26 files changed, 167 insertions, 268 deletions

diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index 6d6dd345bc4d..211c0cc6c3d3 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -1476,22 +1476,20 @@ int inet_ctl_sock_create(struct sock **sk, unsigned short family,
 }
 EXPORT_SYMBOL_GPL(inet_ctl_sock_create);
 
-unsigned long snmp_fold_field(void __percpu *mib[], int offt)
+unsigned long snmp_fold_field(void __percpu *mib, int offt)
 {
 	unsigned long res = 0;
-	int i, j;
+	int i;
 
-	for_each_possible_cpu(i) {
-		for (j = 0; j < SNMP_ARRAY_SZ; j++)
-			res += *(((unsigned long *) per_cpu_ptr(mib[j], i)) + offt);
-	}
+	for_each_possible_cpu(i)
+		res += *(((unsigned long *) per_cpu_ptr(mib, i)) + offt);
 	return res;
 }
 EXPORT_SYMBOL_GPL(snmp_fold_field);
 
 #if BITS_PER_LONG==32
 
-u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t syncp_offset)
+u64 snmp_fold_field64(void __percpu *mib, int offt, size_t syncp_offset)
 {
 	u64 res = 0;
 	int cpu;
@@ -1502,7 +1500,7 @@ u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t syncp_offset)
 		u64 v;
 		unsigned int start;
 
-		bhptr = per_cpu_ptr(mib[0], cpu);
+		bhptr = per_cpu_ptr(mib, cpu);
 		syncp = (struct u64_stats_sync *)(bhptr + syncp_offset);
 		do {
 			start = u64_stats_fetch_begin_irq(syncp);
@@ -1516,25 +1514,6 @@ u64 snmp_fold_field64(void __percpu *mib[], int offt, size_t syncp_offset)
 EXPORT_SYMBOL_GPL(snmp_fold_field64);
 #endif
 
-int snmp_mib_init(void __percpu *ptr[2], size_t mibsize, size_t align)
-{
-	BUG_ON(ptr == NULL);
-	ptr[0] = __alloc_percpu(mibsize, align);
-	if (!ptr[0])
-		return -ENOMEM;
-
-#if SNMP_ARRAY_SZ == 2
-	ptr[1] = __alloc_percpu(mibsize, align);
-	if (!ptr[1]) {
-		free_percpu(ptr[0]);
-		ptr[0] = NULL;
-		return -ENOMEM;
-	}
-#endif
-	return 0;
-}
-EXPORT_SYMBOL_GPL(snmp_mib_init);
-
 #ifdef CONFIG_IP_MULTICAST
 static const struct net_protocol igmp_protocol = {
 	.handler =	igmp_rcv,
@@ -1570,40 +1549,30 @@ static __net_init int ipv4_mib_init_net(struct net *net)
 {
 	int i;
 
-	if (snmp_mib_init((void __percpu **)net->mib.tcp_statistics,
-			  sizeof(struct tcp_mib),
-			  __alignof__(struct tcp_mib)) < 0)
+	net->mib.tcp_statistics = alloc_percpu(struct tcp_mib);
+	if (!net->mib.tcp_statistics)
 		goto err_tcp_mib;
-	if (snmp_mib_init((void __percpu **)net->mib.ip_statistics,
-			  sizeof(struct ipstats_mib),
-			  __alignof__(struct ipstats_mib)) < 0)
+	net->mib.ip_statistics = alloc_percpu(struct ipstats_mib);
+	if (!net->mib.ip_statistics)
 		goto err_ip_mib;
 
 	for_each_possible_cpu(i) {
 		struct ipstats_mib *af_inet_stats;
-		af_inet_stats = per_cpu_ptr(net->mib.ip_statistics[0], i);
+		af_inet_stats = per_cpu_ptr(net->mib.ip_statistics, i);
 		u64_stats_init(&af_inet_stats->syncp);
-#if SNMP_ARRAY_SZ == 2
-		af_inet_stats = per_cpu_ptr(net->mib.ip_statistics[1], i);
-		u64_stats_init(&af_inet_stats->syncp);
-#endif
 	}
 
-	if (snmp_mib_init((void __percpu **)net->mib.net_statistics,
-			  sizeof(struct linux_mib),
-			  __alignof__(struct linux_mib)) < 0)
+	net->mib.net_statistics = alloc_percpu(struct linux_mib);
+	if (!net->mib.net_statistics)
 		goto err_net_mib;
-	if (snmp_mib_init((void __percpu **)net->mib.udp_statistics,
-			  sizeof(struct udp_mib),
-			  __alignof__(struct udp_mib)) < 0)
+	net->mib.udp_statistics = alloc_percpu(struct udp_mib);
+	if (!net->mib.udp_statistics)
 		goto err_udp_mib;
-	if (snmp_mib_init((void __percpu **)net->mib.udplite_statistics,
-			  sizeof(struct udp_mib),
-			  __alignof__(struct udp_mib)) < 0)
+	net->mib.udplite_statistics = alloc_percpu(struct udp_mib);
+	if (!net->mib.udplite_statistics)
 		goto err_udplite_mib;
-	if (snmp_mib_init((void __percpu **)net->mib.icmp_statistics,
-			  sizeof(struct icmp_mib),
-			  __alignof__(struct icmp_mib)) < 0)
+	net->mib.icmp_statistics = alloc_percpu(struct icmp_mib);
+	if (!net->mib.icmp_statistics)
 		goto err_icmp_mib;
 	net->mib.icmpmsg_statistics = kzalloc(sizeof(struct icmpmsg_mib),
 					      GFP_KERNEL);
@@ -1614,17 +1583,17 @@ static __net_init int ipv4_mib_init_net(struct net *net)
 	return 0;
 
