[deliverable/linux.git] / net / ipv6 / syncookies.c

/*
 *  IPv6 Syncookies implementation for the Linux kernel
 *
 *  Authors:
 *  Glenn Griffin	<ggriffin.kernel@gmail.com>
 *
 *  Based on IPv4 implementation by Andi Kleen
 *  linux/net/ipv4/syncookies.c
 *
 *	This program is free software; you can redistribute it and/or
 *      modify it under the terms of the GNU General Public License
 *      as published by the Free Software Foundation; either version
 *      2 of the License, or (at your option) any later version.
 *
 */

#include <linux/tcp.h>
#include <linux/random.h>
#include <linux/cryptohash.h>
#include <linux/kernel.h>
#include <net/ipv6.h>
#include <net/tcp.h>

#define COOKIEBITS 24	/* Upper bits store count */
#define COOKIEMASK (((__u32)1 << COOKIEBITS) - 1)

/* Table must be sorted. */
static __u16 const msstab[] = {
	64,
	512,
	536,
	1280 - 60,
	1480 - 60,
	1500 - 60,
	4460 - 60,
	9000 - 60,
};

/*
 * This (misnamed) value is the age of syncookie which is permitted.
 * Its ideal value should be dependent on TCP_TIMEOUT_INIT and
 * sysctl_tcp_retries1. It's a rather complicated formula (exponential
 * backoff) to compute at runtime so it's currently hardcoded here.
 */
#define COUNTER_TRIES 4

static inline struct sock *get_cookie_sock(struct sock *sk, struct sk_buff *skb,
					   struct request_sock *req,
					   struct dst_entry *dst)
{
	struct inet_connection_sock *icsk = inet_csk(sk);
	struct sock *child;

	child = icsk->icsk_af_ops->syn_recv_sock(sk, skb, req, dst);
	if (child)
		inet_csk_reqsk_queue_add(sk, req, child);
	else
		reqsk_free(req);

	return child;
}

static DEFINE_PER_CPU(__u32 [16 + 5 + SHA_WORKSPACE_WORDS],
		      ipv6_cookie_scratch);

static u32 cookie_hash(const struct in6_addr *saddr, const struct in6_addr *daddr,
		       __be16 sport, __be16 dport, u32 count, int c)
{
	__u32 *tmp = __get_cpu_var(ipv6_cookie_scratch);

	/*
	 * we have 320 bits of information to hash, copy in the remaining
	 * 192 bits required for sha_transform, from the syncookie_secret
	 * and overwrite the digest with the secret
	 */
	memcpy(tmp + 10, syncookie_secret[c], 44);
	memcpy(tmp, saddr, 16);
	memcpy(tmp + 4, daddr, 16);
	tmp[8] = ((__force u32)sport << 16) + (__force u32)dport;
	tmp[9] = count;
	sha_transform(tmp + 16, (__u8 *)tmp, tmp + 16 + 5);

	return tmp[17];
}

static __u32 secure_tcp_syn_cookie(const struct in6_addr *saddr,
				   const struct in6_addr *daddr,
				   __be16 sport, __be16 dport, __u32 sseq,
				   __u32 count, __u32 data)
{
	return (cookie_hash(saddr, daddr, sport, dport, 0, 0) +
		sseq + (count << COOKIEBITS) +
		((cookie_hash(saddr, daddr, sport, dport, count, 1) + data)
		& COOKIEMASK));
}

static __u32 check_tcp_syn_cookie(__u32 cookie, const struct in6_addr *saddr,
				  const struct in6_addr *daddr, __be16 sport,
				  __be16 dport, __u32 sseq, __u32 count,
				  __u32 maxdiff)
{
	__u32 diff;

	cookie -= cookie_hash(saddr, daddr, sport, dport, 0, 0) + sseq;

	diff = (count - (cookie >> COOKIEBITS)) & ((__u32) -1 >> COOKIEBITS);
	if (diff >= maxdiff)
		return (__u32)-1;

	return (cookie -
		cookie_hash(saddr, daddr, sport, dport, count - diff, 1))
		& COOKIEMASK;
}

u32 __cookie_v6_init_sequence(const struct ipv6hdr *iph,
			      const struct tcphdr *th, __u16 *mssp)
{
	int mssind;
	const __u16 mss = *mssp;

	for (mssind = ARRAY_SIZE(msstab) - 1; mssind ; mssind--)
		if (mss >= msstab[mssind])
			break;

