[deliverable/linux.git] / net / ipv4 / syncookies.c

/*
 *  Syncookies implementation for the Linux kernel
 *
 *  Copyright (C) 1997 Andi Kleen
 *  Based on ideas by D.J.Bernstein and Eric Schenk.
 *
 *	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/slab.h>
#include <linux/random.h>
#include <linux/cryptohash.h>
#include <linux/kernel.h>
#include <linux/export.h>
#include <net/tcp.h>
#include <net/route.h>

/* Timestamps: lowest bits store TCP options */
#define TSBITS 6
#define TSMASK (((__u32)1 << TSBITS) - 1)

extern int sysctl_tcp_syncookies;

__u32 syncookie_secret[2][16-4+SHA_DIGEST_WORDS];
EXPORT_SYMBOL(syncookie_secret);

static __init int init_syncookies(void)
{
	get_random_bytes(syncookie_secret, sizeof(syncookie_secret));
	return 0;
}
__initcall(init_syncookies);

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

static DEFINE_PER_CPU(__u32 [16 + 5 + SHA_WORKSPACE_WORDS],
		      ipv4_cookie_scratch);

static u32 cookie_hash(__be32 saddr, __be32 daddr, __be16 sport, __be16 dport,
		       u32 count, int c)
{
	__u32 *tmp = __get_cpu_var(ipv4_cookie_scratch);

	memcpy(tmp + 4, syncookie_secret[c], sizeof(syncookie_secret[c]));
	tmp[0] = (__force u32)saddr;
	tmp[1] = (__force u32)daddr;
	tmp[2] = ((__force u32)sport << 16) + (__force u32)dport;
	tmp[3] = count;
	sha_transform(tmp + 16, (__u8 *)tmp, tmp + 16 + 5);

	return tmp[17];
}


/*
 * when syncookies are in effect and tcp timestamps are enabled we encode
 * tcp options in the lower bits of the timestamp value that will be
 * sent in the syn-ack.
 * Since subsequent timestamps use the normal tcp_time_stamp value, we
 * must make sure that the resulting initial timestamp is <= tcp_time_stamp.
 */
__u32 cookie_init_timestamp(struct request_sock *req)
{
	struct inet_request_sock *ireq;
	u32 ts, ts_now = tcp_time_stamp;
	u32 options = 0;

	ireq = inet_rsk(req);

	options = ireq->wscale_ok ? ireq->snd_wscale : 0xf;
	options |= ireq->sack_ok << 4;
	options |= ireq->ecn_ok << 5;

	ts = ts_now & ~TSMASK;
	ts |= options;
	if (ts > ts_now) {
		ts >>= TSBITS;
		ts--;
		ts <<= TSBITS;
		ts |= options;
	}
	return ts;
}


static __u32 secure_tcp_syn_cookie(__be32 saddr, __be32 daddr, __be16 sport,
				   __be16 dport, __u32 sseq, __u32 data)
{
	/*
	 * Compute the secure sequence number.
	 * The output should be:
	 *   HASH(sec1,saddr,sport,daddr,dport,sec1) + sseq + (count * 2^24)
	 *      + (HASH(sec2,saddr,sport,daddr,dport,count,sec2) % 2^24).
	 * Where sseq is their sequence number and count increases every
	 * minute by 1.
	 * As an extra hack, we add a small "data" value that encodes the
	 * MSS into the second hash value.
	 */
	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));
}

/*
 * This retrieves the small "data" value from the syncookie.
 * If the syncookie is bad, the data returned will be out of
 * range.  This must be checked by the caller.
 *
 * The count value used to generate the cookie must be less than
 * MAX_SYNCOOKIE_AGE minutes in the past.
 * The return value (__u32)-1 if this test fails.
 */
static __u32 check_tcp_syn_cookie(__u32 cookie, __be32 saddr, __be32 daddr,
				  __be16 sport, __be16 dport, __u32 sseq)
{
	u32 diff, count = tcp_cookie_time();

	/* Strip away the layers from the cookie */
	cookie -= cookie_hash(saddr, daddr, sport, dport, 0, 0) + sseq;

	/* Cookie is now reduced to (count * 2^24) ^ (hash % 2^24) */
	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;	/* Leaving the data behind */
}

