[PATCH net-next 0/6] tcp: misc changes

[PATCH net-next 0/6] tcp: misc changes

[Eric Dumazet]

Minor changes, following recent changes in TCP stack.

Eric Dumazet (6):
  tcp: add a drop_reason pointer to tcp_check_req()
  tcp: add four drop reasons to tcp_check_req()
  tcp: convert to dev_net_rcu()
  net: gro: convert four dev_net() calls
  tcp: remove READ_ONCE(req->ts_recent)
  tcp: tcp_set_window_clamp() cleanup

 include/net/dropreason-core.h  |  9 +++++++++
 include/net/inet6_hashtables.h |  2 +-
 include/net/inet_hashtables.h  |  2 +-
 include/net/tcp.h              |  2 +-
 net/ipv4/tcp.c                 | 36 +++++++++++++++++-----------------
 net/ipv4/tcp_input.c           |  5 ++---
 net/ipv4/tcp_ipv4.c            | 17 ++++++++--------
 net/ipv4/tcp_metrics.c         |  6 +++---
 net/ipv4/tcp_minisocks.c       | 17 +++++++++++-----
 net/ipv4/tcp_offload.c         |  2 +-
 net/ipv4/tcp_output.c          |  2 +-
 net/ipv4/udp_offload.c         |  2 +-
 net/ipv6/tcp_ipv6.c            | 27 +++++++++++++------------
 net/ipv6/tcpv6_offload.c       |  2 +-
 net/ipv6/udp_offload.c         |  2 +-
 15 files changed, 75 insertions(+), 58 deletions(-)

-- 
2.48.1.711.g2feabab25a-goog

[PATCH net-next 1/6] tcp: add a drop_reason pointer to tcp_check_req()

[Eric Dumazet]

We want to add new drop reasons for packets dropped in 3WHS in the
following patches.

tcp_rcv_state_process() has to set reason to TCP_FASTOPEN,
because tcp_check_req() will conditionally overwrite the drop_reason.

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 include/net/tcp.h        | 2 +-
 net/ipv4/tcp_input.c     | 5 ++---
 net/ipv4/tcp_ipv4.c      | 3 ++-
 net/ipv4/tcp_minisocks.c | 3 ++-
 net/ipv6/tcp_ipv6.c      | 3 ++-
 5 files changed, 9 insertions(+), 7 deletions(-)

diff --git a/include/net/tcp.h b/include/net/tcp.h
index f9b9377a289740b907594a0993fc5d70ed36aaac..a9bc959fb102fc6697b4a664b3773b47b3309f13 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -392,7 +392,7 @@ enum tcp_tw_status tcp_timewait_state_process(struct inet_timewait_sock *tw,
 					      u32 *tw_isn);
 struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb,
 			   struct request_sock *req, bool fastopen,
-			   bool *lost_race);
+			   bool *lost_race, enum skb_drop_reason *drop_reason);
 enum skb_drop_reason tcp_child_process(struct sock *parent, struct sock *child,
 				       struct sk_buff *skb);
 void tcp_enter_loss(struct sock *sk);
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index d22ad553b45b17218d5362ea58a4f82559afb851..4e221234808898131a462bc93ee4c9c0ae04309e 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -6812,10 +6812,9 @@ tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb)
 		WARN_ON_ONCE(sk->sk_state != TCP_SYN_RECV &&
 		    sk->sk_state != TCP_FIN_WAIT1);
 
-		if (!tcp_check_req(sk, skb, req, true, &req_stolen)) {
-			SKB_DR_SET(reason, TCP_FASTOPEN);
+		SKB_DR_SET(reason, TCP_FASTOPEN);
+		if (!tcp_check_req(sk, skb, req, true, &req_stolen, &reason))
 			goto discard;
-		}
 	}
 
