1 files changed, 412 insertions, 447 deletions

diff --git a/net/ipv4/tcp_metrics.c b/net/ipv4/tcp_metrics.c
index f6a005c485a9..45b6ecd16412 100644
--- a/net/ipv4/tcp_metrics.c
+++ b/net/ipv4/tcp_metrics.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0
 #include <linux/rcupdate.h>
 #include <linux/spinlock.h>
 #include <linux/jiffies.h>
@@ -20,88 +21,88 @@
 #include <net/tcp.h>
 #include <net/genetlink.h>
 
-int sysctl_tcp_nometrics_save __read_mostly;
+static struct tcp_metrics_block *__tcp_get_metrics(const struct inetpeer_addr *saddr,
+						   const struct inetpeer_addr *daddr,
+						   struct net *net, unsigned int hash);
 
 struct tcp_fastopen_metrics {
 	u16	mss;
-	u16	syn_loss:10;		/* Recurring Fast Open SYN losses */
+	u16	syn_loss:10,		/* Recurring Fast Open SYN losses */
+		try_exp:2;		/* Request w/ exp. option (once) */
 	unsigned long	last_syn_loss;	/* Last Fast Open SYN loss */
 	struct	tcp_fastopen_cookie	cookie;
 };
 
+/* TCP_METRIC_MAX includes 2 extra fields for userspace compatibility
+ * Kernel only stores RTT and RTTVAR in usec resolution
+ */
+#define TCP_METRIC_MAX_KERNEL (TCP_METRIC_MAX - 2)
+
 struct tcp_metrics_block {
 	struct tcp_metrics_block __rcu	*tcpm_next;
-	struct inetpeer_addr		tcpm_addr;
+	struct net			*tcpm_net;
+	struct inetpeer_addr		tcpm_saddr;
+	struct inetpeer_addr		tcpm_daddr;
 	unsigned long			tcpm_stamp;
-	u32				tcpm_ts;
-	u32				tcpm_ts_stamp;
 	u32				tcpm_lock;
-	u32				tcpm_vals[TCP_METRIC_MAX + 1];
+	u32				tcpm_vals[TCP_METRIC_MAX_KERNEL + 1];
 	struct tcp_fastopen_metrics	tcpm_fastopen;
 
 	struct rcu_head			rcu_head;
 };
 
-static bool tcp_metric_locked(struct tcp_metrics_block *tm,
-			      enum tcp_metric_index idx)
+static inline struct net *tm_net(const struct tcp_metrics_block *tm)
 {
-	return tm->tcpm_lock & (1 << idx);
+	/* Paired with the WRITE_ONCE() in tcpm_new() */
+	return READ_ONCE(tm->tcpm_net);
 }
 
-static u32 tcp_metric_get(struct tcp_metrics_block *tm,
-			  enum tcp_metric_index idx)
+static bool tcp_metric_locked(struct tcp_metrics_block *tm,
+			      enum tcp_metric_index idx)
 {
-	return tm->tcpm_vals[idx];
+	/* Paired with WRITE_ONCE() in tcpm_suck_dst() */
+	return READ_ONCE(tm->tcpm_lock) & (1 << idx);
 }
 
-static u32 tcp_metric_get_jiffies(struct tcp_metrics_block *tm,
-				  enum tcp_metric_index idx)
+static u32 tcp_metric_get(const struct tcp_metrics_block *tm,
+			  enum tcp_metric_index idx)
 {
-	return msecs_to_jiffies(tm->tcpm_vals[idx]);
+	/* Paired with WRITE_ONCE() in tcp_metric_set() */
+	return READ_ONCE(tm->tcpm_vals[idx]);
 }
 
 static void tcp_metric_set(struct tcp_metrics_block *tm,
 			   enum tcp_metric_index idx,
 			   u32 val)
 {
-	tm->tcpm_vals[idx] = val;
-}
-
-static void tcp_metric_set_msecs(struct tcp_metrics_block *tm,
-				 enum tcp_metric_index idx,
-				 u32 val)
-{
-	tm->tcpm_vals[idx] = jiffies_to_msecs(val);
+	/* Paired with READ_ONCE() in tcp_metric_get() */
+	WRITE_ONCE(tm->tcpm_vals[idx], val);
 }
 
 static bool addr_same(const struct inetpeer_addr *a,
 		      const struct inetpeer_addr *b)
 {
-	const struct in6_addr *a6, *b6;
-
-	if (a->family != b->family)
-		return false;
-	if (a->family == AF_INET)
-		return a->addr.a4 == b->addr.a4;
-
-	a6 = (const struct in6_addr *) &a->addr.a6[0];
-	b6 = (const struct in6_addr *) &b->addr.a6[0];
-
-	return ipv6_addr_equal(a6, b6);
+	return (a->family == b->family) && !inetpeer_addr_cmp(a, b);
 }
 
 struct tcpm_hash_bucket {
 	struct tcp_metrics_block __rcu	*chain;
 };
 
+static struct tcpm_hash_bucket	*tcp_metrics_hash __read_mostly;
+static unsigned int		tcp_metrics_hash_log __read_mostly;
+
 static DEFINE_SPINLOCK(tcp_metrics_lock);
+static DEFINE_SEQLOCK(fastopen_seqlock);
 
-static void tcpm_suck_dst(struct tcp_metrics_block *tm, struct dst_entry *dst,
+static void tcpm_suck_dst(struct tcp_metrics_block *tm,
+			  const struct dst_entry *dst,
 			  bool fastopen_clear)
 {
+	u32 msval;
 	u32 val;
 
-	tm->tcpm_stamp = jiffies;
+	WRITE_ONCE(tm->tcpm_stamp, jiffies);
 
 	val = 0;
 	if (dst_metric_locked(dst, RTAX_RTT))
@@ -114,54 +115,103 @@ static void tcpm_suck_dst(struct tcp_metrics_block *tm, struct dst_entry *dst,
 		val |= 1 << TCP_METRIC_CWND;
 	if (dst_metric_locked(dst, RTAX_REORDERING))
 		val |= 1 << TCP_METRIC_REORDERING;
-	tm->tcpm_lock = val;
-
-	tm->tcpm_vals[TCP_METRIC_RTT] = dst_metric_raw(dst, RTAX_RTT);
-	tm->tcpm_vals[TCP_METRIC_RTTVAR] = dst_metric_raw(dst, RTAX_RTTVAR);
-	tm->tcpm_vals[TCP_METRIC_SSTHRESH] = dst_metric_raw(dst, RTAX_SSTHRESH);
-	tm->tcpm_vals[TCP_METRIC_CWND] = dst_metric_raw(dst, RTAX_CWND);
-	tm->tcpm_vals[TCP_METRIC_REORDERING] = dst_metric_raw(dst, RTAX_REORDERING);
-	tm->tcpm_ts = 0;
-	tm->tcpm_ts_stamp = 0;
+	/* Paired with READ_ONCE() in tcp_metric_locked() */
+	WRITE_ONCE(tm->tcpm_lock, val);
+
+	msval = dst_metric_raw(dst, RTAX_RTT);
+	tcp_metric_set(tm, TCP_METRIC_RTT, msval * USEC_PER_MSEC);
+
+	msval = dst_metric_raw(dst, RTAX_RTTVAR);
+	tcp_metric_set(tm, TCP_METRIC_RTTVAR, msval * USEC_PER_MSEC);
+	tcp_metric_set(tm, TCP_METRIC_SSTHRESH,
+		       dst_metric_raw(dst, RTAX_SSTHRESH));
+	tcp_metric_set(tm, TCP_METRIC_CWND,
+		       dst_metric_raw(dst, RTAX_CWND));
+	tcp_metric_set(tm, TCP_METRIC_REORDERING,
+		       dst_metric_raw(dst, RTAX_REORDERING));
 	if (fastopen_clear) {
+		write_seqlock(&fastopen_seqlock);
 		tm->tcpm_fastopen.mss = 0;
 		tm->tcpm_fastopen.syn_loss = 0;
+		tm->tcpm_fastopen.try_exp = 0;
+		tm->tcpm_fastopen.cookie.exp = false;
 		tm->tcpm_fastopen.cookie.len = 0;
+		write_sequnlock(&fastopen_seqlock);
 	}
 }
 
