[mirror_ubuntu-artful-kernel.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, __be32 saddr, __be32 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 synproxy_net *snet,
		  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)
{
	struct net *net = nf_ct_net(snet->tmpl);

	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(net, nskb, RTN_UNSPEC))
		goto free_nskb;

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

	ip_local_out(nskb->sk, nskb);
	return;

free_nskb:
	kfree_skb(nskb);
}

static void
synproxy_send_client_synack(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;
	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(snet, 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(snet, 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(snet, 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	= htons(ntohs(th->window) >> opts->wscale);
	nth->check	= 0;
	nth->urg_ptr	= 0;

	synproxy_build_options(nth, opts);

	synproxy_send_tcp(snet, skb, nskb, skb->nfct, IP_CT_ESTABLISHED_REPLY,
	                  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;
	opts->options |= XT_SYNPROXY_OPT_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(par->net);
	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;

	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(snet, 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(void *priv,
				       struct sk_buff *skb,
				       const struct nf_hook_state *nhs)
{
	struct synproxy_net *snet = synproxy_pernet(nhs->net);
	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:
		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_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,
	.hooks		= (1 << NF_INET_LOCAL_IN) | (1 << NF_INET_FORWARD),
	.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 *
ba6d0564	21	synproxy_build_ip(struct sk_buff *skb, __be32 saddr, __be32 daddr)
48b1de4c PM	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
6a1d689d EB	42	synproxy_send_tcp(const struct synproxy_net *snet,
6a1d689d EB	43	const struct sk_buff skb, struct sk_buff nskb,
48b1de4c PM	44	struct nf_conntrack *nfct, enum ip_conntrack_info ctinfo,
	45	struct iphdr niph, struct tcphdr nth,
	46	unsigned int tcp_hdr_size)
	47	{
e45f5066 EB	48	struct net *net = nf_ct_net(snet->tmpl);
e45f5066 EB	49
48b1de4c PM	50	nth->check = ~tcp_v4_check(tcp_hdr_size, niph->saddr, niph->daddr, 0);
	51	nskb->ip_summed = CHECKSUM_PARTIAL;
	52	nskb->csum_start = (unsigned char *)nth - nskb->head;
	53	nskb->csum_offset = offsetof(struct tcphdr, check);
	54
	55	skb_dst_set_noref(nskb, skb_dst(skb));
	56	nskb->protocol = htons(ETH_P_IP);
e45f5066	57	if (ip_route_me_harder(net, nskb, RTN_UNSPEC))
48b1de4c PM	58	goto free_nskb;
	59
	60	if (nfct) {
	61	nskb->nfct = nfct;
	62	nskb->nfctinfo = ctinfo;
	63	nf_conntrack_get(nfct);
	64	}
	65
e2cb77db	66	ip_local_out(nskb->sk, nskb);
48b1de4c PM	67	return;
	68
	69	free_nskb:
	70	kfree_skb(nskb);
	71	}
	72
	73	static void
6a1d689d EB	74	synproxy_send_client_synack(const struct synproxy_net *snet,
6a1d689d EB	75	const struct sk_buff skb, const struct tcphdr th,
