[linux-block.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)

/* RFC 2460, Section 8.3:
 * [ipv6 tcp] MSS must be computed as the maximum packet size minus 60 [..]
 *
 * Due to IPV6_MIN_MTU=1280 the lowest possible MSS is 1220, which allows
 * using higher values than ipv4 tcp syncookies.
 * The other values are chosen based on ethernet (1500 and 9k MTU), plus
 * one that accounts for common encap (PPPoe) overhead. Table must be sorted.
 */
static __u16 const msstab[] = {
	1280 - 60, /* IPV6_MIN_MTU - 60 */
	1480 - 60,
	1500 - 60,
	9000 - 60,
};

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 data)
{
	u32 count = tcp_cookie_time();
	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 diff, count = tcp_cookie_time();

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

	diff = (count - (cookie >> COOKIEBITS)) & ((__u32) -1 >> COOKIEBITS);
	if (diff >= MAX_SYNCOOKIE_AGE)
		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), 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);

	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 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);
	treq = tcp_rsk(req);
	treq->listener = NULL;

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

	req->mss = mss;
	ireq->ir_rmt_port = th->source;
	ireq->ir_num = ntohs(th->dest);
	ireq->ir_v6_rmt_addr = ipv6_hdr(skb)->saddr;
	ireq->ir_v6_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);
		ireq->pktopts = skb;
	}