+#define TCP_METRICS_TIMEOUT		(60 * 60 * HZ)
+
+static void tcpm_check_stamp(struct tcp_metrics_block *tm,
+			     const struct dst_entry *dst)
+{
+	unsigned long limit;
+
+	if (!tm)
+		return;
+	limit = READ_ONCE(tm->tcpm_stamp) + TCP_METRICS_TIMEOUT;
+	if (unlikely(time_after(jiffies, limit)))
+		tcpm_suck_dst(tm, dst, false);
+}
+
+#define TCP_METRICS_RECLAIM_DEPTH	5
+#define TCP_METRICS_RECLAIM_PTR		(struct tcp_metrics_block *) 0x1UL
+
+#define deref_locked(p)	\
+	rcu_dereference_protected(p, lockdep_is_held(&tcp_metrics_lock))
+
 static struct tcp_metrics_block *tcpm_new(struct dst_entry *dst,
-					  struct inetpeer_addr *addr,
-					  unsigned int hash,
-					  bool reclaim)
+					  struct inetpeer_addr *saddr,
+					  struct inetpeer_addr *daddr,
+					  unsigned int hash)
 {
 	struct tcp_metrics_block *tm;
+	bool reclaim = false;
 	struct net *net;
 
 	spin_lock_bh(&tcp_metrics_lock);
-	net = dev_net(dst->dev);
+	net = dst_dev_net_rcu(dst);
+
+	/* While waiting for the spin-lock the cache might have been populated
+	 * with this entry and so we have to check again.
+	 */
+	tm = __tcp_get_metrics(saddr, daddr, net, hash);
+	if (tm == TCP_METRICS_RECLAIM_PTR) {
+		reclaim = true;
+		tm = NULL;
+	}
+	if (tm) {
+		tcpm_check_stamp(tm, dst);
+		goto out_unlock;
+	}
+
 	if (unlikely(reclaim)) {
 		struct tcp_metrics_block *oldest;
 
-		oldest = rcu_dereference(net->ipv4.tcp_metrics_hash[hash].chain);
-		for (tm = rcu_dereference(oldest->tcpm_next); tm;
-		     tm = rcu_dereference(tm->tcpm_next)) {
-			if (time_before(tm->tcpm_stamp, oldest->tcpm_stamp))
+		oldest = deref_locked(tcp_metrics_hash[hash].chain);
+		for (tm = deref_locked(oldest->tcpm_next); tm;
+		     tm = deref_locked(tm->tcpm_next)) {
+			if (time_before(READ_ONCE(tm->tcpm_stamp),
+					READ_ONCE(oldest->tcpm_stamp)))
 				oldest = tm;
 		}
 		tm = oldest;
 	} else {
-		tm = kmalloc(sizeof(*tm), GFP_ATOMIC);
+		tm = kzalloc(sizeof(*tm), GFP_ATOMIC);
 		if (!tm)
 			goto out_unlock;
 	}
-	tm->tcpm_addr = *addr;
+	/* Paired with the READ_ONCE() in tm_net() */
+	WRITE_ONCE(tm->tcpm_net, net);
+
+	tm->tcpm_saddr = *saddr;
+	tm->tcpm_daddr = *daddr;
 
-	tcpm_suck_dst(tm, dst, true);
+	tcpm_suck_dst(tm, dst, reclaim);
 
 	if (likely(!reclaim)) {
-		tm->tcpm_next = net->ipv4.tcp_metrics_hash[hash].chain;
-		rcu_assign_pointer(net->ipv4.tcp_metrics_hash[hash].chain, tm);
+		tm->tcpm_next = tcp_metrics_hash[hash].chain;
+		rcu_assign_pointer(tcp_metrics_hash[hash].chain, tm);
 	}
 
 out_unlock:
@@ -169,17 +219,6 @@ out_unlock:
 	return tm;
 }
 
-#define TCP_METRICS_TIMEOUT		(60 * 60 * HZ)
-
-static void tcpm_check_stamp(struct tcp_metrics_block *tm, struct dst_entry *dst)
-{
-	if (tm && unlikely(time_after(jiffies, tm->tcpm_stamp + TCP_METRICS_TIMEOUT)))
-		tcpm_suck_dst(tm, dst, false);
-}
-
-#define TCP_METRICS_RECLAIM_DEPTH	5
-#define TCP_METRICS_RECLAIM_PTR		(struct tcp_metrics_block *) 0x1UL
-
 static struct tcp_metrics_block *tcp_get_encode(struct tcp_metrics_block *tm, int depth)
 {
 	if (tm)
@@ -189,15 +228,18 @@ static struct tcp_metrics_block *tcp_get_encode(struct tcp_metrics_block *tm, in
 	return NULL;
 }
 
-static struct tcp_metrics_block *__tcp_get_metrics(const struct inetpeer_addr *addr,
+static struct tcp_metrics_block *__tcp_get_metrics(const struct inetpeer_addr *saddr,
+						   const struct inetpeer_addr *daddr,
 						   struct net *net, unsigned int hash)
 {
 	struct tcp_metrics_block *tm;
 	int depth = 0;
 
-	for (tm = rcu_dereference(net->ipv4.tcp_metrics_hash[hash].chain); tm;
+	for (tm = rcu_dereference(tcp_metrics_hash[hash].chain); tm;
 	     tm = rcu_dereference(tm->tcpm_next)) {
-		if (addr_same(&tm->tcpm_addr, addr))
+		if (addr_same(&tm->tcpm_saddr, saddr) &&
+		    addr_same(&tm->tcpm_daddr, daddr) &&
+		    net_eq(tm_net(tm), net))
 			break;
 		depth++;
 	}
@@ -208,105 +250,83 @@ static struct tcp_metrics_block *__tcp_get_metrics_req(struct request_sock *req,
 						       struct dst_entry *dst)
 {
 	struct tcp_metrics_block *tm;
-	struct inetpeer_addr addr;
+	struct inetpeer_addr saddr, daddr;
 	unsigned int hash;
 	struct net *net;
 
-	addr.family = req->rsk_ops->family;
-	switch (addr.family) {
+	saddr.family = req->rsk_ops->family;
+	daddr.family = req->rsk_ops->family;
+	switch (daddr.family) {
 	case AF_INET:
-		addr.addr.a4 = inet_rsk(req)->rmt_addr;
-		hash = (__force unsigned int) addr.addr.a4;
+		inetpeer_set_addr_v4(&saddr, inet_rsk(req)->ir_loc_addr);
+		inetpeer_set_addr_v4(&daddr, inet_rsk(req)->ir_rmt_addr);
+		hash = ipv4_addr_hash(inet_rsk(req)->ir_rmt_addr);
 		break;
+#if IS_ENABLED(CONFIG_IPV6)
 	case AF_INET6:
-		*(struct in6_addr *)addr.addr.a6 = inet6_rsk(req)->rmt_addr;
-		hash = ipv6_addr_hash(&inet6_rsk(req)->rmt_addr);
+		inetpeer_set_addr_v6(&saddr, &inet_rsk(req)->ir_v6_loc_addr);
+		inetpeer_set_addr_v6(&daddr, &inet_rsk(req)->ir_v6_rmt_addr);
+		hash = ipv6_addr_hash(&inet_rsk(req)->ir_v6_rmt_addr);
 		break;
+#endif
 	default:
 		return NULL;
 	}
 