 	if (!th->ack && !th->rst && !th->syn) {
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 7900855237d929d8260e65fe95e367345bb3ecd2..218f01a8cc5f6c410043f07293e9e51840c1f1cb 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -2265,7 +2265,8 @@ int tcp_v4_rcv(struct sk_buff *skb)
 			th = (const struct tcphdr *)skb->data;
 			iph = ip_hdr(skb);
 			tcp_v4_fill_cb(skb, iph, th);
-			nsk = tcp_check_req(sk, skb, req, false, &req_stolen);
+			nsk = tcp_check_req(sk, skb, req, false, &req_stolen,
+					    &drop_reason);
 		} else {
 			drop_reason = SKB_DROP_REASON_SOCKET_FILTER;
 		}
diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index 29b54ade757231dc264fd8a2c357eff1b2ccbb6b..46c86c4f80e9f450834c72f28e3d16b0cffbbd1d 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -657,7 +657,8 @@ EXPORT_SYMBOL(tcp_create_openreq_child);
 
 struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb,
 			   struct request_sock *req,
-			   bool fastopen, bool *req_stolen)
+			   bool fastopen, bool *req_stolen,
+			   enum skb_drop_reason *drop_reason)
 {
 	struct tcp_options_received tmp_opt;
 	struct sock *child;
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index a806082602985fd351c5184f52dc3011c71540a9..d01088ab80d24eb0f829166faae791221d95bf9e 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1828,7 +1828,8 @@ INDIRECT_CALLABLE_SCOPE int tcp_v6_rcv(struct sk_buff *skb)
 			th = (const struct tcphdr *)skb->data;
 			hdr = ipv6_hdr(skb);
 			tcp_v6_fill_cb(skb, hdr, th);
-			nsk = tcp_check_req(sk, skb, req, false, &req_stolen);
+			nsk = tcp_check_req(sk, skb, req, false, &req_stolen,
+					    &drop_reason);
 		} else {
 			drop_reason = SKB_DROP_REASON_SOCKET_FILTER;
 		}
-- 
2.48.1.711.g2feabab25a-goog

[PATCH net-next 2/6] tcp: add four drop reasons to tcp_check_req()

[Eric Dumazet]

Use two existing drop reasons in tcp_check_req():

- TCP_RFC7323_PAWS

- TCP_OVERWINDOW

Add two new ones:

- TCP_RFC7323_TSECR (corresponds to LINUX_MIB_TSECRREJECTED)

- TCP_LISTEN_OVERFLOW (when a listener accept queue is full)

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 include/net/dropreason-core.h |  9 +++++++++
 net/ipv4/tcp_minisocks.c      | 10 ++++++++--
 2 files changed, 17 insertions(+), 2 deletions(-)

diff --git a/include/net/dropreason-core.h b/include/net/dropreason-core.h
index 32a34dfe8cc58fb1afda8922a52249080f1183b5..da806c2c620f774155fb1d6d2d5fbfcf483c4c0e 100644
--- a/include/net/dropreason-core.h
+++ b/include/net/dropreason-core.h
@@ -40,6 +40,8 @@
 	FN(TCP_OFOMERGE)		\
 	FN(TCP_RFC7323_PAWS)		\
 	FN(TCP_RFC7323_PAWS_ACK)	\
+	FN(TCP_RFC7323_TSECR)		\
+	FN(TCP_LISTEN_OVERFLOW)		\
 	FN(TCP_OLD_SEQUENCE)		\
 	FN(TCP_INVALID_SEQUENCE)	\
 	FN(TCP_INVALID_ACK_SEQUENCE)	\
@@ -281,6 +283,13 @@ enum skb_drop_reason {
 	 * Corresponds to LINUX_MIB_PAWS_OLD_ACK.
 	 */
 	SKB_DROP_REASON_TCP_RFC7323_PAWS_ACK,
+	/**
+	 * @SKB_DROP_REASON_TCP_RFC7323_TSECR: PAWS check, invalid TSEcr.
+	 * Corresponds to LINUX_MIB_TSECRREJECTED.
+	 */
+	SKB_DROP_REASON_TCP_RFC7323_TSECR,
+	/** @SKB_DROP_TCP_REASON_LISTEN_OVERFLOW: listener queue full. */
+	SKB_DROP_REASON_TCP_LISTEN_OVERFLOW,
 	/** @SKB_DROP_REASON_TCP_OLD_SEQUENCE: Old SEQ field (duplicate packet) */
 	SKB_DROP_REASON_TCP_OLD_SEQUENCE,
 	/** @SKB_DROP_REASON_TCP_INVALID_SEQUENCE: Not acceptable SEQ field */
diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index 46c86c4f80e9f450834c72f28e3d16b0cffbbd1d..ba4a5d7f251d8ed093b38155d9b1a9f50bfcfe32 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -809,10 +809,15 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb,
 					  LINUX_MIB_TCPACKSKIPPEDSYNRECV,
 					  &tcp_rsk(req)->last_oow_ack_time))
 			req->rsk_ops->send_ack(sk, skb, req);
-		if (paws_reject)
+		if (paws_reject) {
+			SKB_DR_SET(*drop_reason, TCP_RFC7323_PAWS);
 			NET_INC_STATS(sock_net(sk), LINUX_MIB_PAWSESTABREJECTED);
-		else if (tsecr_reject)
+		} else if (tsecr_reject) {
+			SKB_DR_SET(*drop_reason, TCP_RFC7323_TSECR);
 			NET_INC_STATS(sock_net(sk), LINUX_MIB_TSECRREJECTED);
+		} else {
+			SKB_DR_SET(*drop_reason, TCP_OVERWINDOW);
+		}
 		return NULL;
 	}
 