/*
 * MSS Values are chosen based on the 2011 paper
 * 'An Analysis of TCP Maximum Segement Sizes' by S. Alcock and R. Nelson.
 * Values ..
 *  .. lower than 536 are rare (< 0.2%)
 *  .. between 537 and 1299 account for less than < 1.5% of observed values
 *  .. in the 1300-1349 range account for about 15 to 20% of observed mss values
 *  .. exceeding 1460 are very rare (< 0.04%)
 *
 *  1460 is the single most frequently announced mss value (30 to 46% depending
 *  on monitor location).  Table must be sorted.
 */
static __u16 const msstab[] = {
	536,
	1300,
	1440,	/* 1440, 1452: PPPoE */
	1460,
};

/*
 * Generate a syncookie.  mssp points to the mss, which is returned
 * rounded down to the value encoded in the cookie.
 */
u32 __cookie_v4_init_sequence(const struct iphdr *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_v4_init_sequence);

__u32 cookie_v4_init_sequence(struct sock *sk, struct sk_buff *skb, __u16 *mssp)
{
	const struct iphdr *iph = ip_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_v4_init_sequence(iph, th, mssp);
}

/*
 * Check if a ack sequence number is a valid syncookie.
 * Return the decoded mss if it is, or 0 if not.
 */
int __cookie_v4_check(const struct iphdr *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_v4_check);

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;
}


/*
 * when syncookies are in effect and tcp timestamps are enabled we stored
 * additional tcp options in the timestamp.
 * This extracts these options from the timestamp echo.
 *
 * The lowest 4 bits store snd_wscale.
 * next 2 bits indicate SACK and ECN support.
 *
 * return false if we decode an option that should not be.
 */
bool cookie_check_timestamp(struct tcp_options_received *tcp_opt,
			struct net *net, bool *ecn_ok)
{
	/* echoed timestamp, lowest bits contain options */
	u32 options = tcp_opt->rcv_tsecr & TSMASK;

	if (!tcp_opt->saw_tstamp)  {
		tcp_clear_options(tcp_opt);
		return true;
	}

	if (!sysctl_tcp_timestamps)
		return false;

	tcp_opt->sack_ok = (options & (1 << 4)) ? TCP_SACK_SEEN : 0;
	*ecn_ok = (options >> 5) & 1;
	if (*ecn_ok && !net->ipv4.sysctl_tcp_ecn)
		return false;

	if (tcp_opt->sack_ok && !sysctl_tcp_sack)
		return false;

	if ((options & 0xf) == 0xf)
		return true; /* no window scaling */

	tcp_opt->wscale_ok = 1;
	tcp_opt->snd_wscale = options & 0xf;
	return sysctl_tcp_window_scaling != 0;
}
EXPORT_SYMBOL(cookie_check_timestamp);

struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb,
			     struct ip_options *opt)
{
	struct tcp_options_received tcp_opt;
	struct inet_request_sock *ireq;
	struct tcp_request_sock *treq;
	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 rtable *rt;
	__u8 rcv_wscale;
	bool ecn_ok = false;
	struct flowi4 fl4;

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

	if (tcp_synq_no_recent_overflow(sk) ||
	    (mss = __cookie_v4_check(ip_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 = inet_reqsk_alloc(&tcp_request_sock_ops); /* for safety */
	if (!req)
		goto out;

	ireq = inet_rsk(req);
	treq = tcp_rsk(req);
	treq->rcv_isn		= ntohl(th->seq) - 1;
	treq->snt_isn		= cookie;
	req->mss		= mss;
	ireq->ir_num		= ntohs(th->dest);
	ireq->ir_rmt_port	= th->source;
	ireq->ir_loc_addr	= ip_hdr(skb)->daddr;
	ireq->ir_rmt_addr	= ip_hdr(skb)->saddr;
	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->listener		= NULL;

	/* We throwed the options of the initial SYN away, so we hope
	 * the ACK carries the same options again (see RFC1122 4.2.3.8)
	 */
	if (opt && opt->optlen) {
		int opt_size = sizeof(struct ip_options_rcu) + opt->optlen;

		ireq->opt = kmalloc(opt_size, GFP_ATOMIC);
		if (ireq->opt != NULL && ip_options_echo(&ireq->opt->opt, skb)) {
			kfree(ireq->opt);
			ireq->opt = NULL;
		}
	}

	if (security_inet_conn_request(sk, skb, req)) {
		reqsk_free(req);
		goto out;
	}

	req->expires	= 0UL;
	req->num_retrans = 0;