-	net = dev_net(dst->dev);
-	hash = hash_32(hash, net->ipv4.tcp_metrics_hash_log);
+	net = dst_dev_net_rcu(dst);
+	hash ^= net_hash_mix(net);
+	hash = hash_32(hash, tcp_metrics_hash_log);
 
-	for (tm = rcu_dereference(net->ipv4.tcp_metrics_hash[hash].chain); tm;
+	for (tm = rcu_dereference(tcp_metrics_hash[hash].chain); tm;
 	     tm = rcu_dereference(tm->tcpm_next)) {
-		if (addr_same(&tm->tcpm_addr, &addr))
+		if (addr_same(&tm->tcpm_saddr, &saddr) &&
+		    addr_same(&tm->tcpm_daddr, &daddr) &&
+		    net_eq(tm_net(tm), net))
 			break;
 	}
 	tcpm_check_stamp(tm, dst);
 	return tm;
 }
 
-static struct tcp_metrics_block *__tcp_get_metrics_tw(struct inet_timewait_sock *tw)
-{
-	struct inet6_timewait_sock *tw6;
-	struct tcp_metrics_block *tm;
-	struct inetpeer_addr addr;
-	unsigned int hash;
-	struct net *net;
-
-	addr.family = tw->tw_family;
-	switch (addr.family) {
-	case AF_INET:
-		addr.addr.a4 = tw->tw_daddr;
-		hash = (__force unsigned int) addr.addr.a4;
-		break;
-	case AF_INET6:
-		tw6 = inet6_twsk((struct sock *)tw);
-		*(struct in6_addr *)addr.addr.a6 = tw6->tw_v6_daddr;
-		hash = ipv6_addr_hash(&tw6->tw_v6_daddr);
-		break;
-	default:
-		return NULL;
-	}
-
-	net = twsk_net(tw);
-	hash = hash_32(hash, net->ipv4.tcp_metrics_hash_log);
-
-	for (tm = rcu_dereference(net->ipv4.tcp_metrics_hash[hash].chain); tm;
-	     tm = rcu_dereference(tm->tcpm_next)) {
-		if (addr_same(&tm->tcpm_addr, &addr))
-			break;
-	}
-	return tm;
-}
-
 static struct tcp_metrics_block *tcp_get_metrics(struct sock *sk,
 						 struct dst_entry *dst,
 						 bool create)
 {
 	struct tcp_metrics_block *tm;
-	struct inetpeer_addr addr;
+	struct inetpeer_addr saddr, daddr;
 	unsigned int hash;
 	struct net *net;
-	bool reclaim;
 
-	addr.family = sk->sk_family;
-	switch (addr.family) {
-	case AF_INET:
-		addr.addr.a4 = inet_sk(sk)->inet_daddr;
-		hash = (__force unsigned int) addr.addr.a4;
-		break;
-	case AF_INET6:
-		*(struct in6_addr *)addr.addr.a6 = inet6_sk(sk)->daddr;
-		hash = ipv6_addr_hash(&inet6_sk(sk)->daddr);
-		break;
-	default:
-		return NULL;
+	if (sk->sk_family == AF_INET) {
+		inetpeer_set_addr_v4(&saddr, inet_sk(sk)->inet_saddr);
+		inetpeer_set_addr_v4(&daddr, inet_sk(sk)->inet_daddr);
+		hash = ipv4_addr_hash(inet_sk(sk)->inet_daddr);
+	}
+#if IS_ENABLED(CONFIG_IPV6)
+	else if (sk->sk_family == AF_INET6) {
+		if (ipv6_addr_v4mapped(&sk->sk_v6_daddr)) {
+			inetpeer_set_addr_v4(&saddr, inet_sk(sk)->inet_saddr);
+			inetpeer_set_addr_v4(&daddr, inet_sk(sk)->inet_daddr);
+			hash = ipv4_addr_hash(inet_sk(sk)->inet_daddr);
+		} else {
+			inetpeer_set_addr_v6(&saddr, &sk->sk_v6_rcv_saddr);
+			inetpeer_set_addr_v6(&daddr, &sk->sk_v6_daddr);
+			hash = ipv6_addr_hash(&sk->sk_v6_daddr);
+		}
 	}
+#endif
+	else
+		return NULL;
 
-	net = dev_net(dst->dev);
-	hash = hash_32(hash, net->ipv4.tcp_metrics_hash_log);
+	net = dst_dev_net_rcu(dst);
+	hash ^= net_hash_mix(net);
+	hash = hash_32(hash, tcp_metrics_hash_log);
 
-	tm = __tcp_get_metrics(&addr, net, hash);
-	reclaim = false;
-	if (tm == TCP_METRICS_RECLAIM_PTR) {
-		reclaim = true;
+	tm = __tcp_get_metrics(&saddr, &daddr, net, hash);
+	if (tm == TCP_METRICS_RECLAIM_PTR)
 		tm = NULL;
-	}
 	if (!tm && create)
-		tm = tcpm_new(dst, &addr, hash, reclaim);
+		tm = tcpm_new(dst, &saddr, &daddr, hash);
 	else
 		tcpm_check_stamp(tm, dst);
 
@@ -322,19 +342,18 @@ void tcp_update_metrics(struct sock *sk)
 	const struct inet_connection_sock *icsk = inet_csk(sk);
 	struct dst_entry *dst = __sk_dst_get(sk);
 	struct tcp_sock *tp = tcp_sk(sk);
+	struct net *net = sock_net(sk);
 	struct tcp_metrics_block *tm;
 	unsigned long rtt;
 	u32 val;
 	int m;
 
-	if (sysctl_tcp_nometrics_save || !dst)
+	sk_dst_confirm(sk);
+	if (READ_ONCE(net->ipv4.sysctl_tcp_nometrics_save) || !dst)
 		return;
 
-	if (dst->flags & DST_HOST)
-		dst_confirm(dst);
-
 	rcu_read_lock();
-	if (icsk->icsk_backoff || !tp->srtt) {
+	if (icsk->icsk_backoff || !tp->srtt_us) {
 		/* This session failed to estimate rtt. Why?
 		 * Probably, no packets returned in time.  Reset our
 		 * results.
@@ -349,8 +368,8 @@ void tcp_update_metrics(struct sock *sk)
 	if (!tm)
 		goto out_unlock;
 
-	rtt = tcp_metric_get_jiffies(tm, TCP_METRIC_RTT);
-	m = rtt - tp->srtt;
+	rtt = tcp_metric_get(tm, TCP_METRIC_RTT);
+	m = rtt - tp->srtt_us;
 
 	/* If newly calculated rtt larger than stored one, store new
 	 * one. Otherwise, use EWMA. Remember, rtt overestimation is
@@ -358,10 +377,10 @@ void tcp_update_metrics(struct sock *sk)
 	 */
 	if (!tcp_metric_locked(tm, TCP_METRIC_RTT)) {
 		if (m <= 0)
-			rtt = tp->srtt;
+			rtt = tp->srtt_us;
 		else
 			rtt -= (m >> 3);
-		tcp_metric_set_msecs(tm, TCP_METRIC_RTT, rtt);
+		tcp_metric_set(tm, TCP_METRIC_RTT, rtt);
 	}
 
 	if (!tcp_metric_locked(tm, TCP_METRIC_RTTVAR)) {
@@ -372,41 +391,43 @@ void tcp_update_metrics(struct sock *sk)
 
 		/* Scale deviation to rttvar fixed point */
 		m >>= 1;
-		if (m < tp->mdev)
-			m = tp->mdev;
+		if (m < tp->mdev_us)
+			m = tp->mdev_us;
 
-		var = tcp_metric_get_jiffies(tm, TCP_METRIC_RTTVAR);
+		var = tcp_metric_get(tm, TCP_METRIC_RTTVAR);
 		if (m >= var)
 			var = m;
 		else
 			var -= (var - m) >> 2;
 