@@ -882,6 +887,7 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb,
 	return inet_csk_complete_hashdance(sk, child, req, own_req);
 
 listen_overflow:
+	SKB_DR_SET(*drop_reason, TCP_LISTEN_OVERFLOW);
 	if (sk != req->rsk_listener)
 		__NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPMIGRATEREQFAILURE);
 
-- 
2.48.1.711.g2feabab25a-goog

Re: [PATCH net-next 2/6] tcp: add four drop reasons to tcp_check_req()

[Jakub Kicinski]

On Fri, 28 Feb 2025 13:22:44 +0000 Eric Dumazet wrote:
> +	/** @SKB_DROP_TCP_REASON_LISTEN_OVERFLOW: listener queue full. */
> +	     SKB_DROP_REASON_TCP_LISTEN_OVERFLOW,

nit: TCP_REASON vs REASON_TCP
-- 
pw-bot: cr

Re: [PATCH net-next 2/6] tcp: add four drop reasons to tcp_check_req()

[Eric Dumazet]

On Fri, Feb 28, 2025 at 11:28 PM Jakub Kicinski <kuba@kernel.org> wrote:
>
> On Fri, 28 Feb 2025 13:22:44 +0000 Eric Dumazet wrote:
> > +     /** @SKB_DROP_TCP_REASON_LISTEN_OVERFLOW: listener queue full. */
> > +          SKB_DROP_REASON_TCP_LISTEN_OVERFLOW,
>
> nit: TCP_REASON vs REASON_TCP

Oops, thanks !

[PATCH net-next 3/6] tcp: convert to dev_net_rcu()

[Eric Dumazet]

TCP uses of dev_net() are under RCU protection, change them
to dev_net_rcu() to get LOCKDEP support.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
---
 include/net/inet6_hashtables.h |  2 +-
 include/net/inet_hashtables.h  |  2 +-
 net/ipv4/tcp_ipv4.c            | 12 ++++++------
 net/ipv4/tcp_metrics.c         |  6 +++---
 net/ipv6/tcp_ipv6.c            | 22 +++++++++++-----------
 5 files changed, 22 insertions(+), 22 deletions(-)

diff --git a/include/net/inet6_hashtables.h b/include/net/inet6_hashtables.h
index 74dd90ff5f129fe4c8adad67a642ae5070410518..c32878c69179dac5a7fcfa098a297420d9adfab2 100644
--- a/include/net/inet6_hashtables.h
+++ b/include/net/inet6_hashtables.h
@@ -150,7 +150,7 @@ static inline struct sock *__inet6_lookup_skb(struct inet_hashinfo *hashinfo,
 					      int iif, int sdif,
 					      bool *refcounted)
 {
-	struct net *net = dev_net(skb_dst(skb)->dev);
+	struct net *net = dev_net_rcu(skb_dst(skb)->dev);
 	const struct ipv6hdr *ip6h = ipv6_hdr(skb);
 	struct sock *sk;
 
diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index 5eea47f135a421ce8275d4cd83c5771b3f448e5c..da818fb0205fed6b4120946bc032e67e046b716f 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -492,7 +492,7 @@ static inline struct sock *__inet_lookup_skb(struct inet_hashinfo *hashinfo,
 					     const int sdif,
 					     bool *refcounted)
 {
-	struct net *net = dev_net(skb_dst(skb)->dev);
+	struct net *net = dev_net_rcu(skb_dst(skb)->dev);
 	const struct iphdr *iph = ip_hdr(skb);
 	struct sock *sk;
 
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 218f01a8cc5f6c410043f07293e9e51840c1f1cb..ae07613e4f335063723f49d7fd70a240412922ef 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -494,14 +494,14 @@ int tcp_v4_err(struct sk_buff *skb, u32 info)
 {
 	const struct iphdr *iph = (const struct iphdr *)skb->data;
 	struct tcphdr *th = (struct tcphdr *)(skb->data + (iph->ihl << 2));
-	struct tcp_sock *tp;
+	struct net *net = dev_net_rcu(skb->dev);
 	const int type = icmp_hdr(skb)->type;
 	const int code = icmp_hdr(skb)->code;
-	struct sock *sk;
 	struct request_sock *fastopen;
+	struct tcp_sock *tp;
 	u32 seq, snd_una;
+	struct sock *sk;
 	int err;
-	struct net *net = dev_net(skb->dev);
 
 	sk = __inet_lookup_established(net, net->ipv4.tcp_death_row.hashinfo,
 				       iph->daddr, th->dest, iph->saddr,
@@ -786,7 +786,7 @@ static void tcp_v4_send_reset(const struct sock *sk, struct sk_buff *skb,
 	arg.iov[0].iov_base = (unsigned char *)&rep;
 	arg.iov[0].iov_len  = sizeof(rep.th);
 
-	net = sk ? sock_net(sk) : dev_net(skb_dst(skb)->dev);
+	net = sk ? sock_net(sk) : dev_net_rcu(skb_dst(skb)->dev);
 
 	/* Invalid TCP option size or twice included auth */
 	if (tcp_parse_auth_options(tcp_hdr(skb), &md5_hash_location, &aoh))
@@ -1961,7 +1961,7 @@ EXPORT_SYMBOL(tcp_v4_do_rcv);
 
 int tcp_v4_early_demux(struct sk_buff *skb)
 {
-	struct net *net = dev_net(skb->dev);
+	struct net *net = dev_net_rcu(skb->dev);
 	const struct iphdr *iph;
 	const struct tcphdr *th;
 	struct sock *sk;
@@ -2172,7 +2172,7 @@ static void tcp_v4_fill_cb(struct sk_buff *skb, const struct iphdr *iph,
 
 int tcp_v4_rcv(struct sk_buff *skb)
 {
-	struct net *net = dev_net(skb->dev);
+	struct net *net = dev_net_rcu(skb->dev);
 	enum skb_drop_reason drop_reason;
 	int sdif = inet_sdif(skb);
 	int dif = inet_iif(skb);
diff --git a/net/ipv4/tcp_metrics.c b/net/ipv4/tcp_metrics.c
index 95669935494ef8003a1877e2b86c76bd27307afd..4251670e328c83b55eff7bbda3cc3d97d78563a8 100644
--- a/net/ipv4/tcp_metrics.c
+++ b/net/ipv4/tcp_metrics.c
@@ -170,7 +170,7 @@ static struct tcp_metrics_block *tcpm_new(struct dst_entry *dst,
 	bool reclaim = false;
 
 	spin_lock_bh(&tcp_metrics_lock);
-	net = dev_net(dst->dev);
+	net = dev_net_rcu(dst->dev);
 