	*mssp = msstab[mssind];

	return secure_tcp_syn_cookie(&iph->saddr, &iph->daddr, th->source,
				     th->dest, ntohl(th->seq),
				     jiffies / (HZ * 60), mssind);
}
EXPORT_SYMBOL_GPL(__cookie_v6_init_sequence);

__u32 cookie_v6_init_sequence(struct sock *sk, const struct sk_buff *skb, __u16 *mssp)
{
	const struct ipv6hdr *iph = ipv6_hdr(skb);
	const struct tcphdr *th = tcp_hdr(skb);

	tcp_synq_overflow(sk);
	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESSENT);

	return __cookie_v6_init_sequence(iph, th, mssp);
}

int __cookie_v6_check(const struct ipv6hdr *iph, const struct tcphdr *th,
		      __u32 cookie)
{
	__u32 seq = ntohl(th->seq) - 1;
	__u32 mssind = check_tcp_syn_cookie(cookie, &iph->saddr, &iph->daddr,
					    th->source, th->dest, seq,
					    jiffies / (HZ * 60), COUNTER_TRIES);

	return mssind < ARRAY_SIZE(msstab) ? msstab[mssind] : 0;
}
EXPORT_SYMBOL_GPL(__cookie_v6_check);

struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb)
{
	struct tcp_options_received tcp_opt;
	struct inet_request_sock *ireq;
	struct inet6_request_sock *ireq6;
	struct tcp_request_sock *treq;
	struct ipv6_pinfo *np = inet6_sk(sk);
	struct tcp_sock *tp = tcp_sk(sk);
	const struct tcphdr *th = tcp_hdr(skb);
	__u32 cookie = ntohl(th->ack_seq) - 1;
	struct sock *ret = sk;
	struct request_sock *req;
	int mss;
	struct dst_entry *dst;
	__u8 rcv_wscale;
	bool ecn_ok = false;

	if (!sysctl_tcp_syncookies || !th->ack || th->rst)
		goto out;

	if (tcp_synq_no_recent_overflow(sk) ||
		(mss = __cookie_v6_check(ipv6_hdr(skb), th, cookie)) == 0) {
		NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESFAILED);
		goto out;
	}

	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESRECV);

	/* check for timestamp cookie support */
	memset(&tcp_opt, 0, sizeof(tcp_opt));
	tcp_parse_options(skb, &tcp_opt, 0, NULL);

	if (!cookie_check_timestamp(&tcp_opt, sock_net(sk), &ecn_ok))
		goto out;

	ret = NULL;
	req = inet6_reqsk_alloc(&tcp6_request_sock_ops);
	if (!req)
		goto out;

	ireq = inet_rsk(req);
	ireq6 = inet6_rsk(req);
	treq = tcp_rsk(req);
	treq->listener = NULL;

	if (security_inet_conn_request(sk, skb, req))
		goto out_free;

	req->mss = mss;
	ireq->rmt_port = th->source;
	ireq->loc_port = th->dest;
	ireq6->rmt_addr = ipv6_hdr(skb)->saddr;
	ireq6->loc_addr = ipv6_hdr(skb)->daddr;
	if (ipv6_opt_accepted(sk, skb) ||
	    np->rxopt.bits.rxinfo || np->rxopt.bits.rxoinfo ||
	    np->rxopt.bits.rxhlim || np->rxopt.bits.rxohlim) {
		atomic_inc(&skb->users);
		ireq6->pktopts = skb;
	}

	ireq6->iif = sk->sk_bound_dev_if;
	/* So that link locals have meaning */
	if (!sk->sk_bound_dev_if &&
	    ipv6_addr_type(&ireq6->rmt_addr) & IPV6_ADDR_LINKLOCAL)
		ireq6->iif = inet6_iif(skb);

