303	synproxy_send_client_synack(skb, th, &opts);
7cc9eb6e JDB	304	return NF_DROP;
	305
	306	} else if (th->ack && !(th->fin \|\| th->rst \|\| th->syn)) {
4ad36228 PM	307	/* ACK from client */
4ad36228 PM	308	synproxy_recv_client_ack(snet, skb, th, &opts, ntohl(th->seq));
7cc9eb6e JDB	309	return NF_DROP;
7cc9eb6e JDB	310	}
4ad36228	311
7cc9eb6e	312	return XT_CONTINUE;
4ad36228 PM	313	}
	314
	315	static unsigned int ipv6_synproxy_hook(unsigned int hooknum,
	316	struct sk_buff *skb,
	317	const struct net_device *in,
	318	const struct net_device *out,
	319	int (okfn)(struct sk_buff ))
	320	{
	321	struct synproxy_net *snet = synproxy_pernet(dev_net(in ? : out));
	322	enum ip_conntrack_info ctinfo;
	323	struct nf_conn *ct;
	324	struct nf_conn_synproxy *synproxy;
	325	struct synproxy_options opts = {};
	326	const struct ip_ct_tcp *state;
	327	struct tcphdr *th, _th;
	328	__be16 frag_off;
	329	u8 nexthdr;
	330	int thoff;
	331
	332	ct = nf_ct_get(skb, &ctinfo);
	333	if (ct == NULL)
	334	return NF_ACCEPT;
	335
	336	synproxy = nfct_synproxy(ct);
	337	if (synproxy == NULL)
	338	return NF_ACCEPT;
	339
	340	if (nf_is_loopback_packet(skb))
	341	return NF_ACCEPT;
	342
	343	nexthdr = ipv6_hdr(skb)->nexthdr;
	344	thoff = ipv6_skip_exthdr(skb, sizeof(struct ipv6hdr), &nexthdr,
	345	&frag_off);
	346	if (thoff < 0)
	347	return NF_ACCEPT;
	348
	349	th = skb_header_pointer(skb, thoff, sizeof(_th), &_th);
	350	if (th == NULL)
	351	return NF_DROP;
	352
	353	state = &ct->proto.tcp;
	354	switch (state->state) {
	355	case TCP_CONNTRACK_CLOSE:
	356	if (th->rst && !test_bit(IPS_SEEN_REPLY_BIT, &ct->status)) {
	357	nf_ct_seqadj_init(ct, ctinfo, synproxy->isn -
	358	ntohl(th->seq) + 1);
	359	break;
	360	}
	361
	362	if (!th->syn \|\| th->ack \|\|
	363	CTINFO2DIR(ctinfo) != IP_CT_DIR_ORIGINAL)
	364	break;
	365
	366	/* Reopened connection - reset the sequence number and timestamp
	367	* adjustments, they will get initialized once the connection is
	368	* reestablished.
	369	*/
	370	nf_ct_seqadj_init(ct, ctinfo, 0);
	371	synproxy->tsoff = 0;
	372	this_cpu_inc(snet->stats->conn_reopened);
	373
	374	/* fall through */
	375	case TCP_CONNTRACK_SYN_SENT:
f4a87e7b PM	376	if (!synproxy_parse_options(skb, thoff, th, &opts))
f4a87e7b PM	377	return NF_DROP;
4ad36228 PM	378
	379	if (!th->syn && th->ack &&
	380	CTINFO2DIR(ctinfo) == IP_CT_DIR_ORIGINAL) {
	381	/* Keep-Alives are sent with SEG.SEQ = SND.NXT-1,
	382	* therefore we need to add 1 to make the SYN sequence
	383	* number match the one of first SYN.