 	/* While waiting for the spin-lock the cache might have been populated
 	 * with this entry and so we have to check again.
@@ -273,7 +273,7 @@ static struct tcp_metrics_block *__tcp_get_metrics_req(struct request_sock *req,
 		return NULL;
 	}
 
-	net = dev_net(dst->dev);
+	net = dev_net_rcu(dst->dev);
 	hash ^= net_hash_mix(net);
 	hash = hash_32(hash, tcp_metrics_hash_log);
 
@@ -318,7 +318,7 @@ static struct tcp_metrics_block *tcp_get_metrics(struct sock *sk,
 	else
 		return NULL;
 
-	net = dev_net(dst->dev);
+	net = dev_net_rcu(dst->dev);
 	hash ^= net_hash_mix(net);
 	hash = hash_32(hash, tcp_metrics_hash_log);
 
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index d01088ab80d24eb0f829166faae791221d95bf9e..fe75ad8e606cbca77d69326dc00273e7b214edee 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -376,7 +376,7 @@ static int tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
 {
 	const struct ipv6hdr *hdr = (const struct ipv6hdr *)skb->data;
 	const struct tcphdr *th = (struct tcphdr *)(skb->data+offset);
-	struct net *net = dev_net(skb->dev);
+	struct net *net = dev_net_rcu(skb->dev);
 	struct request_sock *fastopen;
 	struct ipv6_pinfo *np;
 	struct tcp_sock *tp;
@@ -866,16 +866,16 @@ static void tcp_v6_send_response(const struct sock *sk, struct sk_buff *skb, u32
 				 int oif, int rst, u8 tclass, __be32 label,
 				 u32 priority, u32 txhash, struct tcp_key *key)
 {
-	const struct tcphdr *th = tcp_hdr(skb);
-	struct tcphdr *t1;
-	struct sk_buff *buff;
-	struct flowi6 fl6;
-	struct net *net = sk ? sock_net(sk) : dev_net(skb_dst(skb)->dev);
-	struct sock *ctl_sk = net->ipv6.tcp_sk;
+	struct net *net = sk ? sock_net(sk) : dev_net_rcu(skb_dst(skb)->dev);
 	unsigned int tot_len = sizeof(struct tcphdr);
+	struct sock *ctl_sk = net->ipv6.tcp_sk;
+	const struct tcphdr *th = tcp_hdr(skb);
 	__be32 mrst = 0, *topt;
 	struct dst_entry *dst;
-	__u32 mark = 0;
+	struct sk_buff *buff;
+	struct tcphdr *t1;
+	struct flowi6 fl6;
+	u32 mark = 0;
 
 	if (tsecr)
 		tot_len += TCPOLEN_TSTAMP_ALIGNED;
@@ -1041,7 +1041,7 @@ static void tcp_v6_send_reset(const struct sock *sk, struct sk_buff *skb,
 	if (!sk && !ipv6_unicast_destination(skb))
 		return;
 
-	net = sk ? sock_net(sk) : dev_net(skb_dst(skb)->dev);
+	net = sk ? sock_net(sk) : dev_net_rcu(skb_dst(skb)->dev);
 	/* Invalid TCP option size or twice included auth */
 	if (tcp_parse_auth_options(th, &md5_hash_location, &aoh))
 		return;
@@ -1740,6 +1740,7 @@ static void tcp_v6_fill_cb(struct sk_buff *skb, const struct ipv6hdr *hdr,
 
 INDIRECT_CALLABLE_SCOPE int tcp_v6_rcv(struct sk_buff *skb)
 {
+	struct net *net = dev_net_rcu(skb->dev);
 	enum skb_drop_reason drop_reason;
 	int sdif = inet6_sdif(skb);
 	int dif = inet6_iif(skb);
@@ -1749,7 +1750,6 @@ INDIRECT_CALLABLE_SCOPE int tcp_v6_rcv(struct sk_buff *skb)
 	bool refcounted;
 	int ret;
 	u32 isn;
-	struct net *net = dev_net(skb->dev);
 
 	drop_reason = SKB_DROP_REASON_NOT_SPECIFIED;
 	if (skb->pkt_type != PACKET_HOST)
@@ -2001,7 +2001,7 @@ INDIRECT_CALLABLE_SCOPE int tcp_v6_rcv(struct sk_buff *skb)
 
 void tcp_v6_early_demux(struct sk_buff *skb)
 {
-	struct net *net = dev_net(skb->dev);
+	struct net *net = dev_net_rcu(skb->dev);
 	const struct ipv6hdr *hdr;
 	const struct tcphdr *th;
 	struct sock *sk;
-- 
2.48.1.711.g2feabab25a-goog