	req->expires = 0UL;
	req->num_retrans = 0;
	ireq->ecn_ok		= ecn_ok;
	ireq->snd_wscale	= tcp_opt.snd_wscale;
	ireq->sack_ok		= tcp_opt.sack_ok;
	ireq->wscale_ok		= tcp_opt.wscale_ok;
	ireq->tstamp_ok		= tcp_opt.saw_tstamp;
	req->ts_recent		= tcp_opt.saw_tstamp ? tcp_opt.rcv_tsval : 0;
	treq->snt_synack	= tcp_opt.saw_tstamp ? tcp_opt.rcv_tsecr : 0;
	treq->rcv_isn = ntohl(th->seq) - 1;
	treq->snt_isn = cookie;

	/*
	 * We need to lookup the dst_entry to get the correct window size.
	 * This is taken from tcp_v6_syn_recv_sock.  Somebody please enlighten
	 * me if there is a preferred way.
	 */
	{
		struct in6_addr *final_p, final;
		struct flowi6 fl6;
		memset(&fl6, 0, sizeof(fl6));
		fl6.flowi6_proto = IPPROTO_TCP;
		fl6.daddr = ireq6->rmt_addr;
		final_p = fl6_update_dst(&fl6, np->opt, &final);
		fl6.saddr = ireq6->loc_addr;
		fl6.flowi6_oif = sk->sk_bound_dev_if;
		fl6.flowi6_mark = sk->sk_mark;
		fl6.fl6_dport = inet_rsk(req)->rmt_port;
		fl6.fl6_sport = inet_sk(sk)->inet_sport;
		security_req_classify_flow(req, flowi6_to_flowi(&fl6));

		dst = ip6_dst_lookup_flow(sk, &fl6, final_p, false);
		if (IS_ERR(dst))
			goto out_free;
	}

	req->window_clamp = tp->window_clamp ? :dst_metric(dst, RTAX_WINDOW);
	tcp_select_initial_window(tcp_full_space(sk), req->mss,
				  &req->rcv_wnd, &req->window_clamp,
				  ireq->wscale_ok, &rcv_wscale,
				  dst_metric(dst, RTAX_INITRWND));

	ireq->rcv_wscale = rcv_wscale;

