tcp: randomize timestamps on syncookies

[ Upstream commit 84b114b9 ]

Whole point of randomization was to hide server uptime, but an attacker
can simply start a syn flood and TCP generates 'old style' timestamps,
directly revealing server jiffies value.

Also, TSval sent by the server to a particular remote address vary
depending on syncookies being sent or not, potentially triggering PAWS
drops for innocent clients.

Lets implement proper randomization, including for SYNcookies.

Also we do not need to export sysctl_tcp_timestamps, since it is not
used from a module.

In v2, I added Florian feedback and contribution, adding tsoff to
tcp_get_cookie_sock().

v3 removed one unused variable in tcp_v4_connect() as Florian spotted.

Fixes: 95a22cae ("tcp: randomize tcp timestamp offsets for each connection")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Florian Westphal <fw@strlen.de>
Tested-by: Florian Westphal <fw@strlen.de>
Cc: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

include/net/secure_seq.h

@@ -6,10 +6,12 @@
 u32 secure_ipv4_port_ephemeral(__be32 saddr, __be32 daddr, __be16 dport);
 u32 secure_ipv6_port_ephemeral(const __be32 *saddr, const __be32 *daddr,
 			       __be16 dport);
-u32 secure_tcp_sequence_number(__be32 saddr, __be32 daddr,
-			       __be16 sport, __be16 dport, u32 *tsoff);
-u32 secure_tcpv6_sequence_number(const __be32 *saddr, const __be32 *daddr,
-				 __be16 sport, __be16 dport, u32 *tsoff);
+u32 secure_tcp_seq(__be32 saddr, __be32 daddr,
+		   __be16 sport, __be16 dport);
+u32 secure_tcp_ts_off(__be32 saddr, __be32 daddr);
+u32 secure_tcpv6_seq(const __be32 *saddr, const __be32 *daddr,
+		     __be16 sport, __be16 dport);
+u32 secure_tcpv6_ts_off(const __be32 *saddr, const __be32 *daddr);
 u64 secure_dccp_sequence_number(__be32 saddr, __be32 daddr,
 				__be16 sport, __be16 dport);
 u64 secure_dccpv6_sequence_number(__be32 *saddr, __be32 *daddr,

include/net/tcp.h

@@ -471,7 +471,7 @@ void inet_sk_rx_dst_set(struct sock *sk, const struct sk_buff *skb);
 /* From syncookies.c */
 struct sock *tcp_get_cookie_sock(struct sock *sk, struct sk_buff *skb,
 				 struct request_sock *req,
-				 struct dst_entry *dst);
+				 struct dst_entry *dst, u32 tsoff);
 int __cookie_v4_check(const struct iphdr *iph, const struct tcphdr *th,
 		      u32 cookie);
 struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb);
@@ -1816,7 +1816,8 @@ struct tcp_request_sock_ops {
 	struct dst_entry *(*route_req)(const struct sock *sk, struct flowi *fl,
 				       const struct request_sock *req,
 				       bool *strict);
-	__u32 (*init_seq)(const struct sk_buff *skb, u32 *tsoff);
+	u32 (*init_seq)(const struct sk_buff *skb);
+	u32 (*init_ts_off)(const struct sk_buff *skb);
 	int (*send_synack)(const struct sock *sk, struct dst_entry *dst,
 			   struct flowi *fl, struct request_sock *req,
 			   struct tcp_fastopen_cookie *foc,

net/core/secure_seq.c

@@ -24,9 +24,13 @@ static siphash_key_t ts_secret __read_mostly;
 
 static __always_inline void net_secret_init(void)
 {
-	net_get_random_once(&ts_secret, sizeof(ts_secret));
 	net_get_random_once(&net_secret, sizeof(net_secret));
 }
+
+static __always_inline void ts_secret_init(void)
+{
+	net_get_random_once(&ts_secret, sizeof(ts_secret));
+}
 #endif
 