[PATCH net-next 4/6] net: gro: convert four dev_net() calls

[Eric Dumazet]

tcp4_check_fraglist_gro(), tcp6_check_fraglist_gro(),
udp4_gro_lookup_skb() and udp6_gro_lookup_skb()
assume RCU is held so that the net structure does not disappear.

Use dev_net_rcu() instead of dev_net() to get LOCKDEP support.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
---
 net/ipv4/tcp_offload.c   | 2 +-
 net/ipv4/udp_offload.c   | 2 +-
 net/ipv6/tcpv6_offload.c | 2 +-
 net/ipv6/udp_offload.c   | 2 +-
 4 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/net/ipv4/tcp_offload.c b/net/ipv4/tcp_offload.c
index 2308665b51c5388814e5b61a262a1636d897c4a9..ecef16c58c07146cbeebade0620a5ec7251ddbc5 100644
--- a/net/ipv4/tcp_offload.c
+++ b/net/ipv4/tcp_offload.c
@@ -425,7 +425,7 @@ static void tcp4_check_fraglist_gro(struct list_head *head, struct sk_buff *skb,
 
 	inet_get_iif_sdif(skb, &iif, &sdif);
 	iph = skb_gro_network_header(skb);
-	net = dev_net(skb->dev);
+	net = dev_net_rcu(skb->dev);
 	sk = __inet_lookup_established(net, net->ipv4.tcp_death_row.hashinfo,
 				       iph->saddr, th->source,
 				       iph->daddr, ntohs(th->dest),
diff --git a/net/ipv4/udp_offload.c b/net/ipv4/udp_offload.c
index a5be6e4ed326fbdc6a9b3889db4da903f7f25d37..c1a85b300ee87758ee683a834248a600a3e7f18d 100644
--- a/net/ipv4/udp_offload.c
+++ b/net/ipv4/udp_offload.c
@@ -630,7 +630,7 @@ static struct sock *udp4_gro_lookup_skb(struct sk_buff *skb, __be16 sport,
 					__be16 dport)
 {
 	const struct iphdr *iph = skb_gro_network_header(skb);
-	struct net *net = dev_net(skb->dev);
+	struct net *net = dev_net_rcu(skb->dev);
 	int iif, sdif;
 
 	inet_get_iif_sdif(skb, &iif, &sdif);
diff --git a/net/ipv6/tcpv6_offload.c b/net/ipv6/tcpv6_offload.c
index a45bf17cb2a172d4612cb42f51481b97bbf364cd..91b88daa5b555cb1af591db7680b7d829ce7b1b7 100644
--- a/net/ipv6/tcpv6_offload.c
+++ b/net/ipv6/tcpv6_offload.c
@@ -35,7 +35,7 @@ static void tcp6_check_fraglist_gro(struct list_head *head, struct sk_buff *skb,
 
 	inet6_get_iif_sdif(skb, &iif, &sdif);
 	hdr = skb_gro_network_header(skb);
-	net = dev_net(skb->dev);
+	net = dev_net_rcu(skb->dev);
 	sk = __inet6_lookup_established(net, net->ipv4.tcp_death_row.hashinfo,
 					&hdr->saddr, th->source,
 					&hdr->daddr, ntohs(th->dest),
diff --git a/net/ipv6/udp_offload.c b/net/ipv6/udp_offload.c
index b41152dd424697a9fc3cef13fbb430de49dcb913..404212dfc99abba4d48fc27a574b48ab53731d39 100644
--- a/net/ipv6/udp_offload.c
+++ b/net/ipv6/udp_offload.c
@@ -117,7 +117,7 @@ static struct sock *udp6_gro_lookup_skb(struct sk_buff *skb, __be16 sport,
 					__be16 dport)
 {
 	const struct ipv6hdr *iph = skb_gro_network_header(skb);
-	struct net *net = dev_net(skb->dev);
+	struct net *net = dev_net_rcu(skb->dev);
 	int iif, sdif;
 