-		tcp_metric_set_msecs(tm, TCP_METRIC_RTTVAR, var);
+		tcp_metric_set(tm, TCP_METRIC_RTTVAR, var);
 	}
 
 	if (tcp_in_initial_slowstart(tp)) {
 		/* Slow start still did not finish. */
-		if (!tcp_metric_locked(tm, TCP_METRIC_SSTHRESH)) {
+		if (!READ_ONCE(net->ipv4.sysctl_tcp_no_ssthresh_metrics_save) &&
+		    !tcp_metric_locked(tm, TCP_METRIC_SSTHRESH)) {
 			val = tcp_metric_get(tm, TCP_METRIC_SSTHRESH);
-			if (val && (tp->snd_cwnd >> 1) > val)
+			if (val && (tcp_snd_cwnd(tp) >> 1) > val)
 				tcp_metric_set(tm, TCP_METRIC_SSTHRESH,
-					       tp->snd_cwnd >> 1);
+					       tcp_snd_cwnd(tp) >> 1);
 		}
 		if (!tcp_metric_locked(tm, TCP_METRIC_CWND)) {
 			val = tcp_metric_get(tm, TCP_METRIC_CWND);
-			if (tp->snd_cwnd > val)
+			if (tcp_snd_cwnd(tp) > val)
 				tcp_metric_set(tm, TCP_METRIC_CWND,
-					       tp->snd_cwnd);
+					       tcp_snd_cwnd(tp));
 		}
-	} else if (tp->snd_cwnd > tp->snd_ssthresh &&
+	} else if (!tcp_in_slow_start(tp) &&
 		   icsk->icsk_ca_state == TCP_CA_Open) {
 		/* Cong. avoidance phase, cwnd is reliable. */
-		if (!tcp_metric_locked(tm, TCP_METRIC_SSTHRESH))
+		if (!READ_ONCE(net->ipv4.sysctl_tcp_no_ssthresh_metrics_save) &&
+		    !tcp_metric_locked(tm, TCP_METRIC_SSTHRESH))
 			tcp_metric_set(tm, TCP_METRIC_SSTHRESH,
-				       max(tp->snd_cwnd >> 1, tp->snd_ssthresh));
+				       max(tcp_snd_cwnd(tp) >> 1, tp->snd_ssthresh));
 		if (!tcp_metric_locked(tm, TCP_METRIC_CWND)) {
 			val = tcp_metric_get(tm, TCP_METRIC_CWND);
-			tcp_metric_set(tm, TCP_METRIC_CWND, (val + tp->snd_cwnd) >> 1);
+			tcp_metric_set(tm, TCP_METRIC_CWND, (val + tcp_snd_cwnd(tp)) >> 1);
 		}
 	} else {
 		/* Else slow start did not finish, cwnd is non-sense,
@@ -417,7 +438,8 @@ void tcp_update_metrics(struct sock *sk)
 			tcp_metric_set(tm, TCP_METRIC_CWND,
 				       (val + tp->snd_ssthresh) >> 1);
 		}
-		if (!tcp_metric_locked(tm, TCP_METRIC_SSTHRESH)) {
+		if (!READ_ONCE(net->ipv4.sysctl_tcp_no_ssthresh_metrics_save) &&
+		    !tcp_metric_locked(tm, TCP_METRIC_SSTHRESH)) {
 			val = tcp_metric_get(tm, TCP_METRIC_SSTHRESH);
 			if (val && tp->snd_ssthresh > val)
 				tcp_metric_set(tm, TCP_METRIC_SSTHRESH,
@@ -426,12 +448,13 @@ void tcp_update_metrics(struct sock *sk)
 		if (!tcp_metric_locked(tm, TCP_METRIC_REORDERING)) {
 			val = tcp_metric_get(tm, TCP_METRIC_REORDERING);
 			if (val < tp->reordering &&
-			    tp->reordering != sysctl_tcp_reordering)
+			    tp->reordering !=
+			    READ_ONCE(net->ipv4.sysctl_tcp_reordering))
 				tcp_metric_set(tm, TCP_METRIC_REORDERING,
 					       tp->reordering);
 		}
 	}
-	tm->tcpm_stamp = jiffies;
+	WRITE_ONCE(tm->tcpm_stamp, jiffies);
 out_unlock:
 	rcu_read_unlock();
 }
@@ -442,16 +465,20 @@ void tcp_init_metrics(struct sock *sk)
 {
 	struct dst_entry *dst = __sk_dst_get(sk);
 	struct tcp_sock *tp = tcp_sk(sk);
+	struct net *net = sock_net(sk);
 	struct tcp_metrics_block *tm;
-	u32 val;
+	u32 val, crtt = 0; /* cached RTT scaled by 8 */
 
-	if (dst == NULL)
+	sk_dst_confirm(sk);
+	/* ssthresh may have been reduced unnecessarily during.
+	 * 3WHS. Restore it back to its initial default.
+	 */
+	tp->snd_ssthresh = TCP_INFINITE_SSTHRESH;
+	if (!dst)
 		goto reset;
 
-	dst_confirm(dst);
-
 	rcu_read_lock();
-	tm = tcp_get_metrics(sk, dst, true);
+	tm = tcp_get_metrics(sk, dst, false);
 	if (!tm) {
 		rcu_read_unlock();
 		goto reset;
@@ -460,33 +487,30 @@ void tcp_init_metrics(struct sock *sk)
 	if (tcp_metric_locked(tm, TCP_METRIC_CWND))
 		tp->snd_cwnd_clamp = tcp_metric_get(tm, TCP_METRIC_CWND);
 
-	val = tcp_metric_get(tm, TCP_METRIC_SSTHRESH);
+	val = READ_ONCE(net->ipv4.sysctl_tcp_no_ssthresh_metrics_save) ?
+	      0 : tcp_metric_get(tm, TCP_METRIC_SSTHRESH);
 	if (val) {
 		tp->snd_ssthresh = val;
 		if (tp->snd_ssthresh > tp->snd_cwnd_clamp)
 			tp->snd_ssthresh = tp->snd_cwnd_clamp;
-	} else {
-		/* ssthresh may have been reduced unnecessarily during.
-		 * 3WHS. Restore it back to its initial default.
-		 */
-		tp->snd_ssthresh = TCP_INFINITE_SSTHRESH;
 	}
 	val = tcp_metric_get(tm, TCP_METRIC_REORDERING);
-	if (val && tp->reordering != val) {
-		tcp_disable_fack(tp);
-		tcp_disable_early_retrans(tp);
+	if (val && tp->reordering != val)
 		tp->reordering = val;
-	}
 