 #ifdef CONFIG_INET
@@ -47,7 +51,7 @@ static u32 seq_scale(u32 seq)
 #endif
 
 #if IS_ENABLED(CONFIG_IPV6)
-static u32 secure_tcpv6_ts_off(const __be32 *saddr, const __be32 *daddr)
+u32 secure_tcpv6_ts_off(const __be32 *saddr, const __be32 *daddr)
 {
 	const struct {
 		struct in6_addr saddr;
@@ -60,12 +64,14 @@ static u32 secure_tcpv6_ts_off(const __be32 *saddr, const __be32 *daddr)
 	if (sysctl_tcp_timestamps != 1)
 		return 0;
 
+	ts_secret_init();
 	return siphash(&combined, offsetofend(typeof(combined), daddr),
 		       &ts_secret);
 }
+EXPORT_SYMBOL(secure_tcpv6_ts_off);
 
-u32 secure_tcpv6_sequence_number(const __be32 *saddr, const __be32 *daddr,
-				 __be16 sport, __be16 dport, u32 *tsoff)
+u32 secure_tcpv6_seq(const __be32 *saddr, const __be32 *daddr,
+		     __be16 sport, __be16 dport)
 {
 	const struct {
 		struct in6_addr saddr;
@@ -78,14 +84,14 @@ u32 secure_tcpv6_sequence_number(const __be32 *saddr, const __be32 *daddr,
 		.sport = sport,
 		.dport = dport
 	};
-	u64 hash;
+	u32 hash;
+
 	net_secret_init();
 	hash = siphash(&combined, offsetofend(typeof(combined), dport),
 		       &net_secret);
-	*tsoff = secure_tcpv6_ts_off(saddr, daddr);
 	return seq_scale(hash);
 }
-EXPORT_SYMBOL(secure_tcpv6_sequence_number);
+EXPORT_SYMBOL(secure_tcpv6_seq);
 
 u32 secure_ipv6_port_ephemeral(const __be32 *saddr, const __be32 *daddr,
 			       __be16 dport)
@@ -107,30 +113,30 @@ EXPORT_SYMBOL(secure_ipv6_port_ephemeral);
 #endif
 
 #ifdef CONFIG_INET
-static u32 secure_tcp_ts_off(__be32 saddr, __be32 daddr)
+u32 secure_tcp_ts_off(__be32 saddr, __be32 daddr)
 {
 	if (sysctl_tcp_timestamps != 1)
 		return 0;
 
+	ts_secret_init();
 	return siphash_2u32((__force u32)saddr, (__force u32)daddr,
 			    &ts_secret);
 }
 
-/* secure_tcp_sequence_number(a, b, 0, d) == secure_ipv4_port_ephemeral(a, b, d),
+/* secure_tcp_seq_and_tsoff(a, b, 0, d) == secure_ipv4_port_ephemeral(a, b, d),
  * but fortunately, `sport' cannot be 0 in any circumstances. If this changes,
  * it would be easy enough to have the former function use siphash_4u32, passing
  * the arguments as separate u32.
  */
-
-u32 secure_tcp_sequence_number(__be32 saddr, __be32 daddr,
-			       __be16 sport, __be16 dport, u32 *tsoff)
+u32 secure_tcp_seq(__be32 saddr, __be32 daddr,
+		   __be16 sport, __be16 dport)
 {
-	u64 hash;
+	u32 hash;
+
 	net_secret_init();
 	hash = siphash_3u32((__force u32)saddr, (__force u32)daddr,
 			    (__force u32)sport << 16 | (__force u32)dport,
 			    &net_secret);
-	*tsoff = secure_tcp_ts_off(saddr, daddr);
 	return seq_scale(hash);
 }
 

net/ipv4/syncookies.c

@@ -16,6 +16,7 @@
 #include <linux/siphash.h>
 #include <linux/kernel.h>
 #include <linux/export.h>
+#include <net/secure_seq.h>
 #include <net/tcp.h>
 #include <net/route.h>
 