 err_icmpmsg_mib:
-	snmp_mib_free((void __percpu **)net->mib.icmp_statistics);
+	free_percpu(net->mib.icmp_statistics);
 err_icmp_mib:
-	snmp_mib_free((void __percpu **)net->mib.udplite_statistics);
+	free_percpu(net->mib.udplite_statistics);
 err_udplite_mib:
-	snmp_mib_free((void __percpu **)net->mib.udp_statistics);
+	free_percpu(net->mib.udp_statistics);
 err_udp_mib:
-	snmp_mib_free((void __percpu **)net->mib.net_statistics);
+	free_percpu(net->mib.net_statistics);
 err_net_mib:
-	snmp_mib_free((void __percpu **)net->mib.ip_statistics);
+	free_percpu(net->mib.ip_statistics);
 err_ip_mib:
-	snmp_mib_free((void __percpu **)net->mib.tcp_statistics);
+	free_percpu(net->mib.tcp_statistics);
 err_tcp_mib:
 	return -ENOMEM;
 }
@@ -1632,12 +1601,12 @@ err_tcp_mib:
 static __net_exit void ipv4_mib_exit_net(struct net *net)
 {
 	kfree(net->mib.icmpmsg_statistics);
-	snmp_mib_free((void __percpu **)net->mib.icmp_statistics);
-	snmp_mib_free((void __percpu **)net->mib.udplite_statistics);
-	snmp_mib_free((void __percpu **)net->mib.udp_statistics);
-	snmp_mib_free((void __percpu **)net->mib.net_statistics);
-	snmp_mib_free((void __percpu **)net->mib.ip_statistics);
-	snmp_mib_free((void __percpu **)net->mib.tcp_statistics);
+	free_percpu(net->mib.icmp_statistics);
+	free_percpu(net->mib.udplite_statistics);
+	free_percpu(net->mib.udp_statistics);
+	free_percpu(net->mib.net_statistics);
+	free_percpu(net->mib.ip_statistics);
+	free_percpu(net->mib.tcp_statistics);
 }
 
 static __net_initdata struct pernet_operations ipv4_mib_ops = {
diff --git a/net/ipv4/devinet.c b/net/ipv4/devinet.c
index bdbf68bb2e2d..78692e46b64f 100644
--- a/net/ipv4/devinet.c
+++ b/net/ipv4/devinet.c
@@ -106,7 +106,6 @@ static const struct nla_policy ifa_ipv4_policy[IFA_MAX+1] = {
 #define IN4_ADDR_HSIZE		(1U << IN4_ADDR_HSIZE_SHIFT)
 
 static struct hlist_head inet_addr_lst[IN4_ADDR_HSIZE];
-static DEFINE_SPINLOCK(inet_addr_hash_lock);
 
 static u32 inet_addr_hash(struct net *net, __be32 addr)
 {
@@ -119,16 +118,14 @@ static void inet_hash_insert(struct net *net, struct in_ifaddr *ifa)
 {
 	u32 hash = inet_addr_hash(net, ifa->ifa_local);
 
-	spin_lock(&inet_addr_hash_lock);
+	ASSERT_RTNL();
 	hlist_add_head_rcu(&ifa->hash, &inet_addr_lst[hash]);
-	spin_unlock(&inet_addr_hash_lock);
 }
 
 static void inet_hash_remove(struct in_ifaddr *ifa)
 {
-	spin_lock(&inet_addr_hash_lock);
+	ASSERT_RTNL();
 	hlist_del_init_rcu(&ifa->hash);
-	spin_unlock(&inet_addr_hash_lock);
 }
 
 /**
diff --git a/net/ipv4/gre_demux.c b/net/ipv4/gre_demux.c
index 250be7421ab3..fbfd829f4049 100644
--- a/net/ipv4/gre_demux.c
+++ b/net/ipv4/gre_demux.c
@@ -93,28 +93,6 @@ void gre_build_header(struct sk_buff *skb, const struct tnl_ptk_info *tpi,
 }
 EXPORT_SYMBOL_GPL(gre_build_header);
 
-static __sum16 check_checksum(struct sk_buff *skb)
-{
-	__sum16 csum = 0;
-
-	switch (skb->ip_summed) {
-	case CHECKSUM_COMPLETE:
-		csum = csum_fold(skb->csum);
-
-		if (!csum)
-			break;
-		/* Fall through. */
-
-	case CHECKSUM_NONE:
-		skb->csum = 0;
-		csum = __skb_checksum_complete(skb);
-		skb->ip_summed = CHECKSUM_COMPLETE;
-		break;
-	}
-
-	return csum;
-}
-
 static int parse_gre_header(struct sk_buff *skb, struct tnl_ptk_info *tpi,
 			    bool *csum_err)
 {
@@ -141,7 +119,7 @@ static int parse_gre_header(struct sk_buff *skb, struct tnl_ptk_info *tpi,
 
 	options = (__be32 *)(greh + 1);
 	if (greh->flags & GRE_CSUM) {
-		if (check_checksum(skb)) {
+		if (skb_checksum_simple_validate(skb)) {
 			*csum_err = true;
 			return -EINVAL;
 		}
diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
index 0134663fdbce..fe52666dc43c 100644
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c
@@ -908,16 +908,8 @@ int icmp_rcv(struct sk_buff *skb)
 
 	ICMP_INC_STATS_BH(net, ICMP_MIB_INMSGS);
 
-	switch (skb->ip_summed) {
-	case CHECKSUM_COMPLETE:
-		if (!csum_fold(skb->csum))
-			break;
-		/* fall through */
-	case CHECKSUM_NONE:
-		skb->csum = 0;
-		if (__skb_checksum_complete(skb))
-			goto csum_error;
-	}
+	if (skb_checksum_simple_validate(skb))
+		goto csum_error;
 
 	if (!pskb_pull(skb, sizeof(*icmph)))
 		goto error;
diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c
index 97e4d1655d26..17d34e3c2ac3 100644
--- a/net/ipv4/igmp.c
+++ b/net/ipv4/igmp.c
@@ -988,16 +988,8 @@ int igmp_rcv(struct sk_buff *skb)
 	if (!pskb_may_pull(skb, sizeof(struct igmphdr)))
 		goto drop;
 
