[deliverable/linux.git] / net / ipv4 / netfilter / ipt_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/tcp.h>

#include <linux/netfilter_ipv4/ip_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 iphdr *
synproxy_build_ip(struct sk_buff *skb, u32 saddr, u32 daddr)
{
	struct iphdr *iph;

	skb_reset_network_header(skb);
	iph = (struct iphdr *)skb_put(skb, sizeof(*iph));
	iph->version	= 4;
	iph->ihl	= sizeof(*iph) / 4;
	iph->tos	= 0;
	iph->id		= 0;
	iph->frag_off	= htons(IP_DF);
	iph->ttl	= sysctl_ip_default_ttl;
	iph->protocol	= IPPROTO_TCP;
	iph->check	= 0;
	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 iphdr *niph, struct tcphdr *nth,
		  unsigned int tcp_hdr_size)
{
	nth->check = ~tcp_v4_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);

	skb_dst_set_noref(nskb, skb_dst(skb));
	nskb->protocol = htons(ETH_P_IP);
	if (ip_route_me_harder(nskb, RTN_UNSPEC))
		goto free_nskb;

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

	ip_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 iphdr *iph, *niph;
	struct tcphdr *nth;
	unsigned int tcp_hdr_size;
	u16 mss = opts->mss;

	iph = ip_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_v4_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 iphdr *iph, *niph;
	struct tcphdr *nth;
	unsigned int tcp_hdr_size;

	iph = ip_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 iphdr *iph, *niph;
	struct tcphdr *nth;
	unsigned int tcp_hdr_size;

	iph = ip_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 iphdr *iph, *niph;
	struct tcphdr *nth;
	unsigned int tcp_hdr_size;

	iph = ip_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_v4_check(ip_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_tg4(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_ip_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;

	synproxy_parse_options(skb, par->thoff, th, &opts);

	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 ipv4_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;
	unsigned 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;

	thoff = ip_hdrlen(skb);
	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:
		synproxy_parse_options(skb, thoff, th, &opts);

		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;

		synproxy_parse_options(skb, thoff, th, &opts);
		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_tg4_check(const struct xt_tgchk_param *par)
{
	const struct ipt_entry *e = par->entryinfo;

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

	return nf_ct_l3proto_try_module_get(par->family);
}

static void synproxy_tg4_destroy(const struct xt_tgdtor_param *par)
{
	nf_ct_l3proto_module_put(par->family);
}

static struct xt_target synproxy_tg4_reg __read_mostly = {
	.name		= "SYNPROXY",
	.family		= NFPROTO_IPV4,
	.target		= synproxy_tg4,
	.targetsize	= sizeof(struct xt_synproxy_info),
	.checkentry	= synproxy_tg4_check,
	.destroy	= synproxy_tg4_destroy,
	.me		= THIS_MODULE,
};

static struct nf_hook_ops ipv4_synproxy_ops[] __read_mostly = {
	{
		.hook		= ipv4_synproxy_hook,
		.owner		= THIS_MODULE,
		.pf		= NFPROTO_IPV4,
		.hooknum	= NF_INET_LOCAL_IN,
		.priority	= NF_IP_PRI_CONNTRACK_CONFIRM - 1,
	},
	{
		.hook		= ipv4_synproxy_hook,
		.owner		= THIS_MODULE,
		.pf		= NFPROTO_IPV4,
		.hooknum	= NF_INET_POST_ROUTING,
		.priority	= NF_IP_PRI_CONNTRACK_CONFIRM - 1,
	},
};

static int __init synproxy_tg4_init(void)
{
	int err;

	err = nf_register_hooks(ipv4_synproxy_ops,
				ARRAY_SIZE(ipv4_synproxy_ops));
	if (err < 0)
		goto err1;

	err = xt_register_target(&synproxy_tg4_reg);
	if (err < 0)
		goto err2;

	return 0;

err2:
	nf_unregister_hooks(ipv4_synproxy_ops, ARRAY_SIZE(ipv4_synproxy_ops));
err1:
	return err;
}

static void __exit synproxy_tg4_exit(void)
{
	xt_unregister_target(&synproxy_tg4_reg);
	nf_unregister_hooks(ipv4_synproxy_ops, ARRAY_SIZE(ipv4_synproxy_ops));
}

module_init(synproxy_tg4_init);
module_exit(synproxy_tg4_exit);