@@ -203,7 +204,7 @@ EXPORT_SYMBOL_GPL(__cookie_v4_check);
 
 struct sock *tcp_get_cookie_sock(struct sock *sk, struct sk_buff *skb,
 				 struct request_sock *req,
-				 struct dst_entry *dst)
+				 struct dst_entry *dst, u32 tsoff)
 {
 	struct inet_connection_sock *icsk = inet_csk(sk);
 	struct sock *child;
@@ -213,6 +214,7 @@ struct sock *tcp_get_cookie_sock(struct sock *sk, struct sk_buff *skb,
 						 NULL, &own_req);
 	if (child) {
 		atomic_set(&req->rsk_refcnt, 1);
+		tcp_sk(child)->tsoffset = tsoff;
 		sock_rps_save_rxhash(child, skb);
 		inet_csk_reqsk_queue_add(sk, req, child);
 	} else {
@@ -292,6 +294,7 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb)
 	struct rtable *rt;
 	__u8 rcv_wscale;
 	struct flowi4 fl4;
+	u32 tsoff = 0;
 
 	if (!sock_net(sk)->ipv4.sysctl_tcp_syncookies || !th->ack || th->rst)
 		goto out;
@@ -311,6 +314,11 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb)
 	memset(&tcp_opt, 0, sizeof(tcp_opt));
 	tcp_parse_options(skb, &tcp_opt, 0, NULL);
 
+	if (tcp_opt.saw_tstamp && tcp_opt.rcv_tsecr) {
+		tsoff = secure_tcp_ts_off(ip_hdr(skb)->daddr, ip_hdr(skb)->saddr);
+		tcp_opt.rcv_tsecr -= tsoff;
+	}
+
 	if (!cookie_timestamp_decode(&tcp_opt))
 		goto out;
 
@@ -381,7 +389,7 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb)
 	ireq->rcv_wscale  = rcv_wscale;
 	ireq->ecn_ok = cookie_ecn_ok(&tcp_opt, sock_net(sk), &rt->dst);
 
-	ret = tcp_get_cookie_sock(sk, skb, req, &rt->dst);
+	ret = tcp_get_cookie_sock(sk, skb, req, &rt->dst, tsoff);
 	/* ip_queue_xmit() depends on our flow being setup
 	 * Normal sockets get it right from inet_csk_route_child_sock()
 	 */

net/ipv4/tcp_input.c

@@ -85,7 +85,6 @@ int sysctl_tcp_dsack __read_mostly = 1;
 int sysctl_tcp_app_win __read_mostly = 31;
 int sysctl_tcp_adv_win_scale __read_mostly = 1;
 EXPORT_SYMBOL(sysctl_tcp_adv_win_scale);
-EXPORT_SYMBOL(sysctl_tcp_timestamps);
 
 /* rfc5961 challenge ack rate limiting */
 int sysctl_tcp_challenge_ack_limit = 1000;
@@ -6332,8 +6331,8 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops,
 	if (security_inet_conn_request(sk, skb, req))
 		goto drop_and_free;
 
-	if (isn && tmp_opt.tstamp_ok)
-		af_ops->init_seq(skb, &tcp_rsk(req)->ts_off);
+	if (tmp_opt.tstamp_ok)
+		tcp_rsk(req)->ts_off = af_ops->init_ts_off(skb);
 
 	if (!want_cookie && !isn) {
 		/* VJ's idea. We save last timestamp seen
@@ -6375,7 +6374,7 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops,
 			goto drop_and_release;
 		}
 