	/*
	 * We need to lookup the route here to get at the correct
	 * window size. We should better make sure that the window size
	 * hasn't changed since we received the original syn, but I see
	 * no easy way to do this.
	 */
	flowi4_init_output(&fl4, sk->sk_bound_dev_if, sk->sk_mark,
			   RT_CONN_FLAGS(sk), RT_SCOPE_UNIVERSE, IPPROTO_TCP,
			   inet_sk_flowi_flags(sk),
			   (opt && opt->srr) ? opt->faddr : ireq->ir_rmt_addr,
			   ireq->ir_loc_addr, th->source, th->dest);
	security_req_classify_flow(req, flowi4_to_flowi(&fl4));
	rt = ip_route_output_key(sock_net(sk), &fl4);
	if (IS_ERR(rt)) {
		reqsk_free(req);
		goto out;
	}

	/* Try to redo what tcp_v4_send_synack did. */
	req->window_clamp = tp->window_clamp ? :dst_metric(&rt->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(&rt->dst, RTAX_INITRWND));

	ireq->rcv_wscale  = rcv_wscale;

	ret = get_cookie_sock(sk, skb, req, &rt->dst);
	/* ip_queue_xmit() depends on our flow being setup
	 * Normal sockets get it right from inet_csk_route_child_sock()
	 */
	if (ret)
		inet_sk(ret)->cork.fl.u.ip4 = fl4;
out:	return ret;
}
Commit	Line	Data
1da177e4 LT	1	/*
	2	* Syncookies implementation for the Linux kernel
	3	*
	4	* Copyright (C) 1997 Andi Kleen
e905a9ed	5	* Based on ideas by D.J.Bernstein and Eric Schenk.
1da177e4 LT	6	*
	7	* This program is free software; you can redistribute it and/or
	8	* modify it under the terms of the GNU General Public License
	9	* as published by the Free Software Foundation; either version
	10	* 2 of the License, or (at your option) any later version.
1da177e4 LT	11	*/
	12
	13	#include <linux/tcp.h>
	14	#include <linux/slab.h>
	15	#include <linux/random.h>
	16	#include <linux/cryptohash.h>
	17	#include <linux/kernel.h>
bc3b2d7f	18	#include <linux/export.h>
1da177e4	19	#include <net/tcp.h>
86b08d86	20	#include <net/route.h>
1da177e4	21
734f614b	22	/* Timestamps: lowest bits store TCP options */
172d69e6	23	#define TSBITS 6
4dfc2817 FW	24	#define TSMASK (((__u32)1 << TSBITS) - 1)
4dfc2817 FW	25
1da177e4 LT	26	extern int sysctl_tcp_syncookies;
1da177e4 LT	27
2051f11f	28	__u32 syncookie_secret[2][16-4+SHA_DIGEST_WORDS];
c6aefafb	29	EXPORT_SYMBOL(syncookie_secret);
1da177e4 LT	30
	31	static __init int init_syncookies(void)
	32	{
	33	get_random_bytes(syncookie_secret, sizeof(syncookie_secret));
	34	return 0;
	35	}
c6aefafb	36	__initcall(init_syncookies);
1da177e4 LT	37
	38	#define COOKIEBITS 24 /* Upper bits store count */
	39	#define COOKIEMASK (((__u32)1 << COOKIEBITS) - 1)
	40
245b2e70 TH	41	static DEFINE_PER_CPU(__u32 [16 + 5 + SHA_WORKSPACE_WORDS],
245b2e70 TH	42	ipv4_cookie_scratch);
11baab7a	43
714e85be	44	static u32 cookie_hash(__be32 saddr, __be32 daddr, __be16 sport, __be16 dport,
1da177e4 LT	45	u32 count, int c)
1da177e4 LT	46	{
245b2e70	47	__u32 *tmp = __get_cpu_var(ipv4_cookie_scratch);
1da177e4	48
2051f11f	49	memcpy(tmp + 4, syncookie_secret[c], sizeof(syncookie_secret[c]));
714e85be AV	50	tmp[0] = (__force u32)saddr;
	51	tmp[1] = (__force u32)daddr;