48b1de4c PM	76	const struct synproxy_options *opts)
	77	{
	78	struct sk_buff *nskb;
	79	struct iphdr iph, niph;
	80	struct tcphdr *nth;
	81	unsigned int tcp_hdr_size;
	82	u16 mss = opts->mss;
	83
	84	iph = ip_hdr(skb);
	85
	86	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
	87	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
	88	GFP_ATOMIC);
	89	if (nskb == NULL)
	90	return;
	91	skb_reserve(nskb, MAX_TCP_HEADER);
	92
	93	niph = synproxy_build_ip(nskb, iph->daddr, iph->saddr);
	94
	95	skb_reset_transport_header(nskb);
	96	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
	97	nth->source = th->dest;
	98	nth->dest = th->source;
	99	nth->seq = htonl(__cookie_v4_init_sequence(iph, th, &mss));
	100	nth->ack_seq = htonl(ntohl(th->seq) + 1);
	101	tcp_flag_word(nth) = TCP_FLAG_SYN \| TCP_FLAG_ACK;
	102	if (opts->options & XT_SYNPROXY_OPT_ECN)
	103	tcp_flag_word(nth) \|= TCP_FLAG_ECE;
	104	nth->doff = tcp_hdr_size / 4;
	105	nth->window = 0;
	106	nth->check = 0;
	107	nth->urg_ptr = 0;
	108
	109	synproxy_build_options(nth, opts);
	110
6a1d689d	111	synproxy_send_tcp(snet, skb, nskb, skb->nfct, IP_CT_ESTABLISHED_REPLY,
48b1de4c PM	112	niph, nth, tcp_hdr_size);
	113	}
	114
	115	static void
	116	synproxy_send_server_syn(const struct synproxy_net *snet,
	117	const struct sk_buff skb, const struct tcphdr th,
	118	const struct synproxy_options *opts, u32 recv_seq)
	119	{
	120	struct sk_buff *nskb;
	121	struct iphdr iph, niph;
	122	struct tcphdr *nth;
	123	unsigned int tcp_hdr_size;
	124
	125	iph = ip_hdr(skb);
	126
	127	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
	128	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
	129	GFP_ATOMIC);
	130	if (nskb == NULL)
	131	return;
	132	skb_reserve(nskb, MAX_TCP_HEADER);
	133
	134	niph = synproxy_build_ip(nskb, iph->saddr, iph->daddr);
	135
	136	skb_reset_transport_header(nskb);
	137	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
	138	nth->source = th->source;
	139	nth->dest = th->dest;
	140	nth->seq = htonl(recv_seq - 1);
	141	/* ack_seq is used to relay our ISN to the synproxy hook to initialize
	142	* sequence number translation once a connection tracking entry exists.
	143	*/
	144	nth->ack_seq = htonl(ntohl(th->ack_seq) - 1);
	145	tcp_flag_word(nth) = TCP_FLAG_SYN;
	146	if (opts->options & XT_SYNPROXY_OPT_ECN)
	147	tcp_flag_word(nth) \|= TCP_FLAG_ECE \| TCP_FLAG_CWR;
	148	nth->doff = tcp_hdr_size / 4;
	149	nth->window = th->window;
	150	nth->check = 0;
	151	nth->urg_ptr = 0;
	152
	153	synproxy_build_options(nth, opts);
	154
6a1d689d	155	synproxy_send_tcp(snet, skb, nskb, &snet->tmpl->ct_general, IP_CT_NEW,
48b1de4c PM	156	niph, nth, tcp_hdr_size);
	157	}
	158
	159	static void
	160	synproxy_send_server_ack(const struct synproxy_net *snet,
	161	const struct ip_ct_tcp *state,
	162	const struct sk_buff skb, const struct tcphdr th,
	163	const struct synproxy_options *opts)
	164	{
	165	struct sk_buff *nskb;
	166	struct iphdr iph, niph;
	167	struct tcphdr *nth;
	168	unsigned int tcp_hdr_size;
	169
	170	iph = ip_hdr(skb);
	171
	172	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
	173	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