-	switch (skb->ip_summed) {
-	case CHECKSUM_COMPLETE:
-		if (!csum_fold(skb->csum))
-			break;
-		/* fall through */
-	case CHECKSUM_NONE:
-		skb->csum = 0;
-		if (__skb_checksum_complete(skb))
-			goto drop;
-	}
+	if (skb_checksum_simple_validate(skb))
+		goto drop;
 
 	ih = igmp_hdr(skb);
 	switch (ih->type) {
diff --git a/net/ipv4/inetpeer.c b/net/ipv4/inetpeer.c
index 48f424465112..c98cf141f4ed 100644
--- a/net/ipv4/inetpeer.c
+++ b/net/ipv4/inetpeer.c
@@ -120,7 +120,7 @@ int inet_peer_maxttl __read_mostly = 10 * 60 * HZ;	/* usual time to live: 10 min
 static void inetpeer_gc_worker(struct work_struct *work)
 {
 	struct inet_peer *p, *n, *c;
-	LIST_HEAD(list);
+	struct list_head list;
 
 	spin_lock_bh(&gc_lock);
 	list_replace_init(&gc_list, &list);
diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
index 94213c891565..c5a557a06a31 100644
--- a/net/ipv4/ip_gre.c
+++ b/net/ipv4/ip_gre.c
@@ -410,7 +410,7 @@ static int ipgre_open(struct net_device *dev)
 		struct flowi4 fl4;
 		struct rtable *rt;
 
-		rt = ip_route_output_gre(dev_net(dev), &fl4,
+		rt = ip_route_output_gre(t->net, &fl4,
 					 t->parms.iph.daddr,
 					 t->parms.iph.saddr,
 					 t->parms.o_key,
@@ -434,7 +434,7 @@ static int ipgre_close(struct net_device *dev)
 
 	if (ipv4_is_multicast(t->parms.iph.daddr) && t->mlink) {
 		struct in_device *in_dev;
-		in_dev = inetdev_by_index(dev_net(dev), t->mlink);
+		in_dev = inetdev_by_index(t->net, t->mlink);
 		if (in_dev)
 			ip_mc_dec_group(in_dev, t->parms.iph.daddr);
 	}
@@ -478,7 +478,7 @@ static void __gre_tunnel_init(struct net_device *dev)
 	dev->needed_headroom	= LL_MAX_HEADER + sizeof(struct iphdr) + 4;
 	dev->mtu		= ETH_DATA_LEN - sizeof(struct iphdr) - 4;
 
-	dev->features		|= NETIF_F_NETNS_LOCAL | GRE_FEATURES;
+	dev->features		|= GRE_FEATURES;
 	dev->hw_features	|= GRE_FEATURES;
 
 	if (!(tunnel->parms.o_flags & TUNNEL_SEQ)) {
diff --git a/net/ipv4/ip_options.c b/net/ipv4/ip_options.c
index f4ab72e19af9..5e7aecea05cd 100644
--- a/net/ipv4/ip_options.c
+++ b/net/ipv4/ip_options.c
@@ -364,7 +364,7 @@ int ip_options_compile(struct net *net,
 			}
 			if (optptr[2] <= optlen) {
 				unsigned char *timeptr = NULL;
-				if (optptr[2]+3 > optptr[1]) {
+				if (optptr[2]+3 > optlen) {
 					pp_ptr = optptr + 2;
 					goto error;
 				}
@@ -376,7 +376,7 @@ int ip_options_compile(struct net *net,
 					optptr[2] += 4;
 					break;
 				case IPOPT_TS_TSANDADDR:
-					if (optptr[2]+7 > optptr[1]) {
+					if (optptr[2]+7 > optlen) {
 						pp_ptr = optptr + 2;
 						goto error;
 					}
@@ -390,7 +390,7 @@ int ip_options_compile(struct net *net,
 					optptr[2] += 8;
 					break;
 				case IPOPT_TS_PRESPEC:
-					if (optptr[2]+7 > optptr[1]) {
+					if (optptr[2]+7 > optlen) {
 						pp_ptr = optptr + 2;
 						goto error;
 					}
diff --git a/net/ipv4/proc.c b/net/ipv4/proc.c
index ad737fad6d8b..ae0af9386f7c 100644
--- a/net/ipv4/proc.c
+++ b/net/ipv4/proc.c
@@ -345,15 +345,15 @@ static void icmp_put(struct seq_file *seq)
 	for (i = 0; icmpmibmap[i].name != NULL; i++)
 		seq_printf(seq, " Out%s", icmpmibmap[i].name);
 	seq_printf(seq, "\nIcmp: %lu %lu %lu",
-		snmp_fold_field((void __percpu **) net->mib.icmp_statistics, ICMP_MIB_INMSGS),
-		snmp_fold_field((void __percpu **) net->mib.icmp_statistics, ICMP_MIB_INERRORS),
-		snmp_fold_field((void __percpu **) net->mib.icmp_statistics, ICMP_MIB_CSUMERRORS));
+		snmp_fold_field(net->mib.icmp_statistics, ICMP_MIB_INMSGS),
+		snmp_fold_field(net->mib.icmp_statistics, ICMP_MIB_INERRORS),
+		snmp_fold_field(net->mib.icmp_statistics, ICMP_MIB_CSUMERRORS));
 	for (i = 0; icmpmibmap[i].name != NULL; i++)
 		seq_printf(seq, " %lu",
 			   atomic_long_read(ptr + icmpmibmap[i].index));
 	seq_printf(seq, " %lu %lu",
-		snmp_fold_field((void __percpu **) net->mib.icmp_statistics, ICMP_MIB_OUTMSGS),
-		snmp_fold_field((void __percpu **) net->mib.icmp_statistics, ICMP_MIB_OUTERRORS));
+		snmp_fold_field(net->mib.icmp_statistics, ICMP_MIB_OUTMSGS),
+		snmp_fold_field(net->mib.icmp_statistics, ICMP_MIB_OUTERRORS));
 	for (i = 0; icmpmibmap[i].name != NULL; i++)
 		seq_printf(seq, " %lu",
 			   atomic_long_read(ptr + (icmpmibmap[i].index | 0x100)));
@@ -379,7 +379,7 @@ static int snmp_seq_show(struct seq_file *seq, void *v)
 	BUILD_BUG_ON(offsetof(struct ipstats_mib, mibs) != 0);
 	for (i = 0; snmp4_ipstats_list[i].name != NULL; i++)
 		seq_printf(seq, " %llu",
-			   snmp_fold_field64((void __percpu **)net->mib.ip_statistics,
+			   snmp_fold_field64(net->mib.ip_statistics,
 					     snmp4_ipstats_list[i].entry,
 					     offsetof(struct ipstats_mib, syncp)));
 