 	inet6_get_iif_sdif(skb, &iif, &sdif);
-- 
2.48.1.711.g2feabab25a-goog

[PATCH net-next 5/6] tcp: remove READ_ONCE(req->ts_recent)

[Eric Dumazet]

After commit 8d52da23b6c6 ("tcp: Defer ts_recent changes
until req is owned"), req->ts_recent is not changed anymore.

It is set once in tcp_openreq_init(), bpf_sk_assign_tcp_reqsk()
or cookie_tcp_reqsk_alloc() before the req can be seen by other
cpus/threads.

This completes the revert of eba20811f326 ("tcp: annotate
data-races around tcp_rsk(req)->ts_recent").

Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Wang Hai <wanghai38@huawei.com>
---
 net/ipv4/tcp_ipv4.c      | 2 +-
 net/ipv4/tcp_minisocks.c | 4 ++--
 net/ipv4/tcp_output.c    | 2 +-
 net/ipv6/tcp_ipv6.c      | 2 +-
 4 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index ae07613e4f335063723f49d7fd70a240412922ef..d9405b012dff079f7cafd9d422ff4445a27eb064 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1155,7 +1155,7 @@ static void tcp_v4_reqsk_send_ack(const struct sock *sk, struct sk_buff *skb,
 			tcp_rsk(req)->rcv_nxt,
 			tcp_synack_window(req) >> inet_rsk(req)->rcv_wscale,
 			tcp_rsk_tsval(tcp_rsk(req)),
-			READ_ONCE(req->ts_recent),
+			req->ts_recent,
 			0, &key,
 			inet_rsk(req)->no_srccheck ? IP_REPLY_ARG_NOSRCCHECK : 0,
 			ip_hdr(skb)->tos,
diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c
index ba4a5d7f251d8ed093b38155d9b1a9f50bfcfe32..3cb8f281186b205e2b03d1b78e1750a024b94f6a 100644
--- a/net/ipv4/tcp_minisocks.c
+++ b/net/ipv4/tcp_minisocks.c
@@ -585,7 +585,7 @@ struct sock *tcp_create_openreq_child(const struct sock *sk,
 
 	if (newtp->rx_opt.tstamp_ok) {
 		newtp->tcp_usec_ts = treq->req_usec_ts;
-		newtp->rx_opt.ts_recent = READ_ONCE(req->ts_recent);
+		newtp->rx_opt.ts_recent = req->ts_recent;
 		newtp->rx_opt.ts_recent_stamp = ktime_get_seconds();
 		newtp->tcp_header_len = sizeof(struct tcphdr) + TCPOLEN_TSTAMP_ALIGNED;
 	} else {
@@ -673,7 +673,7 @@ struct sock *tcp_check_req(struct sock *sk, struct sk_buff *skb,
 		tcp_parse_options(sock_net(sk), skb, &tmp_opt, 0, NULL);
 