	ireq->ir_iif = sk->sk_bound_dev_if;
	/* So that link locals have meaning */
	if (!sk->sk_bound_dev_if &&
	    ipv6_addr_type(&ireq->ir_v6_rmt_addr) & IPV6_ADDR_LINKLOCAL)
		ireq->ir_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 = ireq->ir_v6_rmt_addr;
		final_p = fl6_update_dst(&fl6, np->opt, &final);
		fl6.saddr = ireq->ir_v6_loc_addr;
		fl6.flowi6_oif = sk->sk_bound_dev_if;
		fl6.flowi6_mark = sk->sk_mark;
		fl6.fl6_dport = ireq->ir_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
08629354 FW	27	/* RFC 2460, Section 8.3:
	28	* [ipv6 tcp] MSS must be computed as the maximum packet size minus 60 [..]
	29	*
	30	* Due to IPV6_MIN_MTU=1280 the lowest possible MSS is 1220, which allows
	31	* using higher values than ipv4 tcp syncookies.
	32	* The other values are chosen based on ethernet (1500 and 9k MTU), plus
	33	* one that accounts for common encap (PPPoe) overhead. Table must be sorted.
	34	*/
c6aefafb	35	static __u16 const msstab[] = {
08629354	36	1280 - 60, /* IPV6_MIN_MTU - 60 */
5918e2fb FW	37	1480 - 60,
5918e2fb FW	38	1500 - 60,
5918e2fb	39	9000 - 60,
c6aefafb	40	};
c6aefafb	41
c6aefafb GG	42	static inline struct sock get_cookie_sock(struct sock sk, struct sk_buff *skb,
	43	struct request_sock *req,
	44	struct dst_entry *dst)
	45	{
	46	struct inet_connection_sock *icsk = inet_csk(sk);
	47	struct sock *child;
	48
	49	child = icsk->icsk_af_ops->syn_recv_sock(sk, skb, req, dst);
	50	if (child)
	51	inet_csk_reqsk_queue_add(sk, req, child);
	52	else
	53	reqsk_free(req);
	54
	55	return child;
	56	}
	57
245b2e70 TH	58	static DEFINE_PER_CPU(__u32 [16 + 5 + SHA_WORKSPACE_WORDS],
245b2e70 TH	59	ipv6_cookie_scratch);
c6aefafb	60
b71d1d42	61	static u32 cookie_hash(const struct in6_addr saddr, const struct in6_addr daddr,
c6aefafb GG	62	__be16 sport, __be16 dport, u32 count, int c)
c6aefafb GG	63	{
245b2e70	64	__u32 *tmp = __get_cpu_var(ipv6_cookie_scratch);
c6aefafb GG	65
	66	/*
	67	* we have 320 bits of information to hash, copy in the remaining
	68	* 192 bits required for sha_transform, from the syncookie_secret
	69	* and overwrite the digest with the secret
	70	*/
	71	memcpy(tmp + 10, syncookie_secret[c], 44);
	72	memcpy(tmp, saddr, 16);
	73	memcpy(tmp + 4, daddr, 16);
	74	tmp[8] = ((__force u32)sport << 16) + (__force u32)dport;
	75	tmp[9] = count;
	76	sha_transform(tmp + 16, (__u8 *)tmp, tmp + 16 + 5);
	77
	78	return tmp[17];
	79	}
	80
b71d1d42 ED	81	static __u32 secure_tcp_syn_cookie(const struct in6_addr *saddr,
b71d1d42 ED	82	const struct in6_addr *daddr,
c6aefafb	83	__be16 sport, __be16 dport, __u32 sseq,
8c27bd75	84	__u32 data)
c6aefafb	85	{
8c27bd75	86	u32 count = tcp_cookie_time();
c6aefafb GG	87	return (cookie_hash(saddr, daddr, sport, dport, 0, 0) +
	88	sseq + (count << COOKIEBITS) +
	89	((cookie_hash(saddr, daddr, sport, dport, count, 1) + data)
	90	& COOKIEMASK));
	91	}
	92
b71d1d42 ED	93	static __u32 check_tcp_syn_cookie(__u32 cookie, const struct in6_addr *saddr,
b71d1d42 ED	94	const struct in6_addr *daddr, __be16 sport,
8c27bd75	95	__be16 dport, __u32 sseq)
c6aefafb	96	{
8c27bd75	97	__u32 diff, count = tcp_cookie_time();
c6aefafb GG	98
	99	cookie -= cookie_hash(saddr, daddr, sport, dport, 0, 0) + sseq;
	100
	101	diff = (count - (cookie >> COOKIEBITS)) & ((__u32) -1 >> COOKIEBITS);
8c27bd75	102	if (diff >= MAX_SYNCOOKIE_AGE)
c6aefafb GG	103	return (__u32)-1;
	104
	105	return (cookie -
	106	cookie_hash(saddr, daddr, sport, dport, count - diff, 1))
	107	& COOKIEMASK;
	108	}
	109
81eb6a14 PM	110	u32 __cookie_v6_init_sequence(const struct ipv6hdr *iph,
81eb6a14 PM	111	const struct tcphdr th, __u16 mssp)
c6aefafb	112	{
c6aefafb GG	113	int mssind;
	114	const __u16 mss = *mssp;
	115
5918e2fb FW	116	for (mssind = ARRAY_SIZE(msstab) - 1; mssind ; mssind--)
	117	if (mss >= msstab[mssind])
	118	break;
	119
	120	*mssp = msstab[mssind];
c6aefafb	121
c6aefafb	122	return secure_tcp_syn_cookie(&iph->saddr, &iph->daddr, th->source,
8c27bd75	123	th->dest, ntohl(th->seq), mssind);
c6aefafb	124	}
81eb6a14	125	EXPORT_SYMBOL_GPL(__cookie_v6_init_sequence);
c6aefafb	126
81eb6a14	127	__u32 cookie_v6_init_sequence(struct sock sk, const struct sk_buff skb, __u16 *mssp)
c6aefafb	128	{
b71d1d42	129	const struct ipv6hdr *iph = ipv6_hdr(skb);
c6aefafb	130	const struct tcphdr *th = tcp_hdr(skb);
81eb6a14 PM	131
	132	tcp_synq_overflow(sk);
	133	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESSENT);
	134
	135	return __cookie_v6_init_sequence(iph, th, mssp);
	136	}
	137
	138	int __cookie_v6_check(const struct ipv6hdr iph, const struct tcphdr th,
	139	__u32 cookie)
	140	{
c6aefafb GG	141	__u32 seq = ntohl(th->seq) - 1;
c6aefafb GG	142	__u32 mssind = check_tcp_syn_cookie(cookie, &iph->saddr, &iph->daddr,