@@ -395,11 +395,11 @@ static int snmp_seq_show(struct seq_file *seq, void *v)
 		/* MaxConn field is signed, RFC 2012 */
 		if (snmp4_tcp_list[i].entry == TCP_MIB_MAXCONN)
 			seq_printf(seq, " %ld",
-				   snmp_fold_field((void __percpu **)net->mib.tcp_statistics,
+				   snmp_fold_field(net->mib.tcp_statistics,
 						   snmp4_tcp_list[i].entry));
 		else
 			seq_printf(seq, " %lu",
-				   snmp_fold_field((void __percpu **)net->mib.tcp_statistics,
+				   snmp_fold_field(net->mib.tcp_statistics,
 						   snmp4_tcp_list[i].entry));
 	}
 
@@ -410,7 +410,7 @@ static int snmp_seq_show(struct seq_file *seq, void *v)
 	seq_puts(seq, "\nUdp:");
 	for (i = 0; snmp4_udp_list[i].name != NULL; i++)
 		seq_printf(seq, " %lu",
-			   snmp_fold_field((void __percpu **)net->mib.udp_statistics,
+			   snmp_fold_field(net->mib.udp_statistics,
 					   snmp4_udp_list[i].entry));
 
 	/* the UDP and UDP-Lite MIBs are the same */
@@ -421,7 +421,7 @@ static int snmp_seq_show(struct seq_file *seq, void *v)
 	seq_puts(seq, "\nUdpLite:");
 	for (i = 0; snmp4_udp_list[i].name != NULL; i++)
 		seq_printf(seq, " %lu",
-			   snmp_fold_field((void __percpu **)net->mib.udplite_statistics,
+			   snmp_fold_field(net->mib.udplite_statistics,
 					   snmp4_udp_list[i].entry));
 
 	seq_putc(seq, '\n');
@@ -458,7 +458,7 @@ static int netstat_seq_show(struct seq_file *seq, void *v)
 	seq_puts(seq, "\nTcpExt:");
 	for (i = 0; snmp4_net_list[i].name != NULL; i++)
 		seq_printf(seq, " %lu",
-			   snmp_fold_field((void __percpu **)net->mib.net_statistics,
+			   snmp_fold_field(net->mib.net_statistics,
 					   snmp4_net_list[i].entry));
 
 	seq_puts(seq, "\nIpExt:");
@@ -468,7 +468,7 @@ static int netstat_seq_show(struct seq_file *seq, void *v)
 	seq_puts(seq, "\nIpExt:");
 	for (i = 0; snmp4_ipextstats_list[i].name != NULL; i++)
 		seq_printf(seq, " %llu",
-			   snmp_fold_field64((void __percpu **)net->mib.ip_statistics,
+			   snmp_fold_field64(net->mib.ip_statistics,
 					     snmp4_ipextstats_list[i].entry,
 					     offsetof(struct ipstats_mib, syncp)));
 
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 4bd6d52eeffb..eb1dde37e678 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2916,6 +2916,14 @@ static int do_tcp_getsockopt(struct sock *sk, int level,
 	case TCP_USER_TIMEOUT:
 		val = jiffies_to_msecs(icsk->icsk_user_timeout);
 		break;
+
+	case TCP_FASTOPEN:
+		if (icsk->icsk_accept_queue.fastopenq != NULL)
+			val = icsk->icsk_accept_queue.fastopenq->max_qlen;
+		else
+			val = 0;
+		break;
+
 	case TCP_TIMESTAMP:
 		val = tcp_time_stamp + tp->tsoffset;
 		break;
diff --git a/net/ipv4/tcp_bic.c b/net/ipv4/tcp_bic.c
index 821846fb0a7e..d5de69bc04f5 100644
--- a/net/ipv4/tcp_bic.c
+++ b/net/ipv4/tcp_bic.c
@@ -140,13 +140,12 @@ static inline void bictcp_update(struct bictcp *ca, u32 cwnd)
 		ca->cnt = 1;
 }
 
-static void bictcp_cong_avoid(struct sock *sk, u32 ack, u32 acked,
-			      u32 in_flight)
+static void bictcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct bictcp *ca = inet_csk_ca(sk);
 
-	if (!tcp_is_cwnd_limited(sk, in_flight))
+	if (!tcp_is_cwnd_limited(sk))
 		return;
 
 	if (tp->snd_cwnd <= tp->snd_ssthresh)
diff --git a/net/ipv4/tcp_cong.c b/net/ipv4/tcp_cong.c
index 2b9464c93b88..7b09d8b49fa5 100644
--- a/net/ipv4/tcp_cong.c
+++ b/net/ipv4/tcp_cong.c
@@ -276,26 +276,6 @@ int tcp_set_congestion_control(struct sock *sk, const char *name)
 	return err;
 }
 
-/* RFC2861 Check whether we are limited by application or congestion window
- * This is the inverse of cwnd check in tcp_tso_should_defer
- */
-bool tcp_is_cwnd_limited(const struct sock *sk, u32 in_flight)
-{
-	const struct tcp_sock *tp = tcp_sk(sk);
-	u32 left;
-
-	if (in_flight >= tp->snd_cwnd)
-		return true;
-
-	left = tp->snd_cwnd - in_flight;
-	if (sk_can_gso(sk) &&
-	    left * sysctl_tcp_tso_win_divisor < tp->snd_cwnd &&
-	    left < tp->xmit_size_goal_segs)
-		return true;
-	return left <= tcp_max_tso_deferred_mss(tp);
-}
-EXPORT_SYMBOL_GPL(tcp_is_cwnd_limited);
-
 /* Slow start is used when congestion window is no greater than the slow start
  * threshold. We base on RFC2581 and also handle stretch ACKs properly.
  * We do not implement RFC3465 Appropriate Byte Counting (ABC) per se but
@@ -337,11 +317,11 @@ EXPORT_SYMBOL_GPL(tcp_cong_avoid_ai);
 /* This is Jacobson's slow start and congestion avoidance.
  * SIGCOMM '88, p. 328.
  */
-void tcp_reno_cong_avoid(struct sock *sk, u32 ack, u32 acked, u32 in_flight)
+void tcp_reno_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 
-	if (!tcp_is_cwnd_limited(sk, in_flight))
+	if (!tcp_is_cwnd_limited(sk))
 		return;
 