MODULE_LICENSE("GPL");
MODULE_AUTHOR("Patrick McHardy <kaber@trash.net>");
Commit	Line	Data
48b1de4c 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/tcp.h>
	12
	13	#include <linux/netfilter_ipv4/ip_tables.h>
	14	#include <linux/netfilter/x_tables.h>
	15	#include <linux/netfilter/xt_SYNPROXY.h>
	16	#include <net/netfilter/nf_conntrack.h>
	17	#include <net/netfilter/nf_conntrack_seqadj.h>
	18	#include <net/netfilter/nf_conntrack_synproxy.h>
	19
	20	static struct iphdr *
	21	synproxy_build_ip(struct sk_buff *skb, u32 saddr, u32 daddr)
	22	{
	23	struct iphdr *iph;
	24
	25	skb_reset_network_header(skb);
	26	iph = (struct iphdr )skb_put(skb, sizeof(iph));
	27	iph->version = 4;
	28	iph->ihl = sizeof(*iph) / 4;
	29	iph->tos = 0;
	30	iph->id = 0;
	31	iph->frag_off = htons(IP_DF);
	32	iph->ttl = sysctl_ip_default_ttl;
	33	iph->protocol = IPPROTO_TCP;
	34	iph->check = 0;
	35	iph->saddr = saddr;
	36	iph->daddr = daddr;
	37
	38	return iph;
	39	}
	40
	41	static void
	42	synproxy_send_tcp(const struct sk_buff skb, struct sk_buff nskb,
	43	struct nf_conntrack *nfct, enum ip_conntrack_info ctinfo,
	44	struct iphdr niph, struct tcphdr nth,
	45	unsigned int tcp_hdr_size)
	46	{
	47	nth->check = ~tcp_v4_check(tcp_hdr_size, niph->saddr, niph->daddr, 0);
	48	nskb->ip_summed = CHECKSUM_PARTIAL;
	49	nskb->csum_start = (unsigned char *)nth - nskb->head;
	50	nskb->csum_offset = offsetof(struct tcphdr, check);
	51
	52	skb_dst_set_noref(nskb, skb_dst(skb));
	53	nskb->protocol = htons(ETH_P_IP);
	54	if (ip_route_me_harder(nskb, RTN_UNSPEC))
	55	goto free_nskb;
	56
	57	if (nfct) {
	58	nskb->nfct = nfct;
	59	nskb->nfctinfo = ctinfo;
	60	nf_conntrack_get(nfct);
	61	}
	62
	63	ip_local_out(nskb);
	64	return;
65
66	free_nskb:
67	kfree_skb(nskb);
68	}
69
70	static void
71	synproxy_send_client_synack(const struct sk_buff skb, const struct tcphdr th,
72	const struct synproxy_options *opts)
73	{
74	struct sk_buff *nskb;
75	struct iphdr iph, niph;
76	struct tcphdr *nth;
77	unsigned int tcp_hdr_size;
78	u16 mss = opts->mss;
79
80	iph = ip_hdr(skb);
81
82	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
83	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
84	GFP_ATOMIC);
85	if (nskb == NULL)
86	return;
87	skb_reserve(nskb, MAX_TCP_HEADER);
88
89	niph = synproxy_build_ip(nskb, iph->daddr, iph->saddr);
90
91	skb_reset_transport_header(nskb);
92	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
93	nth->source = th->dest;
94	nth->dest = th->source;
95	nth->seq = htonl(__cookie_v4_init_sequence(iph, th, &mss));
96	nth->ack_seq = htonl(ntohl(th->seq) + 1);
97	tcp_flag_word(nth) = TCP_FLAG_SYN \| TCP_FLAG_ACK;
98	if (opts->options & XT_SYNPROXY_OPT_ECN)
99	tcp_flag_word(nth) \|= TCP_FLAG_ECE;
100	nth->doff = tcp_hdr_size / 4;
101	nth->window = 0;
102	nth->check = 0;
103	nth->urg_ptr = 0;
104
105	synproxy_build_options(nth, opts);
106
107	synproxy_send_tcp(skb, nskb, skb->nfct, IP_CT_ESTABLISHED_REPLY,
108	niph, nth, tcp_hdr_size);
109	}
110
111	static void
112	synproxy_send_server_syn(const struct synproxy_net *snet,
113	const struct sk_buff skb, const struct tcphdr th,
114	const struct synproxy_options *opts, u32 recv_seq)
115	{
116	struct sk_buff *nskb;
117	struct iphdr iph, niph;
118	struct tcphdr *nth;
119	unsigned int tcp_hdr_size;
120
121	iph = ip_hdr(skb);
122
123	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
124	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