-	val = tcp_metric_get(tm, TCP_METRIC_RTT);
-	if (val == 0 || tp->srtt == 0) {
-		rcu_read_unlock();
-		goto reset;
-	}
-	/* Initial rtt is determined from SYN,SYN-ACK.
-	 * The segment is small and rtt may appear much
-	 * less than real one. Use per-dst memory
-	 * to make it more realistic.
+	crtt = tcp_metric_get(tm, TCP_METRIC_RTT);
+	rcu_read_unlock();
+reset:
+	/* The initial RTT measurement from the SYN/SYN-ACK is not ideal
+	 * to seed the RTO for later data packets because SYN packets are
+	 * small. Use the per-dst cached values to seed the RTO but keep
+	 * the RTT estimator variables intact (e.g., srtt, mdev, rttvar).
+	 * Later the RTO will be updated immediately upon obtaining the first
+	 * data RTT sample (tcp_rtt_estimator()). Hence the cached RTT only
+	 * influences the first RTO but not later RTT estimation.
+	 *
+	 * But if RTT is not available from the SYN (due to retransmits or
+	 * syn cookies) or the cache, force a conservative 3secs timeout.
 	 *
 	 * A bit of theory. RTT is time passed after "normal" sized packet
 	 * is sent until it is ACKed. In normal circumstances sending small
@@ -497,43 +521,25 @@ void tcp_init_metrics(struct sock *sk)
 	 * to low value, and then abruptly stops to do it and starts to delay
 	 * ACKs, wait for troubles.
 	 */
-	val = msecs_to_jiffies(val);
-	if (val > tp->srtt) {
-		tp->srtt = val;
-		tp->rtt_seq = tp->snd_nxt;
-	}
-	val = tcp_metric_get_jiffies(tm, TCP_METRIC_RTTVAR);
-	if (val > tp->mdev) {
-		tp->mdev = val;
-		tp->mdev_max = tp->rttvar = max(tp->mdev, tcp_rto_min(sk));
-	}
-	rcu_read_unlock();
-
-	tcp_set_rto(sk);
-reset:
-	if (tp->srtt == 0) {
+	if (crtt > tp->srtt_us) {
+		/* Set RTO like tcp_rtt_estimator(), but from cached RTT. */
+		crtt /= 8 * USEC_PER_SEC / HZ;
+		inet_csk(sk)->icsk_rto = crtt + max(2 * crtt, tcp_rto_min(sk));
+	} else if (tp->srtt_us == 0) {
 		/* RFC6298: 5.7 We've failed to get a valid RTT sample from
 		 * 3WHS. This is most likely due to retransmission,
 		 * including spurious one. Reset the RTO back to 3secs
 		 * from the more aggressive 1sec to avoid more spurious
 		 * retransmission.
 		 */
-		tp->mdev = tp->mdev_max = tp->rttvar = TCP_TIMEOUT_FALLBACK;
+		tp->rttvar_us = jiffies_to_usecs(TCP_TIMEOUT_FALLBACK);
+		tp->mdev_us = tp->mdev_max_us = tp->rttvar_us;
+
 		inet_csk(sk)->icsk_rto = TCP_TIMEOUT_FALLBACK;
 	}
-	/* Cut cwnd down to 1 per RFC5681 if SYN or SYN-ACK has been
-	 * retransmitted. In light of RFC6298 more aggressive 1sec
-	 * initRTO, we only reset cwnd when more than 1 SYN/SYN-ACK
-	 * retransmission has occurred.
-	 */
-	if (tp->total_retrans > 1)
-		tp->snd_cwnd = 1;
-	else
-		tp->snd_cwnd = tcp_init_cwnd(tp, dst);
-	tp->snd_cwnd_stamp = tcp_time_stamp;
 }
 
-bool tcp_peer_is_proven(struct request_sock *req, struct dst_entry *dst, bool paws_check)
+bool tcp_peer_is_proven(struct request_sock *req, struct dst_entry *dst)
 {
 	struct tcp_metrics_block *tm;
 	bool ret;
@@ -543,104 +549,17 @@ bool tcp_peer_is_proven(struct request_sock *req, struct dst_entry *dst, bool pa
 
 	rcu_read_lock();
 	tm = __tcp_get_metrics_req(req, dst);
-	if (paws_check) {
-		if (tm &&
-		    (u32)get_seconds() - tm->tcpm_ts_stamp < TCP_PAWS_MSL &&
-		    (s32)(tm->tcpm_ts - req->ts_recent) > TCP_PAWS_WINDOW)
-			ret = false;
-		else
-			ret = true;
-	} else {
-		if (tm && tcp_metric_get(tm, TCP_METRIC_RTT) && tm->tcpm_ts_stamp)
-			ret = true;
-		else
-			ret = false;
-	}
-	rcu_read_unlock();
-
-	return ret;
-}
-EXPORT_SYMBOL_GPL(tcp_peer_is_proven);
-
-void tcp_fetch_timewait_stamp(struct sock *sk, struct dst_entry *dst)
-{
-	struct tcp_metrics_block *tm;
-
-	rcu_read_lock();
-	tm = tcp_get_metrics(sk, dst, true);
-	if (tm) {
-		struct tcp_sock *tp = tcp_sk(sk);
-
-		if ((u32)get_seconds() - tm->tcpm_ts_stamp <= TCP_PAWS_MSL) {
-			tp->rx_opt.ts_recent_stamp = tm->tcpm_ts_stamp;
-			tp->rx_opt.ts_recent = tm->tcpm_ts;
-		}
-	}
-	rcu_read_unlock();
-}
-EXPORT_SYMBOL_GPL(tcp_fetch_timewait_stamp);
-
-/* VJ's idea. Save last timestamp seen from this destination and hold
- * it at least for normal timewait interval to use for duplicate
- * segment detection in subsequent connections, before they enter
- * synchronized state.
- */
-bool tcp_remember_stamp(struct sock *sk)
-{
-	struct dst_entry *dst = __sk_dst_get(sk);
-	bool ret = false;
-
-	if (dst) {
-		struct tcp_metrics_block *tm;
-
-		rcu_read_lock();
-		tm = tcp_get_metrics(sk, dst, true);
-		if (tm) {
-			struct tcp_sock *tp = tcp_sk(sk);
-
-			if ((s32)(tm->tcpm_ts - tp->rx_opt.ts_recent) <= 0 ||
-			    ((u32)get_seconds() - tm->tcpm_ts_stamp > TCP_PAWS_MSL &&
-			     tm->tcpm_ts_stamp <= (u32)tp->rx_opt.ts_recent_stamp)) {
-				tm->tcpm_ts_stamp = (u32)tp->rx_opt.ts_recent_stamp;
-				tm->tcpm_ts = tp->rx_opt.ts_recent;
-			}
-			ret = true;
-		}
-		rcu_read_unlock();
-	}
-	return ret;
-}
-
-bool tcp_tw_remember_stamp(struct inet_timewait_sock *tw)
-{
-	struct tcp_metrics_block *tm;
-	bool ret = false;
-
-	rcu_read_lock();
-	tm = __tcp_get_metrics_tw(tw);
-	if (tm) {
-		const struct tcp_timewait_sock *tcptw;
-		struct sock *sk = (struct sock *) tw;
-
-		tcptw = tcp_twsk(sk);
-		if ((s32)(tm->tcpm_ts - tcptw->tw_ts_recent) <= 0 ||
-		    ((u32)get_seconds() - tm->tcpm_ts_stamp > TCP_PAWS_MSL &&
-		     tm->tcpm_ts_stamp <= (u32)tcptw->tw_ts_recent_stamp)) {
-			tm->tcpm_ts_stamp = (u32)tcptw->tw_ts_recent_stamp;
-			tm->tcpm_ts	   = tcptw->tw_ts_recent;
-		}
+	if (tm && tcp_metric_get(tm, TCP_METRIC_RTT))
 		ret = true;
-	}
+	else
+		ret = false;
 	rcu_read_unlock();
 
 	return ret;
 }
 
-static DEFINE_SEQLOCK(fastopen_seqlock);
-
 void tcp_fastopen_cache_get(struct sock *sk, u16 *mss,
-			    struct tcp_fastopen_cookie *cookie,
-			    int *syn_loss, unsigned long *last_syn_loss)
+			    struct tcp_fastopen_cookie *cookie)
 {
 	struct tcp_metrics_block *tm;
 
@@ -655,27 +574,35 @@ void tcp_fastopen_cache_get(struct sock *sk, u16 *mss,
 			if (tfom->mss)
 				*mss = tfom->mss;
 			*cookie = tfom->cookie;
-			*syn_loss = tfom->syn_loss;
-			*last_syn_loss = *syn_loss ? tfom->last_syn_loss : 0;
+			if (cookie->len <= 0 && tfom->try_exp == 1)
+				cookie->exp = true;
 		} while (read_seqretry(&fastopen_seqlock, seq));
 	}
 	rcu_read_unlock();
 }
 
 void tcp_fastopen_cache_set(struct sock *sk, u16 mss,
-			    struct tcp_fastopen_cookie *cookie, bool syn_lost)
+			    struct tcp_fastopen_cookie *cookie, bool syn_lost,
+			    u16 try_exp)
 {
+	struct dst_entry *dst = __sk_dst_get(sk);
 	struct tcp_metrics_block *tm;
 