	384	*/
	385	if (synproxy_recv_client_ack(snet, skb, th, &opts,
	386	ntohl(th->seq) + 1))
	387	this_cpu_inc(snet->stats->cookie_retrans);
	388
	389	return NF_DROP;
	390	}
	391
	392	synproxy->isn = ntohl(th->ack_seq);
	393	if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
	394	synproxy->its = opts.tsecr;
	395	break;
	396	case TCP_CONNTRACK_SYN_RECV:
	397	if (!th->syn \|\| !th->ack)
	398	break;
	399
f4a87e7b PM	400	if (!synproxy_parse_options(skb, thoff, th, &opts))
	401	return NF_DROP;
	402
4ad36228 PM	403	if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
	404	synproxy->tsoff = opts.tsval - synproxy->its;
	405
	406	opts.options &= ~(XT_SYNPROXY_OPT_MSS \|
	407	XT_SYNPROXY_OPT_WSCALE \|
	408	XT_SYNPROXY_OPT_SACK_PERM);
	409
	410	swap(opts.tsval, opts.tsecr);
	411	synproxy_send_server_ack(snet, state, skb, th, &opts);
	412
	413	nf_ct_seqadj_init(ct, ctinfo, synproxy->isn - ntohl(th->seq));
	414
	415	swap(opts.tsval, opts.tsecr);
	416	synproxy_send_client_ack(snet, skb, th, &opts);
	417
	418	consume_skb(skb);
	419	return NF_STOLEN;
	420	default:
	421	break;
	422	}
	423
	424	synproxy_tstamp_adjust(skb, thoff, th, ct, ctinfo, synproxy);
	425	return NF_ACCEPT;
	426	}
	427
	428	static int synproxy_tg6_check(const struct xt_tgchk_param *par)
	429	{
	430	const struct ip6t_entry *e = par->entryinfo;
	431
	432	if (!(e->ipv6.flags & IP6T_F_PROTO) \|\|
	433	e->ipv6.proto != IPPROTO_TCP \|\|
	434	e->ipv6.invflags & XT_INV_PROTO)
	435	return -EINVAL;
	436
	437	return nf_ct_l3proto_try_module_get(par->family);
	438	}
	439
	440	static void synproxy_tg6_destroy(const struct xt_tgdtor_param *par)
	441	{
	442	nf_ct_l3proto_module_put(par->family);
	443	}
	444
	445	static struct xt_target synproxy_tg6_reg __read_mostly = {
	446	.name = "SYNPROXY",
	447	.family = NFPROTO_IPV6,
	448	.target = synproxy_tg6,
	449	.targetsize = sizeof(struct xt_synproxy_info),
	450	.checkentry = synproxy_tg6_check,
	451	.destroy = synproxy_tg6_destroy,
	452	.me = THIS_MODULE,
	453	};
	454
	455	static struct nf_hook_ops ipv6_synproxy_ops[] __read_mostly = {
	456	{
	457	.hook = ipv6_synproxy_hook,
	458	.owner = THIS_MODULE,
	459	.pf = NFPROTO_IPV6,
	460	.hooknum = NF_INET_LOCAL_IN,
	461	.priority = NF_IP_PRI_CONNTRACK_CONFIRM - 1,
	462	},
	463	{
	464	.hook = ipv6_synproxy_hook,
	465	.owner = THIS_MODULE,
	466	.pf = NFPROTO_IPV6,
467	.hooknum = NF_INET_POST_ROUTING,
468	.priority = NF_IP_PRI_CONNTRACK_CONFIRM - 1,
469	},
470	};
471
472	static int __init synproxy_tg6_init(void)
473	{
474	int err;
475
476	err = nf_register_hooks(ipv6_synproxy_ops,
477	ARRAY_SIZE(ipv6_synproxy_ops));
478	if (err < 0)
479	goto err1;
480
481	err = xt_register_target(&synproxy_tg6_reg);
482	if (err < 0)
483	goto err2;
484
485	return 0;
486
487	err2:
488	nf_unregister_hooks(ipv6_synproxy_ops, ARRAY_SIZE(ipv6_synproxy_ops));
489	err1:
490	return err;
491	}
492
493	static void __exit synproxy_tg6_exit(void)
494	{
495	xt_unregister_target(&synproxy_tg6_reg);
496	nf_unregister_hooks(ipv6_synproxy_ops, ARRAY_SIZE(ipv6_synproxy_ops));
497	}
498
499	module_init(synproxy_tg6_init);
500	module_exit(synproxy_tg6_exit);
501
502	MODULE_LICENSE("GPL");
503	MODULE_AUTHOR("Patrick McHardy <kaber@trash.net>");