	ret = get_cookie_sock(sk, skb, req, dst);
out:
	return ret;
out_free:
	reqsk_free(req);
	return NULL;
}
Commit	Line	Data
c6aefafb GG	1	/*
	2	* IPv6 Syncookies implementation for the Linux kernel
	3	*
	4	* Authors:
	5	* Glenn Griffin <ggriffin.kernel@gmail.com>
	6	*
	7	* Based on IPv4 implementation by Andi Kleen
	8	* linux/net/ipv4/syncookies.c
	9	*
	10	* This program is free software; you can redistribute it and/or
	11	* modify it under the terms of the GNU General Public License
	12	* as published by the Free Software Foundation; either version
	13	* 2 of the License, or (at your option) any later version.
	14	*
	15	*/
	16
	17	#include <linux/tcp.h>
	18	#include <linux/random.h>
	19	#include <linux/cryptohash.h>
	20	#include <linux/kernel.h>
	21	#include <net/ipv6.h>
	22	#include <net/tcp.h>
	23
c6aefafb GG	24	#define COOKIEBITS 24 /* Upper bits store count */
	25	#define COOKIEMASK (((__u32)1 << COOKIEBITS) - 1)
	26
5918e2fb	27	/* Table must be sorted. */
c6aefafb	28	static __u16 const msstab[] = {
5918e2fb FW	29	64,
	30	512,
	31	536,
	32	1280 - 60,
	33	1480 - 60,
	34	1500 - 60,
	35	4460 - 60,
	36	9000 - 60,
c6aefafb	37	};
c6aefafb GG	38
	39	/*
	40	* This (misnamed) value is the age of syncookie which is permitted.
	41	* Its ideal value should be dependent on TCP_TIMEOUT_INIT and
	42	* sysctl_tcp_retries1. It's a rather complicated formula (exponential
	43	* backoff) to compute at runtime so it's currently hardcoded here.
	44	*/
	45	#define COUNTER_TRIES 4
	46
	47	static inline struct sock get_cookie_sock(struct sock sk, struct sk_buff *skb,
	48	struct request_sock *req,
	49	struct dst_entry *dst)
	50	{
	51	struct inet_connection_sock *icsk = inet_csk(sk);
	52	struct sock *child;
	53
	54	child = icsk->icsk_af_ops->syn_recv_sock(sk, skb, req, dst);
	55	if (child)
	56	inet_csk_reqsk_queue_add(sk, req, child);
	57	else
	58	reqsk_free(req);
	59
	60	return child;
	61	}
	62
245b2e70 TH	63	static DEFINE_PER_CPU(__u32 [16 + 5 + SHA_WORKSPACE_WORDS],
245b2e70 TH	64	ipv6_cookie_scratch);
c6aefafb	65
b71d1d42	66	static u32 cookie_hash(const struct in6_addr saddr, const struct in6_addr daddr,
c6aefafb GG	67	__be16 sport, __be16 dport, u32 count, int c)
c6aefafb GG	68	{
245b2e70	69	__u32 *tmp = __get_cpu_var(ipv6_cookie_scratch);
c6aefafb GG	70
	71	/*
	72	* we have 320 bits of information to hash, copy in the remaining
	73	* 192 bits required for sha_transform, from the syncookie_secret
	74	* and overwrite the digest with the secret
	75	*/
	76	memcpy(tmp + 10, syncookie_secret[c], 44);
	77	memcpy(tmp, saddr, 16);
	78	memcpy(tmp + 4, daddr, 16);
	79	tmp[8] = ((__force u32)sport << 16) + (__force u32)dport;
	80	tmp[9] = count;
	81	sha_transform(tmp + 16, (__u8 *)tmp, tmp + 16 + 5);
	82
	83	return tmp[17];
	84	}
	85
b71d1d42 ED	86	static __u32 secure_tcp_syn_cookie(const struct in6_addr *saddr,
b71d1d42 ED	87	const struct in6_addr *daddr,
c6aefafb GG	88	__be16 sport, __be16 dport, __u32 sseq,
	89	__u32 count, __u32 data)
	90	{
	91	return (cookie_hash(saddr, daddr, sport, dport, 0, 0) +
	92	sseq + (count << COOKIEBITS) +
	93	((cookie_hash(saddr, daddr, sport, dport, count, 1) + data)
	94	& COOKIEMASK));
	95	}
	96
