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