 	/* In "safe" area, increase. */
diff --git a/net/ipv4/tcp_cubic.c b/net/ipv4/tcp_cubic.c
index b4f1b29b08bd..a9bd8a4828a9 100644
--- a/net/ipv4/tcp_cubic.c
+++ b/net/ipv4/tcp_cubic.c
@@ -304,13 +304,12 @@ static inline void bictcp_update(struct bictcp *ca, u32 cwnd)
 		ca->cnt = 1;
 }
 
-static void bictcp_cong_avoid(struct sock *sk, u32 ack, u32 acked,
-			      u32 in_flight)
+static void bictcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct bictcp *ca = inet_csk_ca(sk);
 
-	if (!tcp_is_cwnd_limited(sk, in_flight))
+	if (!tcp_is_cwnd_limited(sk))
 		return;
 
 	if (tp->snd_cwnd <= tp->snd_ssthresh) {
diff --git a/net/ipv4/tcp_highspeed.c b/net/ipv4/tcp_highspeed.c
index 8b9e7bad77c0..1c4908280d92 100644
--- a/net/ipv4/tcp_highspeed.c
+++ b/net/ipv4/tcp_highspeed.c
@@ -109,12 +109,12 @@ static void hstcp_init(struct sock *sk)
 	tp->snd_cwnd_clamp = min_t(u32, tp->snd_cwnd_clamp, 0xffffffff/128);
 }
 
-static void hstcp_cong_avoid(struct sock *sk, u32 ack, u32 acked, u32 in_flight)
+static void hstcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct hstcp *ca = inet_csk_ca(sk);
 
-	if (!tcp_is_cwnd_limited(sk, in_flight))
+	if (!tcp_is_cwnd_limited(sk))
 		return;
 
 	if (tp->snd_cwnd <= tp->snd_ssthresh)
diff --git a/net/ipv4/tcp_htcp.c b/net/ipv4/tcp_htcp.c
index 4a194acfd923..031361311a8b 100644
--- a/net/ipv4/tcp_htcp.c
+++ b/net/ipv4/tcp_htcp.c
@@ -227,12 +227,12 @@ static u32 htcp_recalc_ssthresh(struct sock *sk)
 	return max((tp->snd_cwnd * ca->beta) >> 7, 2U);
 }
 
-static void htcp_cong_avoid(struct sock *sk, u32 ack, u32 acked, u32 in_flight)
+static void htcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct htcp *ca = inet_csk_ca(sk);
 
-	if (!tcp_is_cwnd_limited(sk, in_flight))
+	if (!tcp_is_cwnd_limited(sk))
 		return;
 
 	if (tp->snd_cwnd <= tp->snd_ssthresh)
diff --git a/net/ipv4/tcp_hybla.c b/net/ipv4/tcp_hybla.c
index a15a799bf768..d8f8f05a4951 100644
--- a/net/ipv4/tcp_hybla.c
+++ b/net/ipv4/tcp_hybla.c
@@ -87,8 +87,7 @@ static inline u32 hybla_fraction(u32 odds)
  *     o Give cwnd a new value based on the model proposed
  *     o remember increments <1
  */
-static void hybla_cong_avoid(struct sock *sk, u32 ack, u32 acked,
-			     u32 in_flight)
+static void hybla_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct hybla *ca = inet_csk_ca(sk);
@@ -101,11 +100,11 @@ static void hybla_cong_avoid(struct sock *sk, u32 ack, u32 acked,
 		ca->minrtt_us = tp->srtt_us;
 	}
 
-	if (!tcp_is_cwnd_limited(sk, in_flight))
+	if (!tcp_is_cwnd_limited(sk))
 		return;
 
 	if (!ca->hybla_en) {
-		tcp_reno_cong_avoid(sk, ack, acked, in_flight);
+		tcp_reno_cong_avoid(sk, ack, acked);
 		return;
 	}
 
diff --git a/net/ipv4/tcp_illinois.c b/net/ipv4/tcp_illinois.c
index 863d105e3015..5999b3972e64 100644
--- a/net/ipv4/tcp_illinois.c
+++ b/net/ipv4/tcp_illinois.c
@@ -255,8 +255,7 @@ static void tcp_illinois_state(struct sock *sk, u8 new_state)
 /*
  * Increase window in response to successful acknowledgment.
  */
-static void tcp_illinois_cong_avoid(struct sock *sk, u32 ack, u32 acked,
-				    u32 in_flight)
+static void tcp_illinois_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct illinois *ca = inet_csk_ca(sk);
@@ -265,7 +264,7 @@ static void tcp_illinois_cong_avoid(struct sock *sk, u32 ack, u32 acked,
 		update_params(sk);
 
 	/* RFC2861 only increase cwnd if fully utilized */
-	if (!tcp_is_cwnd_limited(sk, in_flight))
+	if (!tcp_is_cwnd_limited(sk))
 		return;
 
 	/* In slow start */
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index d6b46eb2f94c..350b2072f0ab 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -2938,10 +2938,11 @@ static void tcp_synack_rtt_meas(struct sock *sk, const u32 synack_stamp)
 		tcp_ack_update_rtt(sk, FLAG_SYN_ACKED, seq_rtt_us, -1L);
 }
 
-static void tcp_cong_avoid(struct sock *sk, u32 ack, u32 acked, u32 in_flight)
+static void tcp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	const struct inet_connection_sock *icsk = inet_csk(sk);
-	icsk->icsk_ca_ops->cong_avoid(sk, ack, acked, in_flight);
+
+	icsk->icsk_ca_ops->cong_avoid(sk, ack, acked);
 	tcp_sk(sk)->snd_cwnd_stamp = tcp_time_stamp;
 }
 