b71d1d42 ED	97	static __u32 check_tcp_syn_cookie(__u32 cookie, const struct in6_addr *saddr,
b71d1d42 ED	98	const struct in6_addr *daddr, __be16 sport,
c6aefafb GG	99	__be16 dport, __u32 sseq, __u32 count,
	100	__u32 maxdiff)
	101	{
	102	__u32 diff;
	103
	104	cookie -= cookie_hash(saddr, daddr, sport, dport, 0, 0) + sseq;
	105
	106	diff = (count - (cookie >> COOKIEBITS)) & ((__u32) -1 >> COOKIEBITS);
	107	if (diff >= maxdiff)
	108	return (__u32)-1;
	109
	110	return (cookie -
	111	cookie_hash(saddr, daddr, sport, dport, count - diff, 1))
	112	& COOKIEMASK;
	113	}
	114
81eb6a14 PM	115	u32 __cookie_v6_init_sequence(const struct ipv6hdr *iph,
81eb6a14 PM	116	const struct tcphdr th, __u16 mssp)
c6aefafb	117	{
c6aefafb GG	118	int mssind;
	119	const __u16 mss = *mssp;
	120
5918e2fb FW	121	for (mssind = ARRAY_SIZE(msstab) - 1; mssind ; mssind--)
	122	if (mss >= msstab[mssind])
	123	break;
	124
	125	*mssp = msstab[mssind];
c6aefafb	126
c6aefafb GG	127	return secure_tcp_syn_cookie(&iph->saddr, &iph->daddr, th->source,
	128	th->dest, ntohl(th->seq),
	129	jiffies / (HZ * 60), mssind);
	130	}
81eb6a14	131	EXPORT_SYMBOL_GPL(__cookie_v6_init_sequence);
c6aefafb	132
81eb6a14	133	__u32 cookie_v6_init_sequence(struct sock sk, const struct sk_buff skb, __u16 *mssp)
c6aefafb	134	{
b71d1d42	135	const struct ipv6hdr *iph = ipv6_hdr(skb);
c6aefafb	136	const struct tcphdr *th = tcp_hdr(skb);
81eb6a14 PM	137
	138	tcp_synq_overflow(sk);
	139	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESSENT);
	140
	141	return __cookie_v6_init_sequence(iph, th, mssp);
	142	}
	143
	144	int __cookie_v6_check(const struct ipv6hdr iph, const struct tcphdr th,
	145	__u32 cookie)
	146	{
c6aefafb GG	147	__u32 seq = ntohl(th->seq) - 1;
	148	__u32 mssind = check_tcp_syn_cookie(cookie, &iph->saddr, &iph->daddr,
	149	th->source, th->dest, seq,
	150	jiffies / (HZ * 60), COUNTER_TRIES);
	151
5918e2fb	152	return mssind < ARRAY_SIZE(msstab) ? msstab[mssind] : 0;
c6aefafb	153	}
81eb6a14	154	EXPORT_SYMBOL_GPL(__cookie_v6_check);
c6aefafb GG	155
	156	struct sock cookie_v6_check(struct sock sk, struct sk_buff *skb)
	157	{
4957faad	158	struct tcp_options_received tcp_opt;
c6aefafb GG	159	struct inet_request_sock *ireq;
	160	struct inet6_request_sock *ireq6;
	161	struct tcp_request_sock *treq;
	162	struct ipv6_pinfo *np = inet6_sk(sk);
	163	struct tcp_sock *tp = tcp_sk(sk);
	164	const struct tcphdr *th = tcp_hdr(skb);
	165	__u32 cookie = ntohl(th->ack_seq) - 1;
	166	struct sock *ret = sk;
	167	struct request_sock *req;
	168	int mss;
	169	struct dst_entry *dst;
	170	__u8 rcv_wscale;
f0e3d068	171	bool ecn_ok = false;
c6aefafb	172
af9b4738	173	if (!sysctl_tcp_syncookies \|\| !th->ack \|\| th->rst)
c6aefafb GG	174	goto out;
c6aefafb GG	175
a0f82f64	176	if (tcp_synq_no_recent_overflow(sk) \|\|
81eb6a14	177	(mss = __cookie_v6_check(ipv6_hdr(skb), th, cookie)) == 0) {
de0744af	178	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESFAILED);
c6aefafb GG	179	goto out;
	180	}
	181
de0744af	182	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESRECV);
c6aefafb	183
bb5b7c11 DM	184	/* check for timestamp cookie support */
bb5b7c11 DM	185	memset(&tcp_opt, 0, sizeof(tcp_opt));