	174	GFP_ATOMIC);
	175	if (nskb == NULL)
	176	return;
	177	skb_reserve(nskb, MAX_TCP_HEADER);
	178
	179	niph = synproxy_build_ip(nskb, iph->daddr, iph->saddr);
	180
	181	skb_reset_transport_header(nskb);
	182	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
	183	nth->source = th->dest;
	184	nth->dest = th->source;
	185	nth->seq = htonl(ntohl(th->ack_seq));
	186	nth->ack_seq = htonl(ntohl(th->seq) + 1);
	187	tcp_flag_word(nth) = TCP_FLAG_ACK;
	188	nth->doff = tcp_hdr_size / 4;
	189	nth->window = htons(state->seen[IP_CT_DIR_ORIGINAL].td_maxwin);
	190	nth->check = 0;
	191	nth->urg_ptr = 0;
	192
	193	synproxy_build_options(nth, opts);
	194
6a1d689d	195	synproxy_send_tcp(snet, skb, nskb, NULL, 0, niph, nth, tcp_hdr_size);
48b1de4c PM	196	}
	197
	198	static void
	199	synproxy_send_client_ack(const struct synproxy_net *snet,
	200	const struct sk_buff skb, const struct tcphdr th,
	201	const struct synproxy_options *opts)
	202	{
	203	struct sk_buff *nskb;
	204	struct iphdr iph, niph;
	205	struct tcphdr *nth;
	206	unsigned int tcp_hdr_size;
	207
	208	iph = ip_hdr(skb);
	209
	210	tcp_hdr_size = sizeof(*nth) + synproxy_options_size(opts);
	211	nskb = alloc_skb(sizeof(*niph) + tcp_hdr_size + MAX_TCP_HEADER,
	212	GFP_ATOMIC);
	213	if (nskb == NULL)
	214	return;
	215	skb_reserve(nskb, MAX_TCP_HEADER);
	216
	217	niph = synproxy_build_ip(nskb, iph->saddr, iph->daddr);
	218
	219	skb_reset_transport_header(nskb);
	220	nth = (struct tcphdr *)skb_put(nskb, tcp_hdr_size);
	221	nth->source = th->source;
	222	nth->dest = th->dest;
	223	nth->seq = htonl(ntohl(th->seq) + 1);
	224	nth->ack_seq = th->ack_seq;
	225	tcp_flag_word(nth) = TCP_FLAG_ACK;
	226	nth->doff = tcp_hdr_size / 4;
ba6d0564	227	nth->window = htons(ntohs(th->window) >> opts->wscale);
48b1de4c PM	228	nth->check = 0;
	229	nth->urg_ptr = 0;
	230
	231	synproxy_build_options(nth, opts);
	232
6a1d689d	233	synproxy_send_tcp(snet, skb, nskb, skb->nfct, IP_CT_ESTABLISHED_REPLY,
3c16241c	234	niph, nth, tcp_hdr_size);
48b1de4c PM	235	}
	236
	237	static bool
	238	synproxy_recv_client_ack(const struct synproxy_net *snet,
	239	const struct sk_buff skb, const struct tcphdr th,
	240	struct synproxy_options *opts, u32 recv_seq)
	241	{
	242	int mss;
	243
	244	mss = __cookie_v4_check(ip_hdr(skb), th, ntohl(th->ack_seq) - 1);
	245	if (mss == 0) {
	246	this_cpu_inc(snet->stats->cookie_invalid);
	247	return false;
	248	}
	249
	250	this_cpu_inc(snet->stats->cookie_valid);
	251	opts->mss = mss;
a6441b7a	252	opts->options \|= XT_SYNPROXY_OPT_MSS;
48b1de4c PM	253
	254	if (opts->options & XT_SYNPROXY_OPT_TIMESTAMP)
	255	synproxy_check_timestamp_cookie(opts);
	256
	257	synproxy_send_server_syn(snet, skb, th, opts, recv_seq);
	258	return true;
	259	}
	260
	261	static unsigned int
	262	synproxy_tg4(struct sk_buff skb, const struct xt_action_param par)
	263	{
	264	const struct xt_synproxy_info *info = par->targinfo;
686c9b50	265	struct synproxy_net *snet = synproxy_pernet(par->net);
48b1de4c PM	266	struct synproxy_options opts = {};
	267	struct tcphdr *th, _th;
	268
	269	if (nf_ip_checksum(skb, par->hooknum, par->thoff, IPPROTO_TCP))