125	GFP_ATOMIC);
126	if (nskb == NULL)
127	return;
128	skb_reserve(nskb, MAX_TCP_HEADER);
129
130	niph = synproxy_build_ip(nskb, iph->saddr, iph->daddr);
131
132	skb_reset_transport_header(nskb);
133	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
134	nth->source = th->source;
135	nth->dest = th->dest;
136	nth->seq = htonl(recv_seq - 1);
137	/* ack_seq is used to relay our ISN to the synproxy hook to initialize
138	* sequence number translation once a connection tracking entry exists.
139	*/
140	nth->ack_seq = htonl(ntohl(th->ack_seq) - 1);
141	tcp_flag_word(nth) = TCP_FLAG_SYN;
142	if (opts->options & XT_SYNPROXY_OPT_ECN)
143	tcp_flag_word(nth) \|= TCP_FLAG_ECE \| TCP_FLAG_CWR;
144	nth->doff = tcp_hdr_size / 4;
145	nth->window = th->window;
146	nth->check = 0;
147	nth->urg_ptr = 0;
148
149	synproxy_build_options(nth, opts);
150
151	synproxy_send_tcp(skb, nskb, &snet->tmpl->ct_general, IP_CT_NEW,
152	niph, nth, tcp_hdr_size);
153	}
154
155	static void
156	synproxy_send_server_ack(const struct synproxy_net *snet,
157	const struct ip_ct_tcp *state,
158	const struct sk_buff skb, const struct tcphdr th,
159	const struct synproxy_options *opts)
160	{
161	struct sk_buff *nskb;
162	struct iphdr iph, niph;
163	struct tcphdr *nth;
164	unsigned int tcp_hdr_size;
165
166	iph = ip_hdr(skb);
167
168	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
169	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
170	GFP_ATOMIC);
171	if (nskb == NULL)
172	return;
173	skb_reserve(nskb, MAX_TCP_HEADER);
174
175	niph = synproxy_build_ip(nskb, iph->daddr, iph->saddr);
176
177	skb_reset_transport_header(nskb);
178	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
179	nth->source = th->dest;
180	nth->dest = th->source;
181	nth->seq = htonl(ntohl(th->ack_seq));
182	nth->ack_seq = htonl(ntohl(th->seq) + 1);
183	tcp_flag_word(nth) = TCP_FLAG_ACK;
184	nth->doff = tcp_hdr_size / 4;
185	nth->window = htons(state->seen[IP_CT_DIR_ORIGINAL].td_maxwin);
186	nth->check = 0;
187	nth->urg_ptr = 0;
188
189	synproxy_build_options(nth, opts);
190
191	synproxy_send_tcp(skb, nskb, NULL, 0, niph, nth, tcp_hdr_size);
192	}
193
194	static void
195	synproxy_send_client_ack(const struct synproxy_net *snet,
196	const struct sk_buff skb, const struct tcphdr th,
197	const struct synproxy_options *opts)
198	{
199	struct sk_buff *nskb;
200	struct iphdr iph, niph;
201	struct tcphdr *nth;
202	unsigned int tcp_hdr_size;
203
204	iph = ip_hdr(skb);
205
206	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
207	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
208	GFP_ATOMIC);
209	if (nskb == NULL)
210	return;
211	skb_reserve(nskb, MAX_TCP_HEADER);
212
213	niph = synproxy_build_ip(nskb, iph->saddr, iph->daddr);
214
215	skb_reset_transport_header(nskb);
216	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
217	nth->source = th->source;
218	nth->dest = th->dest;
219	nth->seq = htonl(ntohl(th->seq) + 1);
220	nth->ack_seq = th->ack_seq;
221	tcp_flag_word(nth) = TCP_FLAG_ACK;
222	nth->doff = tcp_hdr_size / 4;
223	nth->window = ntohs(htons(th->window) >> opts->wscale);
224	nth->check = 0;
225	nth->urg_ptr = 0;
226
227	synproxy_build_options(nth, opts);
228
229	synproxy_send_tcp(skb, nskb, NULL, 0, niph, nth, tcp_hdr_size);
230	}
231
232	static bool
233	synproxy_recv_client_ack(const struct synproxy_net *snet,
234	const struct sk_buff skb, const struct tcphdr th,
235	struct synproxy_options *opts, u32 recv_seq)