	52	tmp[2] = ((__force u32)sport << 16) + (__force u32)dport;
1da177e4 LT	53	tmp[3] = count;
	54	sha_transform(tmp + 16, (__u8 *)tmp, tmp + 16 + 5);
	55
	56	return tmp[17];
	57	}
	58
4dfc2817 FW	59
	60	/*
	61	* when syncookies are in effect and tcp timestamps are enabled we encode
734f614b	62	* tcp options in the lower bits of the timestamp value that will be
4dfc2817 FW	63	* sent in the syn-ack.
	64	* Since subsequent timestamps use the normal tcp_time_stamp value, we
	65	* must make sure that the resulting initial timestamp is <= tcp_time_stamp.
	66	*/
	67	__u32 cookie_init_timestamp(struct request_sock *req)
	68	{
	69	struct inet_request_sock *ireq;
	70	u32 ts, ts_now = tcp_time_stamp;
	71	u32 options = 0;
	72
	73	ireq = inet_rsk(req);
734f614b FW	74
	75	options = ireq->wscale_ok ? ireq->snd_wscale : 0xf;
	76	options \|= ireq->sack_ok << 4;
172d69e6	77	options \|= ireq->ecn_ok << 5;
4dfc2817 FW	78
	79	ts = ts_now & ~TSMASK;
	80	ts \|= options;
	81	if (ts > ts_now) {
	82	ts >>= TSBITS;
	83	ts--;
	84	ts <<= TSBITS;
	85	ts \|= options;
	86	}
	87	return ts;
	88	}
	89
	90
714e85be	91	static __u32 secure_tcp_syn_cookie(__be32 saddr, __be32 daddr, __be16 sport,
8c27bd75	92	__be16 dport, __u32 sseq, __u32 data)
1da177e4 LT	93	{
	94	/*
	95	* Compute the secure sequence number.
	96	* The output should be:
e905a9ed	97	* HASH(sec1,saddr,sport,daddr,dport,sec1) + sseq + (count * 2^24)
1da177e4 LT	98	* + (HASH(sec2,saddr,sport,daddr,dport,count,sec2) % 2^24).
	99	* Where sseq is their sequence number and count increases every
	100	* minute by 1.
	101	* As an extra hack, we add a small "data" value that encodes the
	102	* MSS into the second hash value.
	103	*/
8c27bd75	104	u32 count = tcp_cookie_time();
1da177e4 LT	105	return (cookie_hash(saddr, daddr, sport, dport, 0, 0) +
	106	sseq + (count << COOKIEBITS) +
	107	((cookie_hash(saddr, daddr, sport, dport, count, 1) + data)
	108	& COOKIEMASK));
	109	}
	110
	111	/*
	112	* This retrieves the small "data" value from the syncookie.
	113	* If the syncookie is bad, the data returned will be out of
	114	* range. This must be checked by the caller.
	115	*
8c27bd75 FW	116	* The count value used to generate the cookie must be less than
	117	* MAX_SYNCOOKIE_AGE minutes in the past.
	118	* The return value (__u32)-1 if this test fails.
1da177e4	119	*/
714e85be	120	static __u32 check_tcp_syn_cookie(__u32 cookie, __be32 saddr, __be32 daddr,
8c27bd75	121	__be16 sport, __be16 dport, __u32 sseq)
1da177e4	122	{
8c27bd75	123	u32 diff, count = tcp_cookie_time();
1da177e4 LT	124
	125	/* Strip away the layers from the cookie */
	126	cookie -= cookie_hash(saddr, daddr, sport, dport, 0, 0) + sseq;
	127
	128	/* Cookie is now reduced to (count * 2^24) ^ (hash % 2^24) */
	129	diff = (count - (cookie >> COOKIEBITS)) & ((__u32) - 1 >> COOKIEBITS);
8c27bd75	130	if (diff >= MAX_SYNCOOKIE_AGE)
1da177e4 LT	131	return (__u32)-1;
	132
	133	return (cookie -
	134	cookie_hash(saddr, daddr, sport, dport, count - diff, 1))
	135	& COOKIEMASK; /* Leaving the data behind */
	136	}
	137
e905a9ed	138	/*