 		if (tmp_opt.saw_tstamp) {
-			tmp_opt.ts_recent = READ_ONCE(req->ts_recent);
+			tmp_opt.ts_recent = req->ts_recent;
 			if (tmp_opt.rcv_tsecr) {
 				if (inet_rsk(req)->tstamp_ok && !fastopen)
 					tsecr_reject = !between(tmp_opt.rcv_tsecr,
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c
index 0a660075add5bea05a61b4fe2d9d334a89d956a7..24e56bf96747253c1a508ddfe27ebd38da7c219e 100644
--- a/net/ipv4/tcp_output.c
+++ b/net/ipv4/tcp_output.c
@@ -949,7 +949,7 @@ static unsigned int tcp_synack_options(const struct sock *sk,
 			tcp_rsk(req)->snt_tsval_first = opts->tsval;
 		}
 		WRITE_ONCE(tcp_rsk(req)->snt_tsval_last, opts->tsval);
-		opts->tsecr = READ_ONCE(req->ts_recent);
+		opts->tsecr = req->ts_recent;
 		remaining -= TCPOLEN_TSTAMP_ALIGNED;
 	}
 	if (likely(ireq->sack_ok)) {
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index fe75ad8e606cbca77d69326dc00273e7b214edee..85c4820bfe1588e4553784129d13408dea70763a 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -1279,7 +1279,7 @@ static void tcp_v6_reqsk_send_ack(const struct sock *sk, struct sk_buff *skb,
 			tcp_rsk(req)->rcv_nxt,
 			tcp_synack_window(req) >> inet_rsk(req)->rcv_wscale,
 			tcp_rsk_tsval(tcp_rsk(req)),
-			READ_ONCE(req->ts_recent), sk->sk_bound_dev_if,
+			req->ts_recent, sk->sk_bound_dev_if,
 			&key, ipv6_get_dsfield(ipv6_hdr(skb)), 0,
 			READ_ONCE(sk->sk_priority),
 			READ_ONCE(tcp_rsk(req)->txhash));
-- 
2.48.1.711.g2feabab25a-goog

[PATCH net-next 6/6] tcp: tcp_set_window_clamp() cleanup

[Eric Dumazet]

Remove one indentation level.

Use max_t() and clamp() macros.

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 net/ipv4/tcp.c | 36 ++++++++++++++++++------------------
 1 file changed, 18 insertions(+), 18 deletions(-)

diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 1184866922130aff0f4a4e6d5c0d95fd42713b7d..eb5a60c7a9ccdd23fb78a74d614c18c4f7e281c9 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -3693,33 +3693,33 @@ EXPORT_SYMBOL(tcp_sock_set_keepcnt);
 
 int tcp_set_window_clamp(struct sock *sk, int val)
 {
+	u32 old_window_clamp, new_window_clamp;
 	struct tcp_sock *tp = tcp_sk(sk);
 
 	if (!val) {
 		if (sk->sk_state != TCP_CLOSE)
 			return -EINVAL;
 		WRITE_ONCE(tp->window_clamp, 0);
-	} else {
-		u32 new_rcv_ssthresh, old_window_clamp = tp->window_clamp;
-		u32 new_window_clamp = val < SOCK_MIN_RCVBUF / 2 ?
-						SOCK_MIN_RCVBUF / 2 : val;
+		return 0;
+	}
 
-		if (new_window_clamp == old_window_clamp)
-			return 0;
+	old_window_clamp = tp->window_clamp;
+	new_window_clamp = max_t(int, SOCK_MIN_RCVBUF / 2, val);
 
-		WRITE_ONCE(tp->window_clamp, new_window_clamp);
-		if (new_window_clamp < old_window_clamp) {
-			/* need to apply the reserved mem provisioning only
-			 * when shrinking the window clamp
-			 */
-			__tcp_adjust_rcv_ssthresh(sk, tp->window_clamp);
+	if (new_window_clamp == old_window_clamp)
+		return 0;
 
-		} else {
-			new_rcv_ssthresh = min(tp->rcv_wnd, tp->window_clamp);
-			tp->rcv_ssthresh = max(new_rcv_ssthresh,
-					       tp->rcv_ssthresh);
-		}
-	}
+	WRITE_ONCE(tp->window_clamp, new_window_clamp);
+
+	/* Need to apply the reserved mem provisioning only
+	 * when shrinking the window clamp.
+	 */
+	if (new_window_clamp < old_window_clamp)
+		__tcp_adjust_rcv_ssthresh(sk, new_window_clamp);
+	else
+		tp->rcv_ssthresh = clamp(new_window_clamp,
+					 tp->rcv_ssthresh,
+					 tp->rcv_wnd);
 	return 0;
 }
 
-- 
2.48.1.711.g2feabab25a-goog