236	{
237	int mss;
238
239	mss = __cookie_v4_check(ip_hdr(skb), th, ntohl(th->ack_seq) - 1);
240	if (mss == 0) {
241	this_cpu_inc(snet->stats->cookie_invalid);
242	return false;
243	}
244
245	this_cpu_inc(snet->stats->cookie_valid);
246	opts->mss = mss;
247
248	if (opts->options & XT_SYNPROXY_OPT_TIMESTAMP)
249	synproxy_check_timestamp_cookie(opts);
250
251	synproxy_send_server_syn(snet, skb, th, opts, recv_seq);
252	return true;
253	}
254
255	static unsigned int
256	synproxy_tg4(struct sk_buff skb, const struct xt_action_param par)
257	{
258	const struct xt_synproxy_info *info = par->targinfo;
259	struct synproxy_net *snet = synproxy_pernet(dev_net(par->in));
260	struct synproxy_options opts = {};
261	struct tcphdr *th, _th;
262
263	if (nf_ip_checksum(skb, par->hooknum, par->thoff, IPPROTO_TCP))
264	return NF_DROP;
265
266	th = skb_header_pointer(skb, par->thoff, sizeof(_th), &_th);
267	if (th == NULL)
268	return NF_DROP;
269
270	synproxy_parse_options(skb, par->thoff, th, &opts);
271
775ada6d	272	if (th->syn && !(th->ack \|\| th->fin \|\| th->rst)) {
48b1de4c PM	273	/* Initial SYN from client */
	274	this_cpu_inc(snet->stats->syn_received);
	275
	276	if (th->ece && th->cwr)
	277	opts.options \|= XT_SYNPROXY_OPT_ECN;
	278
	279	opts.options &= info->options;
	280	if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
	281	synproxy_init_timestamp_cookie(info, &opts);
	282	else
	283	opts.options &= ~(XT_SYNPROXY_OPT_WSCALE \|
	284	XT_SYNPROXY_OPT_SACK_PERM \|
	285	XT_SYNPROXY_OPT_ECN);
	286
	287	synproxy_send_client_synack(skb, th, &opts);
7cc9eb6e JDB	288	return NF_DROP;
	289
	290	} else if (th->ack && !(th->fin \|\| th->rst \|\| th->syn)) {
48b1de4c PM	291	/* ACK from client */
48b1de4c PM	292	synproxy_recv_client_ack(snet, skb, th, &opts, ntohl(th->seq));
7cc9eb6e JDB	293	return NF_DROP;
7cc9eb6e JDB	294	}
48b1de4c	295
7cc9eb6e	296	return XT_CONTINUE;
48b1de4c PM	297	}
	298
	299	static unsigned int ipv4_synproxy_hook(unsigned int hooknum,
	300	struct sk_buff *skb,
	301	const struct net_device *in,
	302	const struct net_device *out,
	303	int (okfn)(struct sk_buff ))
	304	{
	305	struct synproxy_net *snet = synproxy_pernet(dev_net(in ? : out));
	306	enum ip_conntrack_info ctinfo;
	307	struct nf_conn *ct;
	308	struct nf_conn_synproxy *synproxy;
	309	struct synproxy_options opts = {};
	310	const struct ip_ct_tcp *state;
	311	struct tcphdr *th, _th;
	312	unsigned int thoff;
	313
	314	ct = nf_ct_get(skb, &ctinfo);
	315	if (ct == NULL)
	316	return NF_ACCEPT;
	317
	318	synproxy = nfct_synproxy(ct);
	319	if (synproxy == NULL)
	320	return NF_ACCEPT;
	321
	322	if (nf_is_loopback_packet(skb))
	323	return NF_ACCEPT;
	324
	325	thoff = ip_hdrlen(skb);
	326	th = skb_header_pointer(skb, thoff, sizeof(_th), &_th);
	327	if (th == NULL)
	328	return NF_DROP;
	329
	330	state = &ct->proto.tcp;
	331	switch (state->state) {
	332	case TCP_CONNTRACK_CLOSE:
	333	if (th->rst && !test_bit(IPS_SEEN_REPLY_BIT, &ct->status)) {
	334	nf_ct_seqadj_init(ct, ctinfo, synproxy->isn -
	335	ntohl(th->seq) + 1);
	336	break;
	337	}
	338
	339	if (!th->syn \|\| th->ack \|\|
	340	CTINFO2DIR(ctinfo) != IP_CT_DIR_ORIGINAL)
	341	break;
	342
	343	/* Reopened connection - reset the sequence number and timestamp
	344	* adjustments, they will get initialized once the connection is
	345	* reestablished.
	346	*/
	347	nf_ct_seqadj_init(ct, ctinfo, 0);
	348	synproxy->tsoff = 0;