8c27bd75	143	th->source, th->dest, seq);
c6aefafb	144
5918e2fb	145	return mssind < ARRAY_SIZE(msstab) ? msstab[mssind] : 0;
c6aefafb	146	}
81eb6a14	147	EXPORT_SYMBOL_GPL(__cookie_v6_check);
c6aefafb GG	148
	149	struct sock cookie_v6_check(struct sock sk, struct sk_buff *skb)
	150	{
4957faad	151	struct tcp_options_received tcp_opt;
c6aefafb	152	struct inet_request_sock *ireq;
c6aefafb GG	153	struct tcp_request_sock *treq;
	154	struct ipv6_pinfo *np = inet6_sk(sk);
	155	struct tcp_sock *tp = tcp_sk(sk);
	156	const struct tcphdr *th = tcp_hdr(skb);
	157	__u32 cookie = ntohl(th->ack_seq) - 1;
	158	struct sock *ret = sk;
	159	struct request_sock *req;
	160	int mss;
	161	struct dst_entry *dst;
	162	__u8 rcv_wscale;
f0e3d068	163	bool ecn_ok = false;
c6aefafb	164
af9b4738	165	if (!sysctl_tcp_syncookies \|\| !th->ack \|\| th->rst)
c6aefafb GG	166	goto out;
c6aefafb GG	167
a0f82f64	168	if (tcp_synq_no_recent_overflow(sk) \|\|
81eb6a14	169	(mss = __cookie_v6_check(ipv6_hdr(skb), th, cookie)) == 0) {
de0744af	170	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESFAILED);
c6aefafb GG	171	goto out;
	172	}
	173
de0744af	174	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESRECV);
c6aefafb	175
bb5b7c11 DM	176	/* check for timestamp cookie support */
bb5b7c11 DM	177	memset(&tcp_opt, 0, sizeof(tcp_opt));
1a2c6181	178	tcp_parse_options(skb, &tcp_opt, 0, NULL);
bb5b7c11	179
5d134f1c	180	if (!cookie_check_timestamp(&tcp_opt, sock_net(sk), &ecn_ok))
8c763681	181	goto out;
bb5b7c11	182
c6aefafb GG	183	ret = NULL;
	184	req = inet6_reqsk_alloc(&tcp6_request_sock_ops);
	185	if (!req)
	186	goto out;
	187
	188	ireq = inet_rsk(req);
c6aefafb	189	treq = tcp_rsk(req);
8336886f	190	treq->listener = NULL;
c6aefafb	191
1730554f FW	192	if (security_inet_conn_request(sk, skb, req))
1730554f FW	193	goto out_free;
c6aefafb GG	194
c6aefafb GG	195	req->mss = mss;
634fb979	196	ireq->ir_rmt_port = th->source;
b44084c2	197	ireq->ir_num = ntohs(th->dest);
634fb979 ED	198	ireq->ir_v6_rmt_addr = ipv6_hdr(skb)->saddr;
634fb979 ED	199	ireq->ir_v6_loc_addr = ipv6_hdr(skb)->daddr;
c6aefafb GG	200	if (ipv6_opt_accepted(sk, skb) \|\|
	201	np->rxopt.bits.rxinfo \|\| np->rxopt.bits.rxoinfo \|\|
	202	np->rxopt.bits.rxhlim \|\| np->rxopt.bits.rxohlim) {
	203	atomic_inc(&skb->users);
634fb979	204	ireq->pktopts = skb;
c6aefafb GG	205	}
c6aefafb GG	206
634fb979	207	ireq->ir_iif = sk->sk_bound_dev_if;
c6aefafb GG	208	/* So that link locals have meaning */
c6aefafb GG	209	if (!sk->sk_bound_dev_if &&
634fb979 ED	210	ipv6_addr_type(&ireq->ir_v6_rmt_addr) & IPV6_ADDR_LINKLOCAL)
634fb979 ED	211	ireq->ir_iif = inet6_iif(skb);
c6aefafb GG	212
c6aefafb GG	213	req->expires = 0UL;
e6c022a4	214	req->num_retrans = 0;
172d69e6	215	ireq->ecn_ok = ecn_ok;
bb5b7c11	216	ireq->snd_wscale = tcp_opt.snd_wscale;
bb5b7c11 DM	217	ireq->sack_ok = tcp_opt.sack_ok;
	218	ireq->wscale_ok = tcp_opt.wscale_ok;
	219	ireq->tstamp_ok = tcp_opt.saw_tstamp;
	220	req->ts_recent = tcp_opt.saw_tstamp ? tcp_opt.rcv_tsval : 0;
9ad7c049	221	treq->snt_synack = tcp_opt.saw_tstamp ? tcp_opt.rcv_tsecr : 0;
c6aefafb GG	222	treq->rcv_isn = ntohl(th->seq) - 1;
	223	treq->snt_isn = cookie;
	224
	225	/*
	226	* We need to lookup the dst_entry to get the correct window size.
	227	* This is taken from tcp_v6_syn_recv_sock. Somebody please enlighten
	228	* me if there is a preferred way.
	229	*/
	230	{
20c59de2	231	struct in6_addr *final_p, final;
4c9483b2 DM	232	struct flowi6 fl6;
	233	memset(&fl6, 0, sizeof(fl6));
	234	fl6.flowi6_proto = IPPROTO_TCP;
634fb979	235	fl6.daddr = ireq->ir_v6_rmt_addr;
4c9483b2	236	final_p = fl6_update_dst(&fl6, np->opt, &final);
634fb979	237	fl6.saddr = ireq->ir_v6_loc_addr;
4c9483b2 DM	238	fl6.flowi6_oif = sk->sk_bound_dev_if;
4c9483b2 DM	239	fl6.flowi6_mark = sk->sk_mark;
634fb979	240	fl6.fl6_dport = ireq->ir_rmt_port;
1958b856	241	fl6.fl6_sport = inet_sk(sk)->inet_sport;
4c9483b2 DM	242	security_req_classify_flow(req, flowi6_to_flowi(&fl6));
	243
	244	dst = ip6_dst_lookup_flow(sk, &fl6, final_p, false);
68d0c6d3	245	if (IS_ERR(dst))
1730554f	246	goto out_free;
c6aefafb GG	247	}
c6aefafb GG	248
4dfc2817	249	req->window_clamp = tp->window_clamp ? :dst_metric(dst, RTAX_WINDOW);
c6aefafb GG	250	tcp_select_initial_window(tcp_full_space(sk), req->mss,
c6aefafb GG	251	&req->rcv_wnd, &req->window_clamp,
31d12926	252	ireq->wscale_ok, &rcv_wscale,
31d12926	253	dst_metric(dst, RTAX_INITRWND));
c6aefafb GG	254
	255	ireq->rcv_wscale = rcv_wscale;
	256
	257	ret = get_cookie_sock(sk, skb, req, dst);
1730554f FW	258	out:
	259	return ret;
	260	out_free:
	261	reqsk_free(req);
	262	return NULL;
c6aefafb GG	263	}
c6aefafb GG	264