	270	return NF_DROP;
	271
	272	th = skb_header_pointer(skb, par->thoff, sizeof(_th), &_th);
	273	if (th == NULL)
	274	return NF_DROP;
	275
f4a87e7b PM	276	if (!synproxy_parse_options(skb, par->thoff, th, &opts))
f4a87e7b PM	277	return NF_DROP;
48b1de4c	278
775ada6d	279	if (th->syn && !(th->ack \|\| th->fin \|\| th->rst)) {
48b1de4c PM	280	/* Initial SYN from client */
	281	this_cpu_inc(snet->stats->syn_received);
	282
	283	if (th->ece && th->cwr)
	284	opts.options \|= XT_SYNPROXY_OPT_ECN;
	285
	286	opts.options &= info->options;
	287	if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
	288	synproxy_init_timestamp_cookie(info, &opts);
	289	else
	290	opts.options &= ~(XT_SYNPROXY_OPT_WSCALE \|
	291	XT_SYNPROXY_OPT_SACK_PERM \|
	292	XT_SYNPROXY_OPT_ECN);
	293
6a1d689d	294	synproxy_send_client_synack(snet, skb, th, &opts);
7cc9eb6e JDB	295	return NF_DROP;
	296
	297	} else if (th->ack && !(th->fin \|\| th->rst \|\| th->syn)) {
48b1de4c PM	298	/* ACK from client */
48b1de4c PM	299	synproxy_recv_client_ack(snet, skb, th, &opts, ntohl(th->seq));
7cc9eb6e JDB	300	return NF_DROP;
7cc9eb6e JDB	301	}
48b1de4c	302
7cc9eb6e	303	return XT_CONTINUE;
48b1de4c PM	304	}
48b1de4c PM	305
06198b34	306	static unsigned int ipv4_synproxy_hook(void *priv,
48b1de4c	307	struct sk_buff *skb,
238e54c9	308	const struct nf_hook_state *nhs)
48b1de4c	309	{
9dff2c96	310	struct synproxy_net *snet = synproxy_pernet(nhs->net);
48b1de4c PM	311	enum ip_conntrack_info ctinfo;
	312	struct nf_conn *ct;
	313	struct nf_conn_synproxy *synproxy;
	314	struct synproxy_options opts = {};
	315	const struct ip_ct_tcp *state;
	316	struct tcphdr *th, _th;
	317	unsigned int thoff;
	318
	319	ct = nf_ct_get(skb, &ctinfo);
	320	if (ct == NULL)
	321	return NF_ACCEPT;
	322
	323	synproxy = nfct_synproxy(ct);
	324	if (synproxy == NULL)
	325	return NF_ACCEPT;
	326
	327	if (nf_is_loopback_packet(skb))
	328	return NF_ACCEPT;
	329
	330	thoff = ip_hdrlen(skb);
	331	th = skb_header_pointer(skb, thoff, sizeof(_th), &_th);
	332	if (th == NULL)
	333	return NF_DROP;
	334
	335	state = &ct->proto.tcp;
	336	switch (state->state) {
	337	case TCP_CONNTRACK_CLOSE:
	338	if (th->rst && !test_bit(IPS_SEEN_REPLY_BIT, &ct->status)) {
	339	nf_ct_seqadj_init(ct, ctinfo, synproxy->isn -
	340	ntohl(th->seq) + 1);
	341	break;
	342	}
	343
	344	if (!th->syn \|\| th->ack \|\|
	345	CTINFO2DIR(ctinfo) != IP_CT_DIR_ORIGINAL)
	346	break;
	347
	348	/* Reopened connection - reset the sequence number and timestamp
	349	* adjustments, they will get initialized once the connection is
	350	* reestablished.
	351	*/
	352	nf_ct_seqadj_init(ct, ctinfo, 0);
	353	synproxy->tsoff = 0;
	354	this_cpu_inc(snet->stats->conn_reopened);
	355
	356	/* fall through */
	357	case TCP_CONNTRACK_SYN_SENT:
f4a87e7b PM	358	if (!synproxy_parse_options(skb, thoff, th, &opts))
f4a87e7b PM	359	return NF_DROP;
48b1de4c PM	360
	361	if (!th->syn && th->ack &&
	362	CTINFO2DIR(ctinfo) == IP_CT_DIR_ORIGINAL) {
	363	/* Keep-Alives are sent with SEG.SEQ = SND.NXT-1,
	364	* therefore we need to add 1 to make the SYN sequence
	365	* number match the one of first SYN.