	349	this_cpu_inc(snet->stats->conn_reopened);
	350
	351	/* fall through */
	352	case TCP_CONNTRACK_SYN_SENT:
	353	synproxy_parse_options(skb, thoff, th, &opts);
	354
	355	if (!th->syn && th->ack &&
	356	CTINFO2DIR(ctinfo) == IP_CT_DIR_ORIGINAL) {
	357	/* Keep-Alives are sent with SEG.SEQ = SND.NXT-1,
	358	* therefore we need to add 1 to make the SYN sequence
	359	* number match the one of first SYN.
	360	*/
361	if (synproxy_recv_client_ack(snet, skb, th, &opts,
362	ntohl(th->seq) + 1))
363	this_cpu_inc(snet->stats->cookie_retrans);
364
365	return NF_DROP;
366	}
367
368	synproxy->isn = ntohl(th->ack_seq);
369	if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
370	synproxy->its = opts.tsecr;
371	break;
372	case TCP_CONNTRACK_SYN_RECV:
373	if (!th->syn \|\| !th->ack)
374	break;
375
376	synproxy_parse_options(skb, thoff, th, &opts);
377	if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
378	synproxy->tsoff = opts.tsval - synproxy->its;
379
380	opts.options &= ~(XT_SYNPROXY_OPT_MSS \|
381	XT_SYNPROXY_OPT_WSCALE \|
382	XT_SYNPROXY_OPT_SACK_PERM);
383
384	swap(opts.tsval, opts.tsecr);
385	synproxy_send_server_ack(snet, state, skb, th, &opts);
386
387	nf_ct_seqadj_init(ct, ctinfo, synproxy->isn - ntohl(th->seq));
388
389	swap(opts.tsval, opts.tsecr);
390	synproxy_send_client_ack(snet, skb, th, &opts);
391
392	consume_skb(skb);
393	return NF_STOLEN;
394	default:
395	break;
396	}
397
398	synproxy_tstamp_adjust(skb, thoff, th, ct, ctinfo, synproxy);
399	return NF_ACCEPT;
400	}
401
402	static int synproxy_tg4_check(const struct xt_tgchk_param *par)
403	{
404	const struct ipt_entry *e = par->entryinfo;
405
406	if (e->ip.proto != IPPROTO_TCP \|\|
407	e->ip.invflags & XT_INV_PROTO)
408	return -EINVAL;
409
410	return nf_ct_l3proto_try_module_get(par->family);
411	}
412
413	static void synproxy_tg4_destroy(const struct xt_tgdtor_param *par)
414	{
415	nf_ct_l3proto_module_put(par->family);
416	}
417
418	static struct xt_target synproxy_tg4_reg __read_mostly = {
419	.name = "SYNPROXY",
420	.family = NFPROTO_IPV4,
421	.target = synproxy_tg4,
422	.targetsize = sizeof(struct xt_synproxy_info),
423	.checkentry = synproxy_tg4_check,
424	.destroy = synproxy_tg4_destroy,
425	.me = THIS_MODULE,
426	};
427
428	static struct nf_hook_ops ipv4_synproxy_ops[] __read_mostly = {
429	{
430	.hook = ipv4_synproxy_hook,
431	.owner = THIS_MODULE,
432	.pf = NFPROTO_IPV4,
433	.hooknum = NF_INET_LOCAL_IN,
434	.priority = NF_IP_PRI_CONNTRACK_CONFIRM - 1,
435	},
436	{
437	.hook = ipv4_synproxy_hook,
438	.owner = THIS_MODULE,
439	.pf = NFPROTO_IPV4,
440	.hooknum = NF_INET_POST_ROUTING,
441	.priority = NF_IP_PRI_CONNTRACK_CONFIRM - 1,
442	},
443	};
444
445	static int __init synproxy_tg4_init(void)
446	{
447	int err;
448
449	err = nf_register_hooks(ipv4_synproxy_ops,
450	ARRAY_SIZE(ipv4_synproxy_ops));
451	if (err < 0)
452	goto err1;
453
454	err = xt_register_target(&synproxy_tg4_reg);
455	if (err < 0)
456	goto err2;
457
458	return 0;
459
460	err2:
461	nf_unregister_hooks(ipv4_synproxy_ops, ARRAY_SIZE(ipv4_synproxy_ops));
462	err1:
463	return err;
464	}
465
466	static void __exit synproxy_tg4_exit(void)
467	{
468	xt_unregister_target(&synproxy_tg4_reg);
469	nf_unregister_hooks(ipv4_synproxy_ops, ARRAY_SIZE(ipv4_synproxy_ops));
470	}
471
472	module_init(synproxy_tg4_init);
473	module_exit(synproxy_tg4_exit);
474
475	MODULE_LICENSE("GPL");
476	MODULE_AUTHOR("Patrick McHardy <kaber@trash.net>");