08629354 FW	139	* MSS Values are chosen based on the 2011 paper
	140	* 'An Analysis of TCP Maximum Segement Sizes' by S. Alcock and R. Nelson.
	141	* Values ..
	142	* .. lower than 536 are rare (< 0.2%)
	143	* .. between 537 and 1299 account for less than < 1.5% of observed values
	144	* .. in the 1300-1349 range account for about 15 to 20% of observed mss values
	145	* .. exceeding 1460 are very rare (< 0.04%)
5918e2fb	146	*
08629354 FW	147	* 1460 is the single most frequently announced mss value (30 to 46% depending
08629354 FW	148	* on monitor location). Table must be sorted.
1da177e4 LT	149	*/
1da177e4 LT	150	static __u16 const msstab[] = {
5918e2fb	151	536,
08629354 FW	152	1300,
08629354 FW	153	1440, /* 1440, 1452: PPPoE */
5918e2fb	154	1460,
1da177e4	155	};
1da177e4 LT	156
	157	/*
	158	* Generate a syncookie. mssp points to the mss, which is returned
	159	* rounded down to the value encoded in the cookie.
	160	*/
0198230b PM	161	u32 __cookie_v4_init_sequence(const struct iphdr iph, const struct tcphdr th,
0198230b PM	162	u16 *mssp)
1da177e4	163	{
1da177e4 LT	164	int mssind;
	165	const __u16 mss = *mssp;
	166
5918e2fb FW	167	for (mssind = ARRAY_SIZE(msstab) - 1; mssind ; mssind--)
	168	if (mss >= msstab[mssind])
	169	break;
	170	*mssp = msstab[mssind];
1da177e4	171
aa8223c7 ACM	172	return secure_tcp_syn_cookie(iph->saddr, iph->daddr,
aa8223c7 ACM	173	th->source, th->dest, ntohl(th->seq),
8c27bd75	174	mssind);
1da177e4	175	}
0198230b PM	176	EXPORT_SYMBOL_GPL(__cookie_v4_init_sequence);
	177
	178	__u32 cookie_v4_init_sequence(struct sock sk, struct sk_buff skb, __u16 *mssp)
	179	{
	180	const struct iphdr *iph = ip_hdr(skb);
	181	const struct tcphdr *th = tcp_hdr(skb);
	182
	183	tcp_synq_overflow(sk);
	184	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESSENT);
	185
	186	return __cookie_v4_init_sequence(iph, th, mssp);
	187	}
1da177e4	188
e905a9ed YH	189	/*
e905a9ed YH	190	* Check if a ack sequence number is a valid syncookie.
1da177e4 LT	191	* Return the decoded mss if it is, or 0 if not.
1da177e4 LT	192	*/
0198230b PM	193	int __cookie_v4_check(const struct iphdr iph, const struct tcphdr th,
0198230b PM	194	u32 cookie)
1da177e4	195	{
aa8223c7 ACM	196	__u32 seq = ntohl(th->seq) - 1;
aa8223c7 ACM	197	__u32 mssind = check_tcp_syn_cookie(cookie, iph->saddr, iph->daddr,
8c27bd75	198	th->source, th->dest, seq);
1da177e4	199
5918e2fb	200	return mssind < ARRAY_SIZE(msstab) ? msstab[mssind] : 0;
1da177e4	201	}
0198230b	202	EXPORT_SYMBOL_GPL(__cookie_v4_check);
1da177e4	203
1da177e4	204	static inline struct sock get_cookie_sock(struct sock sk, struct sk_buff *skb,
60236fdd	205	struct request_sock *req,
1da177e4 LT	206	struct dst_entry *dst)
1da177e4 LT	207	{
8292a17a	208	struct inet_connection_sock *icsk = inet_csk(sk);
1da177e4 LT	209	struct sock *child;
1da177e4 LT	210
8292a17a	211	child = icsk->icsk_af_ops->syn_recv_sock(sk, skb, req, dst);
1da177e4	212	if (child)
463c84b9	213	inet_csk_reqsk_queue_add(sk, req, child);
1da177e4	214	else
60236fdd	215	reqsk_free(req);