@@ -3364,7 +3365,6 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag)
 	u32 ack_seq = TCP_SKB_CB(skb)->seq;
 	u32 ack = TCP_SKB_CB(skb)->ack_seq;
 	bool is_dupack = false;
-	u32 prior_in_flight;
 	u32 prior_fackets;
 	int prior_packets = tp->packets_out;
 	const int prior_unsacked = tp->packets_out - tp->sacked_out;
@@ -3397,7 +3397,6 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag)
 		flag |= FLAG_SND_UNA_ADVANCED;
 
 	prior_fackets = tp->fackets_out;
-	prior_in_flight = tcp_packets_in_flight(tp);
 
 	/* ts_recent update must be made after we are sure that the packet
 	 * is in window.
@@ -3452,7 +3451,7 @@ static int tcp_ack(struct sock *sk, const struct sk_buff *skb, int flag)
 
 	/* Advance cwnd if state allows */
 	if (tcp_may_raise_cwnd(sk, flag))
-		tcp_cong_avoid(sk, ack, acked, prior_in_flight);
+		tcp_cong_avoid(sk, ack, acked);
 
 	if (tcp_ack_is_dubious(sk, flag)) {
 		is_dupack = !(flag & (FLAG_SND_UNA_ADVANCED | FLAG_NOT_DUP));
@@ -4703,28 +4702,6 @@ static int tcp_prune_queue(struct sock *sk)
 	return -1;
 }
 
-/* RFC2861, slow part. Adjust cwnd, after it was not full during one rto.
- * As additional protections, we do not touch cwnd in retransmission phases,
- * and if application hit its sndbuf limit recently.
- */
-void tcp_cwnd_application_limited(struct sock *sk)
-{
-	struct tcp_sock *tp = tcp_sk(sk);
-
-	if (inet_csk(sk)->icsk_ca_state == TCP_CA_Open &&
-	    sk->sk_socket && !test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
-		/* Limited by application or receiver window. */
-		u32 init_win = tcp_init_cwnd(tp, __sk_dst_get(sk));
-		u32 win_used = max(tp->snd_cwnd_used, init_win);
-		if (win_used < tp->snd_cwnd) {
-			tp->snd_ssthresh = tcp_current_ssthresh(sk);
-			tp->snd_cwnd = (tp->snd_cwnd + win_used) >> 1;
-		}
-		tp->snd_cwnd_used = 0;
-	}
-	tp->snd_cwnd_stamp = tcp_time_stamp;
-}
-
 static bool tcp_should_expand_sndbuf(const struct sock *sk)
 {
 	const struct tcp_sock *tp = tcp_sk(sk);
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 438f3b95143d..ad166dcc278f 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1744,28 +1744,6 @@ static struct sock *tcp_v4_hnd_req(struct sock *sk, struct sk_buff *skb)
 	return sk;
 }
 
-static __sum16 tcp_v4_checksum_init(struct sk_buff *skb)
-{
-	const struct iphdr *iph = ip_hdr(skb);
-
-	if (skb->ip_summed == CHECKSUM_COMPLETE) {
-		if (!tcp_v4_check(skb->len, iph->saddr,
-				  iph->daddr, skb->csum)) {
-			skb->ip_summed = CHECKSUM_UNNECESSARY;
-			return 0;
-		}
-	}
-
-	skb->csum = csum_tcpudp_nofold(iph->saddr, iph->daddr,
-				       skb->len, IPPROTO_TCP, 0);
-
-	if (skb->len <= 76) {
-		return __skb_checksum_complete(skb);
-	}
-	return 0;
-}
-
-
 /* The socket must have it's spinlock held when we get
  * here.
  *
@@ -1960,7 +1938,8 @@ int tcp_v4_rcv(struct sk_buff *skb)
 	 * Packet length and doff are validated by header prediction,
 	 * provided case of th->doff==0 is eliminated.
 	 * So, we defer the checks. */
-	if (!skb_csum_unnecessary(skb) && tcp_v4_checksum_init(skb))
+
+	if (skb_checksum_init(skb, IPPROTO_TCP, inet_compute_pseudo))
 		goto csum_error;
 
 	th = tcp_hdr(skb);
diff --git a/net/ipv4/tcp_lp.c b/net/ipv4/tcp_lp.c
index c9aecae31327..1e70fa8fa793 100644
--- a/net/ipv4/tcp_lp.c
+++ b/net/ipv4/tcp_lp.c
@@ -115,13 +115,12 @@ static void tcp_lp_init(struct sock *sk)
  * Will only call newReno CA when away from inference.
  * From TCP-LP's paper, this will be handled in additive increasement.
  */
-static void tcp_lp_cong_avoid(struct sock *sk, u32 ack, u32 acked,
-			      u32 in_flight)
+static void tcp_lp_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct lp *lp = inet_csk_ca(sk);
 
 	if (!(lp->flag & LP_WITHIN_INF))
-		tcp_reno_cong_avoid(sk, ack, acked, in_flight);
+		tcp_reno_cong_avoid(sk, ack, acked);
 }
 
 /**
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 12d6016bdd9a..694711a140d4 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -878,15 +878,8 @@ static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it,
 	BUG_ON(!skb || !tcp_skb_pcount(skb));
 
 	if (clone_it) {
-		const struct sk_buff *fclone = skb + 1;
-
 		skb_mstamp_get(&skb->skb_mstamp);
 
-		if (unlikely(skb->fclone == SKB_FCLONE_ORIG &&
-			     fclone->fclone == SKB_FCLONE_CLONE))
-			NET_INC_STATS(sock_net(sk),
-				      LINUX_MIB_TCPSPURIOUS_RTX_HOSTQUEUES);
-
 		if (unlikely(skb_cloned(skb)))
 			skb = pskb_copy(skb, gfp_mask);
 		else
@@ -1387,12 +1380,35 @@ unsigned int tcp_current_mss(struct sock *sk)
 	return mss_now;
 }
 