+	if (!dst)
+		return;
 	rcu_read_lock();
-	tm = tcp_get_metrics(sk, __sk_dst_get(sk), true);
+	tm = tcp_get_metrics(sk, dst, true);
 	if (tm) {
 		struct tcp_fastopen_metrics *tfom = &tm->tcpm_fastopen;
 
 		write_seqlock_bh(&fastopen_seqlock);
-		tfom->mss = mss;
-		if (cookie->len > 0)
+		if (mss)
+			tfom->mss = mss;
+		if (cookie && cookie->len > 0)
 			tfom->cookie = *cookie;
+		else if (try_exp > tfom->try_exp &&
+			 tfom->cookie.len <= 0 && !tfom->cookie.exp)
+			tfom->try_exp = try_exp;
 		if (syn_lost) {
 			++tfom->syn_loss;
 			tfom->last_syn_loss = jiffies;
@@ -686,19 +613,17 @@ void tcp_fastopen_cache_set(struct sock *sk, u16 mss,
 	rcu_read_unlock();
 }
 
-static struct genl_family tcp_metrics_nl_family = {
-	.id		= GENL_ID_GENERATE,
-	.hdrsize	= 0,
-	.name		= TCP_METRICS_GENL_NAME,
-	.version	= TCP_METRICS_GENL_VERSION,
-	.maxattr	= TCP_METRICS_ATTR_MAX,
-	.netnsok	= true,
-};
+static struct genl_family tcp_metrics_nl_family;
 
-static struct nla_policy tcp_metrics_nl_policy[TCP_METRICS_ATTR_MAX + 1] = {
+static const struct nla_policy tcp_metrics_nl_policy[TCP_METRICS_ATTR_MAX + 1] = {
 	[TCP_METRICS_ATTR_ADDR_IPV4]	= { .type = NLA_U32, },
-	[TCP_METRICS_ATTR_ADDR_IPV6]	= { .type = NLA_BINARY,
-					    .len = sizeof(struct in6_addr), },
+	[TCP_METRICS_ATTR_ADDR_IPV6]	=
+		NLA_POLICY_EXACT_LEN(sizeof(struct in6_addr)),
+
+	[TCP_METRICS_ATTR_SADDR_IPV4]	= { .type = NLA_U32, },
+	[TCP_METRICS_ATTR_SADDR_IPV6]	=
+		NLA_POLICY_EXACT_LEN(sizeof(struct in6_addr)),
+
 	/* Following attributes are not received for GET/DEL,
 	 * we keep them for reference
 	 */
@@ -722,15 +647,21 @@ static int tcp_metrics_fill_info(struct sk_buff *msg,
 	struct nlattr *nest;
 	int i;
 
-	switch (tm->tcpm_addr.family) {
+	switch (tm->tcpm_daddr.family) {
 	case AF_INET:
-		if (nla_put_be32(msg, TCP_METRICS_ATTR_ADDR_IPV4,
-				tm->tcpm_addr.addr.a4) < 0)
+		if (nla_put_in_addr(msg, TCP_METRICS_ATTR_ADDR_IPV4,
+				    inetpeer_get_addr_v4(&tm->tcpm_daddr)) < 0)
+			goto nla_put_failure;
+		if (nla_put_in_addr(msg, TCP_METRICS_ATTR_SADDR_IPV4,
+				    inetpeer_get_addr_v4(&tm->tcpm_saddr)) < 0)
 			goto nla_put_failure;
 		break;
 	case AF_INET6:
-		if (nla_put(msg, TCP_METRICS_ATTR_ADDR_IPV6, 16,
-			    tm->tcpm_addr.addr.a6) < 0)
+		if (nla_put_in6_addr(msg, TCP_METRICS_ATTR_ADDR_IPV6,
+				     inetpeer_get_addr_v6(&tm->tcpm_daddr)) < 0)
+			goto nla_put_failure;
+		if (nla_put_in6_addr(msg, TCP_METRICS_ATTR_SADDR_IPV6,
+				     inetpeer_get_addr_v6(&tm->tcpm_saddr)) < 0)
 			goto nla_put_failure;
 		break;
 	default:
@@ -738,27 +669,36 @@ static int tcp_metrics_fill_info(struct sk_buff *msg,
 	}
 
 	if (nla_put_msecs(msg, TCP_METRICS_ATTR_AGE,
-			  jiffies - tm->tcpm_stamp) < 0)
+			  jiffies - READ_ONCE(tm->tcpm_stamp),
+			  TCP_METRICS_ATTR_PAD) < 0)
 		goto nla_put_failure;
-	if (tm->tcpm_ts_stamp) {
-		if (nla_put_s32(msg, TCP_METRICS_ATTR_TW_TS_STAMP,
-				(s32) (get_seconds() - tm->tcpm_ts_stamp)) < 0)
-			goto nla_put_failure;
-		if (nla_put_u32(msg, TCP_METRICS_ATTR_TW_TSVAL,
-				tm->tcpm_ts) < 0)
-			goto nla_put_failure;
-	}
 
 	{
 		int n = 0;
 
-		nest = nla_nest_start(msg, TCP_METRICS_ATTR_VALS);
+		nest = nla_nest_start_noflag(msg, TCP_METRICS_ATTR_VALS);
 		if (!nest)
 			goto nla_put_failure;
-		for (i = 0; i < TCP_METRIC_MAX + 1; i++) {
-			if (!tm->tcpm_vals[i])
+		for (i = 0; i < TCP_METRIC_MAX_KERNEL + 1; i++) {
+			u32 val = tcp_metric_get(tm, i);
+
+			if (!val)
 				continue;
-			if (nla_put_u32(msg, i + 1, tm->tcpm_vals[i]) < 0)
+			if (i == TCP_METRIC_RTT) {
+				if (nla_put_u32(msg, TCP_METRIC_RTT_US + 1,
+						val) < 0)
+					goto nla_put_failure;
+				n++;
+				val = max(val / 1000, 1U);
+			}
+			if (i == TCP_METRIC_RTTVAR) {
+				if (nla_put_u32(msg, TCP_METRIC_RTTVAR_US + 1,
+						val) < 0)
+					goto nla_put_failure;
+				n++;
+				val = max(val / 1000, 1U);
+			}
+			if (nla_put_u32(msg, i + 1, val) < 0)
 				goto nla_put_failure;
 			n++;
 		}
@@ -786,7 +726,8 @@ static int tcp_metrics_fill_info(struct sk_buff *msg,
 		    (nla_put_u16(msg, TCP_METRICS_ATTR_FOPEN_SYN_DROPS,
 				tfom->syn_loss) < 0 ||
 		     nla_put_msecs(msg, TCP_METRICS_ATTR_FOPEN_SYN_DROP_TS,
-				jiffies - tfom->last_syn_loss) < 0))
+				jiffies - tfom->last_syn_loss,
+				TCP_METRICS_ATTR_PAD) < 0))
 			goto nla_put_failure;
 		if (tfom->cookie.len > 0 &&
 		    nla_put(msg, TCP_METRICS_ATTR_FOPEN_COOKIE,
@@ -815,7 +756,8 @@ static int tcp_metrics_dump_info(struct sk_buff *skb,
 	if (tcp_metrics_fill_info(skb, tm) < 0)
 		goto nla_put_failure;
 