1da177e4 LT	216
	217	return child;
	218	}
	219
4dfc2817 FW	220
	221	/*
	222	* when syncookies are in effect and tcp timestamps are enabled we stored
	223	* additional tcp options in the timestamp.
	224	* This extracts these options from the timestamp echo.
	225	*
734f614b	226	* The lowest 4 bits store snd_wscale.
172d69e6	227	* next 2 bits indicate SACK and ECN support.
8c763681 FW	228	*
8c763681 FW	229	* return false if we decode an option that should not be.
4dfc2817	230	*/
5d134f1c HFS	231	bool cookie_check_timestamp(struct tcp_options_received *tcp_opt,
5d134f1c HFS	232	struct net net, bool ecn_ok)
4dfc2817	233	{
734f614b	234	/* echoed timestamp, lowest bits contain options */
4dfc2817 FW	235	u32 options = tcp_opt->rcv_tsecr & TSMASK;
4dfc2817 FW	236
8c763681 FW	237	if (!tcp_opt->saw_tstamp) {
	238	tcp_clear_options(tcp_opt);
	239	return true;
	240	}
	241
	242	if (!sysctl_tcp_timestamps)
	243	return false;
	244
ab56222a	245	tcp_opt->sack_ok = (options & (1 << 4)) ? TCP_SACK_SEEN : 0;
172d69e6	246	*ecn_ok = (options >> 5) & 1;
5d134f1c	247	if (*ecn_ok && !net->ipv4.sysctl_tcp_ecn)
172d69e6	248	return false;
4dfc2817	249
8c763681 FW	250	if (tcp_opt->sack_ok && !sysctl_tcp_sack)
8c763681 FW	251	return false;
4dfc2817	252
734f614b FW	253	if ((options & 0xf) == 0xf)
	254	return true; /* no window scaling */
	255
	256	tcp_opt->wscale_ok = 1;
	257	tcp_opt->snd_wscale = options & 0xf;
	258	return sysctl_tcp_window_scaling != 0;
4dfc2817 FW	259	}
	260	EXPORT_SYMBOL(cookie_check_timestamp);
	261
1da177e4 LT	262	struct sock cookie_v4_check(struct sock sk, struct sk_buff *skb,
	263	struct ip_options *opt)
	264	{
4957faad	265	struct tcp_options_received tcp_opt;
2e6599cb ACM	266	struct inet_request_sock *ireq;
2e6599cb ACM	267	struct tcp_request_sock *treq;
1da177e4	268	struct tcp_sock *tp = tcp_sk(sk);
aa8223c7 ACM	269	const struct tcphdr *th = tcp_hdr(skb);
aa8223c7 ACM	270	__u32 cookie = ntohl(th->ack_seq) - 1;
1da177e4	271	struct sock *ret = sk;
e905a9ed YH	272	struct request_sock *req;
	273	int mss;
	274	struct rtable *rt;
1da177e4	275	__u8 rcv_wscale;
f0e3d068	276	bool ecn_ok = false;
dfd25fff	277	struct flowi4 fl4;
1da177e4	278
af9b4738	279	if (!sysctl_tcp_syncookies \|\| !th->ack \|\| th->rst)
1da177e4 LT	280	goto out;
1da177e4 LT	281
a0f82f64	282	if (tcp_synq_no_recent_overflow(sk) \|\|
0198230b	283	(mss = __cookie_v4_check(ip_hdr(skb), th, cookie)) == 0) {
de0744af	284	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESFAILED);
1da177e4 LT	285	goto out;
	286	}
	287
de0744af	288	NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_SYNCOOKIESRECV);
1da177e4	289
bb5b7c11 DM	290	/* check for timestamp cookie support */
bb5b7c11 DM	291	memset(&tcp_opt, 0, sizeof(tcp_opt));
1a2c6181	292	tcp_parse_options(skb, &tcp_opt, 0, NULL);
bb5b7c11	293
5d134f1c	294	if (!cookie_check_timestamp(&tcp_opt, sock_net(sk), &ecn_ok))
8c763681	295	goto out;
bb5b7c11	296
1da177e4	297	ret = NULL;
ce4a7d0d	298	req = inet_reqsk_alloc(&tcp_request_sock_ops); /* for safety */
1da177e4 LT	299	if (!req)