	366	*/
	367	if (synproxy_recv_client_ack(snet, skb, th, &opts,
	368	ntohl(th->seq) + 1))
	369	this_cpu_inc(snet->stats->cookie_retrans);
	370
	371	return NF_DROP;
	372	}
	373
	374	synproxy->isn = ntohl(th->ack_seq);
	375	if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
	376	synproxy->its = opts.tsecr;
	377	break;
	378	case TCP_CONNTRACK_SYN_RECV:
	379	if (!th->syn \|\| !th->ack)
	380	break;
	381
f4a87e7b PM	382	if (!synproxy_parse_options(skb, thoff, th, &opts))
	383	return NF_DROP;
	384
48b1de4c PM	385	if (opts.options & XT_SYNPROXY_OPT_TIMESTAMP)
	386	synproxy->tsoff = opts.tsval - synproxy->its;
	387
	388	opts.options &= ~(XT_SYNPROXY_OPT_MSS \|
	389	XT_SYNPROXY_OPT_WSCALE \|
	390	XT_SYNPROXY_OPT_SACK_PERM);
	391
	392	swap(opts.tsval, opts.tsecr);
	393	synproxy_send_server_ack(snet, state, skb, th, &opts);
	394
	395	nf_ct_seqadj_init(ct, ctinfo, synproxy->isn - ntohl(th->seq));
	396
	397	swap(opts.tsval, opts.tsecr);
	398	synproxy_send_client_ack(snet, skb, th, &opts);
	399
	400	consume_skb(skb);
	401	return NF_STOLEN;
	402	default:
	403	break;
	404	}
	405
	406	synproxy_tstamp_adjust(skb, thoff, th, ct, ctinfo, synproxy);
	407	return NF_ACCEPT;
	408	}
	409
	410	static int synproxy_tg4_check(const struct xt_tgchk_param *par)
	411	{
	412	const struct ipt_entry *e = par->entryinfo;
	413
	414	if (e->ip.proto != IPPROTO_TCP \|\|
	415	e->ip.invflags & XT_INV_PROTO)
	416	return -EINVAL;
	417
	418	return nf_ct_l3proto_try_module_get(par->family);
	419	}
	420
	421	static void synproxy_tg4_destroy(const struct xt_tgdtor_param *par)
	422	{
	423	nf_ct_l3proto_module_put(par->family);
	424	}
	425
	426	static struct xt_target synproxy_tg4_reg __read_mostly = {
	427	.name = "SYNPROXY",
	428	.family = NFPROTO_IPV4,
f01b3926	429	.hooks = (1 << NF_INET_LOCAL_IN) \| (1 << NF_INET_FORWARD),
48b1de4c PM	430	.target = synproxy_tg4,
	431	.targetsize = sizeof(struct xt_synproxy_info),
	432	.checkentry = synproxy_tg4_check,
	433	.destroy = synproxy_tg4_destroy,
	434	.me = THIS_MODULE,
	435	};
	436
	437	static struct nf_hook_ops ipv4_synproxy_ops[] __read_mostly = {
	438	{
	439	.hook = ipv4_synproxy_hook,
	440	.owner = THIS_MODULE,
	441	.pf = NFPROTO_IPV4,
	442	.hooknum = NF_INET_LOCAL_IN,
	443	.priority = NF_IP_PRI_CONNTRACK_CONFIRM - 1,
	444	},
	445	{
	446	.hook = ipv4_synproxy_hook,
	447	.owner = THIS_MODULE,
	448	.pf = NFPROTO_IPV4,
	449	.hooknum = NF_INET_POST_ROUTING,
	450	.priority = NF_IP_PRI_CONNTRACK_CONFIRM - 1,
	451	},
	452	};
	453
	454	static int __init synproxy_tg4_init(void)
	455	{
	456	int err;
	457
	458	err = nf_register_hooks(ipv4_synproxy_ops,
	459	ARRAY_SIZE(ipv4_synproxy_ops));
	460	if (err < 0)
	461	goto err1;
	462
	463	err = xt_register_target(&synproxy_tg4_reg);
	464	if (err < 0)
	465	goto err2;
	466
	467	return 0;
	468
	469	err2:
	470	nf_unregister_hooks(ipv4_synproxy_ops, ARRAY_SIZE(ipv4_synproxy_ops));
	471	err1:
	472	return err;
	473	}
	474
	475	static void __exit synproxy_tg4_exit(void)
	476	{
	477	xt_unregister_target(&synproxy_tg4_reg);
	478	nf_unregister_hooks(ipv4_synproxy_ops, ARRAY_SIZE(ipv4_synproxy_ops));
	479	}
	480
	481	module_init(synproxy_tg4_init);
	482	module_exit(synproxy_tg4_exit);
	483
	484	MODULE_LICENSE("GPL");
	485	MODULE_AUTHOR("Patrick McHardy <kaber@trash.net>");