-/* Congestion window validation. (RFC2861) */
-static void tcp_cwnd_validate(struct sock *sk)
+/* RFC2861, slow part. Adjust cwnd, after it was not full during one rto.
+ * As additional protections, we do not touch cwnd in retransmission phases,
+ * and if application hit its sndbuf limit recently.
+ */
+static void tcp_cwnd_application_limited(struct sock *sk)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 
-	if (tp->packets_out >= tp->snd_cwnd) {
+	if (inet_csk(sk)->icsk_ca_state == TCP_CA_Open &&
+	    sk->sk_socket && !test_bit(SOCK_NOSPACE, &sk->sk_socket->flags)) {
+		/* Limited by application or receiver window. */
+		u32 init_win = tcp_init_cwnd(tp, __sk_dst_get(sk));
+		u32 win_used = max(tp->snd_cwnd_used, init_win);
+		if (win_used < tp->snd_cwnd) {
+			tp->snd_ssthresh = tcp_current_ssthresh(sk);
+			tp->snd_cwnd = (tp->snd_cwnd + win_used) >> 1;
+		}
+		tp->snd_cwnd_used = 0;
+	}
+	tp->snd_cwnd_stamp = tcp_time_stamp;
+}
+
+static void tcp_cwnd_validate(struct sock *sk, u32 unsent_segs)
+{
+	struct tcp_sock *tp = tcp_sk(sk);
+
+	tp->lsnd_pending = tp->packets_out + unsent_segs;
+
+	if (tcp_is_cwnd_limited(sk)) {
 		/* Network is feed fully. */
 		tp->snd_cwnd_used = 0;
 		tp->snd_cwnd_stamp = tcp_time_stamp;
@@ -1865,7 +1881,7 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct sk_buff *skb;
-	unsigned int tso_segs, sent_pkts;
+	unsigned int tso_segs, sent_pkts, unsent_segs = 0;
 	int cwnd_quota;
 	int result;
 
@@ -1909,7 +1925,7 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 				break;
 		} else {
 			if (!push_one && tcp_tso_should_defer(sk, skb))
-				break;
+				goto compute_unsent_segs;
 		}
 
 		/* TCP Small Queues :
@@ -1934,8 +1950,14 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 			 * there is no smp_mb__after_set_bit() yet
 			 */
 			smp_mb__after_clear_bit();
-			if (atomic_read(&sk->sk_wmem_alloc) > limit)
+			if (atomic_read(&sk->sk_wmem_alloc) > limit) {
+				u32 unsent_bytes;
+
+compute_unsent_segs:
+				unsent_bytes = tp->write_seq - tp->snd_nxt;
+				unsent_segs = DIV_ROUND_UP(unsent_bytes, mss_now);
 				break;
+			}
 		}
 
 		limit = mss_now;
@@ -1975,7 +1997,7 @@ repair:
 		/* Send one loss probe per tail loss episode. */
 		if (push_one != 2)
 			tcp_schedule_loss_probe(sk);
-		tcp_cwnd_validate(sk);
+		tcp_cwnd_validate(sk, unsent_segs);
 		return false;
 	}
 	return (push_one == 2) || (!tp->packets_out && tcp_send_head(sk));
@@ -2039,6 +2061,25 @@ bool tcp_schedule_loss_probe(struct sock *sk)
 	return true;
 }
 
+/* Thanks to skb fast clones, we can detect if a prior transmit of
+ * a packet is still in a qdisc or driver queue.
+ * In this case, there is very little point doing a retransmit !
+ * Note: This is called from BH context only.
+ */
+static bool skb_still_in_host_queue(const struct sock *sk,
+				    const struct sk_buff *skb)
+{
+	const struct sk_buff *fclone = skb + 1;
+
+	if (unlikely(skb->fclone == SKB_FCLONE_ORIG &&
+		     fclone->fclone == SKB_FCLONE_CLONE)) {
+		NET_INC_STATS_BH(sock_net(sk),
+				 LINUX_MIB_TCPSPURIOUS_RTX_HOSTQUEUES);
+		return true;
+	}
+	return false;
+}
+
 /* When probe timeout (PTO) fires, send a new segment if one exists, else
  * retransmit the last segment.
  */
@@ -2064,6 +2105,9 @@ void tcp_send_loss_probe(struct sock *sk)
 	if (WARN_ON(!skb))
 		goto rearm_timer;
 
+	if (skb_still_in_host_queue(sk, skb))
+		goto rearm_timer;
+
 	pcount = tcp_skb_pcount(skb);
 	if (WARN_ON(!pcount))
 		goto rearm_timer;
@@ -2385,6 +2429,9 @@ int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb)
 	    min(sk->sk_wmem_queued + (sk->sk_wmem_queued >> 2), sk->sk_sndbuf))
 		return -EAGAIN;
 
+	if (skb_still_in_host_queue(sk, skb))
+		return -EBUSY;
+
 	if (before(TCP_SKB_CB(skb)->seq, tp->snd_una)) {
 		if (before(TCP_SKB_CB(skb)->end_seq, tp->snd_una))
 			BUG();
@@ -2478,7 +2525,7 @@ int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb)
 		 * see tcp_input.c tcp_sacktag_write_queue().
 		 */
 		TCP_SKB_CB(skb)->ack_seq = tp->snd_nxt;
-	} else {
+	} else if (err != -EBUSY) {
 		NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPRETRANSFAIL);
 	}
 	return err;
diff --git a/net/ipv4/tcp_scalable.c b/net/ipv4/tcp_scalable.c
index 0ac50836da4d..8250949b8853 100644
--- a/net/ipv4/tcp_scalable.c
+++ b/net/ipv4/tcp_scalable.c
@@ -15,12 +15,11 @@
 #define TCP_SCALABLE_AI_CNT	50U
 #define TCP_SCALABLE_MD_SCALE	3
 