	300	goto out;
	301
2e6599cb ACM	302	ireq = inet_rsk(req);
2e6599cb ACM	303	treq = tcp_rsk(req);
aa8223c7	304	treq->rcv_isn = ntohl(th->seq) - 1;
e905a9ed	305	treq->snt_isn = cookie;
1da177e4	306	req->mss = mss;
b44084c2 ED	307	ireq->ir_num = ntohs(th->dest);
	308	ireq->ir_rmt_port = th->source;
	309	ireq->ir_loc_addr = ip_hdr(skb)->daddr;
	310	ireq->ir_rmt_addr = ip_hdr(skb)->saddr;
172d69e6	311	ireq->ecn_ok = ecn_ok;
bb5b7c11	312	ireq->snd_wscale = tcp_opt.snd_wscale;
bb5b7c11 DM	313	ireq->sack_ok = tcp_opt.sack_ok;
	314	ireq->wscale_ok = tcp_opt.wscale_ok;
	315	ireq->tstamp_ok = tcp_opt.saw_tstamp;
	316	req->ts_recent = tcp_opt.saw_tstamp ? tcp_opt.rcv_tsval : 0;
9ad7c049	317	treq->snt_synack = tcp_opt.saw_tstamp ? tcp_opt.rcv_tsecr : 0;
8336886f	318	treq->listener = NULL;
1da177e4 LT	319
	320	/* We throwed the options of the initial SYN away, so we hope
	321	* the ACK carries the same options again (see RFC1122 4.2.3.8)
	322	*/
	323	if (opt && opt->optlen) {
f6d8bd05	324	int opt_size = sizeof(struct ip_options_rcu) + opt->optlen;
1da177e4	325
2e6599cb	326	ireq->opt = kmalloc(opt_size, GFP_ATOMIC);
f6d8bd05	327	if (ireq->opt != NULL && ip_options_echo(&ireq->opt->opt, skb)) {
2e6599cb ACM	328	kfree(ireq->opt);
2e6599cb ACM	329	ireq->opt = NULL;
1da177e4 LT	330	}
	331	}
	332
284904aa PM	333	if (security_inet_conn_request(sk, skb, req)) {
	334	reqsk_free(req);
	335	goto out;
	336	}
	337
e905a9ed	338	req->expires = 0UL;
e6c022a4	339	req->num_retrans = 0;
e905a9ed	340
1da177e4 LT	341	/*
	342	* We need to lookup the route here to get at the correct
	343	* window size. We should better make sure that the window size
	344	* hasn't changed since we received the original syn, but I see
e905a9ed	345	* no easy way to do this.
1da177e4	346	*/
d66954a0 DP	347	flowi4_init_output(&fl4, sk->sk_bound_dev_if, sk->sk_mark,
d66954a0 DP	348	RT_CONN_FLAGS(sk), RT_SCOPE_UNIVERSE, IPPROTO_TCP,
dfd25fff	349	inet_sk_flowi_flags(sk),
634fb979 ED	350	(opt && opt->srr) ? opt->faddr : ireq->ir_rmt_addr,
634fb979 ED	351	ireq->ir_loc_addr, th->source, th->dest);
dfd25fff ED	352	security_req_classify_flow(req, flowi4_to_flowi(&fl4));
	353	rt = ip_route_output_key(sock_net(sk), &fl4);
	354	if (IS_ERR(rt)) {
	355	reqsk_free(req);
	356	goto out;
1da177e4 LT	357	}
	358
	359	/* Try to redo what tcp_v4_send_synack did. */
d8d1f30b	360	req->window_clamp = tp->window_clamp ? :dst_metric(&rt->dst, RTAX_WINDOW);
4dfc2817	361
1da177e4	362	tcp_select_initial_window(tcp_full_space(sk), req->mss,
e905a9ed	363	&req->rcv_wnd, &req->window_clamp,
31d12926	364	ireq->wscale_ok, &rcv_wscale,
d8d1f30b	365	dst_metric(&rt->dst, RTAX_INITRWND));
4dfc2817	366
e905a9ed	367	ireq->rcv_wscale = rcv_wscale;
1da177e4	368
d8d1f30b	369	ret = get_cookie_sock(sk, skb, req, &rt->dst);
dfd25fff ED	370	/* ip_queue_xmit() depends on our flow being setup
	371	* Normal sockets get it right from inet_csk_route_child_sock()
	372	*/
	373	if (ret)
	374	inet_sk(ret)->cork.fl.u.ip4 = fl4;
1da177e4 LT	375	out: return ret;
1da177e4 LT	376	}