-	return genlmsg_end(skb, hdr);
+	genlmsg_end(skb, hdr);
+	return 0;
 
 nla_put_failure:
 	genlmsg_cancel(skb, hdr);
@@ -826,20 +768,24 @@ static int tcp_metrics_nl_dump(struct sk_buff *skb,
 			       struct netlink_callback *cb)
 {
 	struct net *net = sock_net(skb->sk);
-	unsigned int max_rows = 1U << net->ipv4.tcp_metrics_hash_log;
+	unsigned int max_rows = 1U << tcp_metrics_hash_log;
 	unsigned int row, s_row = cb->args[0];
 	int s_col = cb->args[1], col = s_col;
+	int res = 0;
 
 	for (row = s_row; row < max_rows; row++, s_col = 0) {
 		struct tcp_metrics_block *tm;
-		struct tcpm_hash_bucket *hb = net->ipv4.tcp_metrics_hash + row;
+		struct tcpm_hash_bucket *hb = tcp_metrics_hash + row;
 
 		rcu_read_lock();
 		for (col = 0, tm = rcu_dereference(hb->chain); tm;
 		     tm = rcu_dereference(tm->tcpm_next), col++) {
+			if (!net_eq(tm_net(tm), net))
+				continue;
 			if (col < s_col)
 				continue;
-			if (tcp_metrics_dump_info(skb, cb, tm) < 0) {
+			res = tcp_metrics_dump_info(skb, cb, tm);
+			if (res < 0) {
 				rcu_read_unlock();
 				goto done;
 			}
@@ -850,47 +796,68 @@ static int tcp_metrics_nl_dump(struct sk_buff *skb,
 done:
 	cb->args[0] = row;
 	cb->args[1] = col;
-	return skb->len;
+	return res;
 }
 
-static int parse_nl_addr(struct genl_info *info, struct inetpeer_addr *addr,
-			 unsigned int *hash, int optional)
+static int __parse_nl_addr(struct genl_info *info, struct inetpeer_addr *addr,
+			   unsigned int *hash, int optional, int v4, int v6)
 {
 	struct nlattr *a;
 
-	a = info->attrs[TCP_METRICS_ATTR_ADDR_IPV4];
+	a = info->attrs[v4];
 	if (a) {
-		addr->family = AF_INET;
-		addr->addr.a4 = nla_get_be32(a);
-		*hash = (__force unsigned int) addr->addr.a4;
+		inetpeer_set_addr_v4(addr, nla_get_in_addr(a));
+		if (hash)
+			*hash = ipv4_addr_hash(inetpeer_get_addr_v4(addr));
 		return 0;
 	}
-	a = info->attrs[TCP_METRICS_ATTR_ADDR_IPV6];
+	a = info->attrs[v6];
 	if (a) {
-		if (nla_len(a) != sizeof(struct in6_addr))
-			return -EINVAL;
-		addr->family = AF_INET6;
-		memcpy(addr->addr.a6, nla_data(a), sizeof(addr->addr.a6));
-		*hash = ipv6_addr_hash((struct in6_addr *) addr->addr.a6);
+		struct in6_addr in6;
+
+		in6 = nla_get_in6_addr(a);
+		inetpeer_set_addr_v6(addr, &in6);
+		if (hash)
+			*hash = ipv6_addr_hash(inetpeer_get_addr_v6(addr));
 		return 0;
 	}
 	return optional ? 1 : -EAFNOSUPPORT;
 }
 
+static int parse_nl_addr(struct genl_info *info, struct inetpeer_addr *addr,
+			 unsigned int *hash, int optional)
+{
+	return __parse_nl_addr(info, addr, hash, optional,
+			       TCP_METRICS_ATTR_ADDR_IPV4,
+			       TCP_METRICS_ATTR_ADDR_IPV6);
+}
+
+static int parse_nl_saddr(struct genl_info *info, struct inetpeer_addr *addr)
+{
+	return __parse_nl_addr(info, addr, NULL, 0,
+			       TCP_METRICS_ATTR_SADDR_IPV4,
+			       TCP_METRICS_ATTR_SADDR_IPV6);
+}
+
 static int tcp_metrics_nl_cmd_get(struct sk_buff *skb, struct genl_info *info)
 {
 	struct tcp_metrics_block *tm;
-	struct inetpeer_addr addr;
+	struct inetpeer_addr saddr, daddr;
 	unsigned int hash;
 	struct sk_buff *msg;
 	struct net *net = genl_info_net(info);
 	void *reply;
 	int ret;
+	bool src = true;
 
-	ret = parse_nl_addr(info, &addr, &hash, 0);
+	ret = parse_nl_addr(info, &daddr, &hash, 0);
 	if (ret < 0)
 		return ret;
 
+	ret = parse_nl_saddr(info, &saddr);
+	if (ret < 0)
+		src = false;
+
 	msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
 	if (!msg)
 		return -ENOMEM;
@@ -900,12 +867,15 @@ static int tcp_metrics_nl_cmd_get(struct sk_buff *skb, struct genl_info *info)
 	if (!reply)
 		goto nla_put_failure;
 
-	hash = hash_32(hash, net->ipv4.tcp_metrics_hash_log);
+	hash ^= net_hash_mix(net);
+	hash = hash_32(hash, tcp_metrics_hash_log);
 	ret = -ESRCH;
 	rcu_read_lock();
-	for (tm = rcu_dereference(net->ipv4.tcp_metrics_hash[hash].chain); tm;
+	for (tm = rcu_dereference(tcp_metrics_hash[hash].chain); tm;
 	     tm = rcu_dereference(tm->tcpm_next)) {
-		if (addr_same(&tm->tcpm_addr, &addr)) {
+		if (addr_same(&tm->tcpm_daddr, &daddr) &&
+		    (!src || addr_same(&tm->tcpm_saddr, &saddr)) &&
+		    net_eq(tm_net(tm), net)) {
 			ret = tcp_metrics_fill_info(msg, tm);
 			break;
 		}
@@ -925,34 +895,34 @@ out_free:
 	return ret;
 }
 
-#define deref_locked_genl(p)	\
-	rcu_dereference_protected(p, lockdep_genl_is_held() && \
-				     lockdep_is_held(&tcp_metrics_lock))
-
-#define deref_genl(p)	rcu_dereference_protected(p, lockdep_genl_is_held())
-
-static int tcp_metrics_flush_all(struct net *net)
+static void tcp_metrics_flush_all(struct net *net)
 {
-	unsigned int max_rows = 1U << net->ipv4.tcp_metrics_hash_log;
-	struct tcpm_hash_bucket *hb = net->ipv4.tcp_metrics_hash;
+	unsigned int max_rows = 1U << tcp_metrics_hash_log;
+	struct tcpm_hash_bucket *hb = tcp_metrics_hash;
 	struct tcp_metrics_block *tm;
 	unsigned int row;
 
 	for (row = 0; row < max_rows; row++, hb++) {
-		spin_lock_bh(&tcp_metrics_lock);
-		tm = deref_locked_genl(hb->chain);
-		if (tm)
-			hb->chain = NULL;
-		spin_unlock_bh(&tcp_metrics_lock);
-		while (tm) {
-			struct tcp_metrics_block *next;
+		struct tcp_metrics_block __rcu **pp = &hb->chain;
+		bool match;
 
-			next = deref_genl(tm->tcpm_next);
-			kfree_rcu(tm, rcu_head);
-			tm = next;
+		if (!rcu_access_pointer(*pp))
+			continue;
+
+		spin_lock_bh(&tcp_metrics_lock);
+		for (tm = deref_locked(*pp); tm; tm = deref_locked(*pp)) {
+			match = net ? net_eq(tm_net(tm), net) :
+				!check_net(tm_net(tm));
+			if (match) {
+				rcu_assign_pointer(*pp, tm->tcpm_next);
+				kfree_rcu(tm, rcu_head);
+			} else {
+				pp = &tm->tcpm_next;
+			}
 		}
+		spin_unlock_bh(&tcp_metrics_lock);
+		cond_resched();
 	}
-	return 0;
 }
 
 static int tcp_metrics_nl_cmd_del(struct sk_buff *skb, struct genl_info *info)
@@ -960,52 +930,75 @@ static int tcp_metrics_nl_cmd_del(struct sk_buff *skb, struct genl_info *info)
 	struct tcpm_hash_bucket *hb;
 	struct tcp_metrics_block *tm;
 	struct tcp_metrics_block __rcu **pp;
-	struct inetpeer_addr addr;
+	struct inetpeer_addr saddr, daddr;
 	unsigned int hash;
 	struct net *net = genl_info_net(info);
 	int ret;
+	bool src = true, found = false;
 