1a2c6181	186	tcp_parse_options(skb, &tcp_opt, 0, NULL);
bb5b7c11	187
5d134f1c	188	if (!cookie_check_timestamp(&tcp_opt, sock_net(sk), &ecn_ok))
8c763681	189	goto out;
bb5b7c11	190
c6aefafb GG	191	ret = NULL;
	192	req = inet6_reqsk_alloc(&tcp6_request_sock_ops);
	193	if (!req)
	194	goto out;
	195
	196	ireq = inet_rsk(req);
	197	ireq6 = inet6_rsk(req);
	198	treq = tcp_rsk(req);
8336886f	199	treq->listener = NULL;
c6aefafb	200
1730554f FW	201	if (security_inet_conn_request(sk, skb, req))
1730554f FW	202	goto out_free;
c6aefafb GG	203
	204	req->mss = mss;
	205	ireq->rmt_port = th->source;
fd507037	206	ireq->loc_port = th->dest;
4e3fd7a0 AD	207	ireq6->rmt_addr = ipv6_hdr(skb)->saddr;
4e3fd7a0 AD	208	ireq6->loc_addr = ipv6_hdr(skb)->daddr;
c6aefafb GG	209	if (ipv6_opt_accepted(sk, skb) \|\|
	210	np->rxopt.bits.rxinfo \|\| np->rxopt.bits.rxoinfo \|\|
	211	np->rxopt.bits.rxhlim \|\| np->rxopt.bits.rxohlim) {
	212	atomic_inc(&skb->users);
	213	ireq6->pktopts = skb;
	214	}
	215
	216	ireq6->iif = sk->sk_bound_dev_if;
	217	/* So that link locals have meaning */
	218	if (!sk->sk_bound_dev_if &&
	219	ipv6_addr_type(&ireq6->rmt_addr) & IPV6_ADDR_LINKLOCAL)
	220	ireq6->iif = inet6_iif(skb);
	221
	222	req->expires = 0UL;
e6c022a4	223	req->num_retrans = 0;
172d69e6	224	ireq->ecn_ok = ecn_ok;
bb5b7c11	225	ireq->snd_wscale = tcp_opt.snd_wscale;
bb5b7c11 DM	226	ireq->sack_ok = tcp_opt.sack_ok;
	227	ireq->wscale_ok = tcp_opt.wscale_ok;
	228	ireq->tstamp_ok = tcp_opt.saw_tstamp;
	229	req->ts_recent = tcp_opt.saw_tstamp ? tcp_opt.rcv_tsval : 0;
9ad7c049	230	treq->snt_synack = tcp_opt.saw_tstamp ? tcp_opt.rcv_tsecr : 0;
c6aefafb GG	231	treq->rcv_isn = ntohl(th->seq) - 1;
	232	treq->snt_isn = cookie;
	233
	234	/*
	235	* We need to lookup the dst_entry to get the correct window size.
	236	* This is taken from tcp_v6_syn_recv_sock. Somebody please enlighten
	237	* me if there is a preferred way.
	238	*/
	239	{
20c59de2	240	struct in6_addr *final_p, final;
4c9483b2 DM	241	struct flowi6 fl6;
	242	memset(&fl6, 0, sizeof(fl6));
	243	fl6.flowi6_proto = IPPROTO_TCP;
4e3fd7a0	244	fl6.daddr = ireq6->rmt_addr;
4c9483b2	245	final_p = fl6_update_dst(&fl6, np->opt, &final);
4e3fd7a0	246	fl6.saddr = ireq6->loc_addr;
4c9483b2 DM	247	fl6.flowi6_oif = sk->sk_bound_dev_if;
4c9483b2 DM	248	fl6.flowi6_mark = sk->sk_mark;
1958b856 DM	249	fl6.fl6_dport = inet_rsk(req)->rmt_port;
1958b856 DM	250	fl6.fl6_sport = inet_sk(sk)->inet_sport;
4c9483b2 DM	251	security_req_classify_flow(req, flowi6_to_flowi(&fl6));
	252
	253	dst = ip6_dst_lookup_flow(sk, &fl6, final_p, false);
68d0c6d3	254	if (IS_ERR(dst))
1730554f	255	goto out_free;
c6aefafb GG	256	}
c6aefafb GG	257
4dfc2817	258	req->window_clamp = tp->window_clamp ? :dst_metric(dst, RTAX_WINDOW);
c6aefafb GG	259	tcp_select_initial_window(tcp_full_space(sk), req->mss,
c6aefafb GG	260	&req->rcv_wnd, &req->window_clamp,
31d12926	261	ireq->wscale_ok, &rcv_wscale,
31d12926	262	dst_metric(dst, RTAX_INITRWND));
c6aefafb GG	263
	264	ireq->rcv_wscale = rcv_wscale;
	265
	266	ret = get_cookie_sock(sk, skb, req, dst);
1730554f FW	267	out:
	268	return ret;
	269	out_free:
	270	reqsk_free(req);
	271	return NULL;
c6aefafb GG	272	}
c6aefafb GG	273