-		isn = af_ops->init_seq(skb, &tcp_rsk(req)->ts_off);
+		isn = af_ops->init_seq(skb);
 	}
 	if (!dst) {
 		dst = af_ops->route_req(sk, &fl, req, NULL);
@@ -6387,7 +6386,6 @@ int tcp_conn_request(struct request_sock_ops *rsk_ops,
 
 	if (want_cookie) {
 		isn = cookie_init_sequence(af_ops, sk, skb, &req->mss);
-		tcp_rsk(req)->ts_off = 0;
 		req->cookie_ts = tmp_opt.tstamp_ok;
 		if (!tmp_opt.tstamp_ok)
 			inet_rsk(req)->ecn_ok = 0;

net/ipv4/tcp_ipv4.c

@@ -94,12 +94,18 @@ static int tcp_v4_md5_hash_hdr(char *md5_hash, const struct tcp_md5sig_key *key,
 struct inet_hashinfo tcp_hashinfo;
 EXPORT_SYMBOL(tcp_hashinfo);
 
-static u32 tcp_v4_init_sequence(const struct sk_buff *skb, u32 *tsoff)
+static u32 tcp_v4_init_seq(const struct sk_buff *skb)
 {
-	return secure_tcp_sequence_number(ip_hdr(skb)->daddr,
-					  ip_hdr(skb)->saddr,
-					  tcp_hdr(skb)->dest,
-					  tcp_hdr(skb)->source, tsoff);
+	return secure_tcp_seq(ip_hdr(skb)->daddr,
+			      ip_hdr(skb)->saddr,
+			      tcp_hdr(skb)->dest,
+			      tcp_hdr(skb)->source);
+}
+
+static u32 tcp_v4_init_ts_off(const struct sk_buff *skb)
+{
+	return secure_tcp_ts_off(ip_hdr(skb)->daddr,
+				 ip_hdr(skb)->saddr);
 }
 
 int tcp_twsk_unique(struct sock *sk, struct sock *sktw, void *twp)
@@ -145,7 +151,6 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
 	struct flowi4 *fl4;
 	struct rtable *rt;
 	int err;
-	u32 seq;
 	struct ip_options_rcu *inet_opt;
 	struct inet_timewait_death_row *tcp_death_row = &sock_net(sk)->ipv4.tcp_death_row;
 
@@ -236,13 +241,13 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len)
 	rt = NULL;
 
 	if (likely(!tp->repair)) {
-		seq = secure_tcp_sequence_number(inet->inet_saddr,
-						 inet->inet_daddr,
-						 inet->inet_sport,
-						 usin->sin_port,
-						 &tp->tsoffset);
 		if (!tp->write_seq)
-			tp->write_seq = seq;
+			tp->write_seq = secure_tcp_seq(inet->inet_saddr,
+						       inet->inet_daddr,
+						       inet->inet_sport,
+						       usin->sin_port);
+		tp->tsoffset = secure_tcp_ts_off(inet->inet_saddr,
+						 inet->inet_daddr);
 	}
 
 	inet->inet_id = tp->write_seq ^ jiffies;
@@ -1253,7 +1258,8 @@ static const struct tcp_request_sock_ops tcp_request_sock_ipv4_ops = {
 	.cookie_init_seq =	cookie_v4_init_sequence,
 #endif
 	.route_req	=	tcp_v4_route_req,
-	.init_seq	=	tcp_v4_init_sequence,
+	.init_seq	=	tcp_v4_init_seq,
+	.init_ts_off	=	tcp_v4_init_ts_off,
 	.send_synack	=	tcp_v4_send_synack,
 };
 

net/ipv6/syncookies.c

@@ -18,6 +18,7 @@
 #include <linux/random.h>
 #include <linux/siphash.h>
 #include <linux/kernel.h>
+#include <net/secure_seq.h>
 #include <net/ipv6.h>
 #include <net/tcp.h>
 
@@ -143,6 +144,7 @@ struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb)
 	int mss;
 	struct dst_entry *dst;
 	__u8 rcv_wscale;
+	u32 tsoff = 0;
 
 	if (!sock_net(sk)->ipv4.sysctl_tcp_syncookies || !th->ack || th->rst)
 		goto out;
@@ -162,6 +164,12 @@ struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb)
 	memset(&tcp_opt, 0, sizeof(tcp_opt));
 	tcp_parse_options(skb, &tcp_opt, 0, NULL);
 