-	ret = parse_nl_addr(info, &addr, &hash, 1);
+	ret = parse_nl_addr(info, &daddr, &hash, 1);
 	if (ret < 0)
 		return ret;
-	if (ret > 0)
-		return tcp_metrics_flush_all(net);
+	if (ret > 0) {
+		tcp_metrics_flush_all(net);
+		return 0;
+	}
+	ret = parse_nl_saddr(info, &saddr);
+	if (ret < 0)
+		src = false;
 
-	hash = hash_32(hash, net->ipv4.tcp_metrics_hash_log);
-	hb = net->ipv4.tcp_metrics_hash + hash;
+	hash ^= net_hash_mix(net);
+	hash = hash_32(hash, tcp_metrics_hash_log);
+	hb = tcp_metrics_hash + hash;
 	pp = &hb->chain;
 	spin_lock_bh(&tcp_metrics_lock);
-	for (tm = deref_locked_genl(*pp); tm;
-	     pp = &tm->tcpm_next, tm = deref_locked_genl(*pp)) {
-		if (addr_same(&tm->tcpm_addr, &addr)) {
-			*pp = tm->tcpm_next;
-			break;
+	for (tm = deref_locked(*pp); tm; tm = deref_locked(*pp)) {
+		if (addr_same(&tm->tcpm_daddr, &daddr) &&
+		    (!src || addr_same(&tm->tcpm_saddr, &saddr)) &&
+		    net_eq(tm_net(tm), net)) {
+			rcu_assign_pointer(*pp, tm->tcpm_next);
+			kfree_rcu(tm, rcu_head);
+			found = true;
+		} else {
+			pp = &tm->tcpm_next;
 		}
 	}
 	spin_unlock_bh(&tcp_metrics_lock);
-	if (!tm)
+	if (!found)
 		return -ESRCH;
-	kfree_rcu(tm, rcu_head);
 	return 0;
 }
 
-static struct genl_ops tcp_metrics_nl_ops[] = {
+static const struct genl_small_ops tcp_metrics_nl_ops[] = {
 	{
 		.cmd = TCP_METRICS_CMD_GET,
+		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
 		.doit = tcp_metrics_nl_cmd_get,
 		.dumpit = tcp_metrics_nl_dump,
-		.policy = tcp_metrics_nl_policy,
-		.flags = GENL_ADMIN_PERM,
 	},
 	{
 		.cmd = TCP_METRICS_CMD_DEL,
+		.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
 		.doit = tcp_metrics_nl_cmd_del,
-		.policy = tcp_metrics_nl_policy,
 		.flags = GENL_ADMIN_PERM,
 	},
 };
 
-static unsigned int tcpmhash_entries;
+static struct genl_family tcp_metrics_nl_family __ro_after_init = {
+	.hdrsize	= 0,
+	.name		= TCP_METRICS_GENL_NAME,
+	.version	= TCP_METRICS_GENL_VERSION,
+	.maxattr	= TCP_METRICS_ATTR_MAX,
+	.policy = tcp_metrics_nl_policy,
+	.netnsok	= true,
+	.parallel_ops	= true,
+	.module		= THIS_MODULE,
+	.small_ops	= tcp_metrics_nl_ops,
+	.n_small_ops	= ARRAY_SIZE(tcp_metrics_nl_ops),
+	.resv_start_op	= TCP_METRICS_CMD_DEL + 1,
+};
+
+static unsigned int tcpmhash_entries __initdata;
 static int __init set_tcpmhash_entries(char *str)
 {
 	ssize_t ret;
@@ -1021,74 +1014,46 @@ static int __init set_tcpmhash_entries(char *str)
 }
 __setup("tcpmhash_entries=", set_tcpmhash_entries);
 
-static int __net_init tcp_net_metrics_init(struct net *net)
+static void __init tcp_metrics_hash_alloc(void)
 {
+	unsigned int slots = tcpmhash_entries;
 	size_t size;
-	unsigned int slots;
 
-	slots = tcpmhash_entries;
 	if (!slots) {
-		if (totalram_pages >= 128 * 1024)
+		if (totalram_pages() >= 128 * 1024)
 			slots = 16 * 1024;
 		else
 			slots = 8 * 1024;
 	}
 
-	net->ipv4.tcp_metrics_hash_log = order_base_2(slots);
-	size = sizeof(struct tcpm_hash_bucket) << net->ipv4.tcp_metrics_hash_log;
-
-	net->ipv4.tcp_metrics_hash = kzalloc(size, GFP_KERNEL | __GFP_NOWARN);
-	if (!net->ipv4.tcp_metrics_hash)
-		net->ipv4.tcp_metrics_hash = vzalloc(size);
-
-	if (!net->ipv4.tcp_metrics_hash)
-		return -ENOMEM;
+	tcp_metrics_hash_log = order_base_2(slots);
+	size = sizeof(struct tcpm_hash_bucket) << tcp_metrics_hash_log;
 
-	return 0;
+	tcp_metrics_hash = kvzalloc(size, GFP_KERNEL);
+	if (!tcp_metrics_hash)
+		panic("Could not allocate the tcp_metrics hash table\n");
 }
 
-static void __net_exit tcp_net_metrics_exit(struct net *net)
+static void __net_exit tcp_net_metrics_exit_batch(struct list_head *net_exit_list)
 {
-	unsigned int i;
-
-	for (i = 0; i < (1U << net->ipv4.tcp_metrics_hash_log) ; i++) {
-		struct tcp_metrics_block *tm, *next;
-
-		tm = rcu_dereference_protected(net->ipv4.tcp_metrics_hash[i].chain, 1);
-		while (tm) {
-			next = rcu_dereference_protected(tm->tcpm_next, 1);
-			kfree(tm);
-			tm = next;
-		}
-	}
-	if (is_vmalloc_addr(net->ipv4.tcp_metrics_hash))
-		vfree(net->ipv4.tcp_metrics_hash);
-	else
-		kfree(net->ipv4.tcp_metrics_hash);
+	tcp_metrics_flush_all(NULL);
 }
 
 static __net_initdata struct pernet_operations tcp_net_metrics_ops = {
-	.init	=	tcp_net_metrics_init,
-	.exit	=	tcp_net_metrics_exit,
+	.exit_batch	=	tcp_net_metrics_exit_batch,
 };
 
 void __init tcp_metrics_init(void)
 {
 	int ret;
 
+	tcp_metrics_hash_alloc();
+
 	ret = register_pernet_subsys(&tcp_net_metrics_ops);
 	if (ret < 0)
-		goto cleanup;
-	ret = genl_register_family_with_ops(&tcp_metrics_nl_family,
-					    tcp_metrics_nl_ops,
-					    ARRAY_SIZE(tcp_metrics_nl_ops));
-	if (ret < 0)
-		goto cleanup_subsys;
-	return;
-
-cleanup_subsys:
-	unregister_pernet_subsys(&tcp_net_metrics_ops);
+		panic("Could not register tcp_net_metrics_ops\n");
 
-cleanup:
-	return;
+	ret = genl_register_family(&tcp_metrics_nl_family);
+	if (ret < 0)
+		panic("Could not register tcp_metrics generic netlink\n");
 }