-static void tcp_scalable_cong_avoid(struct sock *sk, u32 ack, u32 acked,
-				    u32 in_flight)
+static void tcp_scalable_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 
-	if (!tcp_is_cwnd_limited(sk, in_flight))
+	if (!tcp_is_cwnd_limited(sk))
 		return;
 
 	if (tp->snd_cwnd <= tp->snd_ssthresh)
diff --git a/net/ipv4/tcp_vegas.c b/net/ipv4/tcp_vegas.c
index 48539fff6357..9a5e05f27f4f 100644
--- a/net/ipv4/tcp_vegas.c
+++ b/net/ipv4/tcp_vegas.c
@@ -163,14 +163,13 @@ static inline u32 tcp_vegas_ssthresh(struct tcp_sock *tp)
 	return  min(tp->snd_ssthresh, tp->snd_cwnd-1);
 }
 
-static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 acked,
-				 u32 in_flight)
+static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct vegas *vegas = inet_csk_ca(sk);
 
 	if (!vegas->doing_vegas_now) {
-		tcp_reno_cong_avoid(sk, ack, acked, in_flight);
+		tcp_reno_cong_avoid(sk, ack, acked);
 		return;
 	}
 
@@ -195,7 +194,7 @@ static void tcp_vegas_cong_avoid(struct sock *sk, u32 ack, u32 acked,
 			/* We don't have enough RTT samples to do the Vegas
 			 * calculation, so we'll behave like Reno.
 			 */
-			tcp_reno_cong_avoid(sk, ack, acked, in_flight);
+			tcp_reno_cong_avoid(sk, ack, acked);
 		} else {
 			u32 rtt, diff;
 			u64 target_cwnd;
diff --git a/net/ipv4/tcp_veno.c b/net/ipv4/tcp_veno.c
index 1b8e28fcd7e1..27b9825753d1 100644
--- a/net/ipv4/tcp_veno.c
+++ b/net/ipv4/tcp_veno.c
@@ -114,19 +114,18 @@ static void tcp_veno_cwnd_event(struct sock *sk, enum tcp_ca_event event)
 		tcp_veno_init(sk);
 }
 
-static void tcp_veno_cong_avoid(struct sock *sk, u32 ack, u32 acked,
-				u32 in_flight)
+static void tcp_veno_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct veno *veno = inet_csk_ca(sk);
 
 	if (!veno->doing_veno_now) {
-		tcp_reno_cong_avoid(sk, ack, acked, in_flight);
+		tcp_reno_cong_avoid(sk, ack, acked);
 		return;
 	}
 
 	/* limited by applications */
-	if (!tcp_is_cwnd_limited(sk, in_flight))
+	if (!tcp_is_cwnd_limited(sk))
 		return;
 
 	/* We do the Veno calculations only if we got enough rtt samples */
@@ -134,7 +133,7 @@ static void tcp_veno_cong_avoid(struct sock *sk, u32 ack, u32 acked,
 		/* We don't have enough rtt samples to do the Veno
 		 * calculation, so we'll behave like Reno.
 		 */
-		tcp_reno_cong_avoid(sk, ack, acked, in_flight);
+		tcp_reno_cong_avoid(sk, ack, acked);
 	} else {
 		u64 target_cwnd;
 		u32 rtt;
diff --git a/net/ipv4/tcp_yeah.c b/net/ipv4/tcp_yeah.c
index 5ede0e727945..599b79b8eac0 100644
--- a/net/ipv4/tcp_yeah.c
+++ b/net/ipv4/tcp_yeah.c
@@ -69,13 +69,12 @@ static void tcp_yeah_pkts_acked(struct sock *sk, u32 pkts_acked, s32 rtt_us)
 	tcp_vegas_pkts_acked(sk, pkts_acked, rtt_us);
 }
 
-static void tcp_yeah_cong_avoid(struct sock *sk, u32 ack, u32 acked,
-				u32 in_flight)
+static void tcp_yeah_cong_avoid(struct sock *sk, u32 ack, u32 acked)
 {
 	struct tcp_sock *tp = tcp_sk(sk);
 	struct yeah *yeah = inet_csk_ca(sk);
 
-	if (!tcp_is_cwnd_limited(sk, in_flight))
+	if (!tcp_is_cwnd_limited(sk))
 		return;
 
 	if (tp->snd_cwnd <= tp->snd_ssthresh)
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index 4468e1adc094..54ea0a3a48f1 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -1495,6 +1495,10 @@ int udp_queue_rcv_skb(struct sock *sk, struct sk_buff *skb)
 		if (skb->len > sizeof(struct udphdr) && encap_rcv != NULL) {
 			int ret;
 
+			/* Verify checksum before giving to encap */
+			if (udp_lib_checksum_complete(skb))
+				goto csum_error;
+
 			ret = encap_rcv(sk, skb);
 			if (ret <= 0) {
 				UDP_INC_STATS_BH(sock_net(sk),
@@ -1672,7 +1676,6 @@ static int __udp4_lib_mcast_deliver(struct net *net, struct sk_buff *skb,
 static inline int udp4_csum_init(struct sk_buff *skb, struct udphdr *uh,
 				 int proto)
 {
-	const struct iphdr *iph;
 	int err;
 
 	UDP_SKB_CB(skb)->partial_cov = 0;
@@ -1684,22 +1687,8 @@ static inline int udp4_csum_init(struct sk_buff *skb, struct udphdr *uh,
 			return err;
 	}
 
-	iph = ip_hdr(skb);
-	if (uh->check == 0) {
-		skb->ip_summed = CHECKSUM_UNNECESSARY;
-	} else if (skb->ip_summed == CHECKSUM_COMPLETE) {
-		if (!csum_tcpudp_magic(iph->saddr, iph->daddr, skb->len,
-				      proto, skb->csum))
-			skb->ip_summed = CHECKSUM_UNNECESSARY;
-	}
-	if (!skb_csum_unnecessary(skb))
-		skb->csum = csum_tcpudp_nofold(iph->saddr, iph->daddr,
-					       skb->len, proto, 0);
-	/* Probably, we should checksum udp header (it should be in cache
-	 * in any case) and data in tiny packets (< rx copybreak).
-	 */
-
-	return 0;
+	return skb_checksum_init_zero_check(skb, proto, uh->check,
+					    inet_compute_pseudo);
 }
 
 /*