+	if (tcp_opt.saw_tstamp && tcp_opt.rcv_tsecr) {
+		tsoff = secure_tcpv6_ts_off(ipv6_hdr(skb)->daddr.s6_addr32,
+					    ipv6_hdr(skb)->saddr.s6_addr32);
+		tcp_opt.rcv_tsecr -= tsoff;
+	}
+
 	if (!cookie_timestamp_decode(&tcp_opt))
 		goto out;
 
@@ -242,7 +250,7 @@ struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb)
 	ireq->rcv_wscale = rcv_wscale;
 	ireq->ecn_ok = cookie_ecn_ok(&tcp_opt, sock_net(sk), dst);
 
-	ret = tcp_get_cookie_sock(sk, skb, req, dst);
+	ret = tcp_get_cookie_sock(sk, skb, req, dst, tsoff);
 out:
 	return ret;
 out_free:

net/ipv6/tcp_ipv6.c

@@ -101,12 +101,18 @@ static void inet6_sk_rx_dst_set(struct sock *sk, const struct sk_buff *skb)
 	}
 }
 
-static u32 tcp_v6_init_sequence(const struct sk_buff *skb, u32 *tsoff)
+static u32 tcp_v6_init_seq(const struct sk_buff *skb)
 {
-	return secure_tcpv6_sequence_number(ipv6_hdr(skb)->daddr.s6_addr32,
-					    ipv6_hdr(skb)->saddr.s6_addr32,
-					    tcp_hdr(skb)->dest,
-					    tcp_hdr(skb)->source, tsoff);
+	return secure_tcpv6_seq(ipv6_hdr(skb)->daddr.s6_addr32,
+				ipv6_hdr(skb)->saddr.s6_addr32,
+				tcp_hdr(skb)->dest,
+				tcp_hdr(skb)->source);
+}
+
+static u32 tcp_v6_init_ts_off(const struct sk_buff *skb)
+{
+	return secure_tcpv6_ts_off(ipv6_hdr(skb)->daddr.s6_addr32,
+				   ipv6_hdr(skb)->saddr.s6_addr32);
 }
 
 static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
@@ -122,7 +128,6 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 	struct flowi6 fl6;
 	struct dst_entry *dst;
 	int addr_type;
-	u32 seq;
 	int err;
 	struct inet_timewait_death_row *tcp_death_row = &sock_net(sk)->ipv4.tcp_death_row;
 
@@ -287,13 +292,13 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr,
 	sk_set_txhash(sk);
 
 	if (likely(!tp->repair)) {
-		seq = secure_tcpv6_sequence_number(np->saddr.s6_addr32,
-						   sk->sk_v6_daddr.s6_addr32,
-						   inet->inet_sport,
-						   inet->inet_dport,
-						   &tp->tsoffset);
 		if (!tp->write_seq)
-			tp->write_seq = seq;
+			tp->write_seq = secure_tcpv6_seq(np->saddr.s6_addr32,
+							 sk->sk_v6_daddr.s6_addr32,
+							 inet->inet_sport,
+							 inet->inet_dport);
+		tp->tsoffset = secure_tcpv6_ts_off(np->saddr.s6_addr32,
+						   sk->sk_v6_daddr.s6_addr32);
 	}
 
 	if (tcp_fastopen_defer_connect(sk, &err))
@@ -757,7 +762,8 @@ static const struct tcp_request_sock_ops tcp_request_sock_ipv6_ops = {
 	.cookie_init_seq =	cookie_v6_init_sequence,
 #endif
 	.route_req	=	tcp_v6_route_req,
-	.init_seq	=	tcp_v6_init_sequence,
+	.init_seq	=	tcp_v6_init_seq,
+	.init_ts_off	=	tcp_v6_init_ts_off,
 	.send_synack	=	tcp_v6_send_synack,
 };