tcp: tsq: move tsq_flags close to sk_wmem_alloc

tsq_flags being in the same cache line than sk_wmem_alloc
makes a lot of sense. Both fields are changed from tcp_wfree()
and more generally by various TSQ related functions.

Prior patch made room in struct sock and added sk_tsq_flags,
this patch deletes tsq_flags from struct tcp_sock.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

include/linux/tcp.h

@@ -186,7 +186,6 @@ struct tcp_sock {
 	u32	tsoffset;	/* timestamp offset */
 
 	struct list_head tsq_node; /* anchor in tsq_tasklet.head list */
-	unsigned long	tsq_flags;
 
 	/* Data for direct copy to user */
 	struct {

net/ipv4/tcp.c

@@ -663,9 +663,9 @@ static void tcp_push(struct sock *sk, int flags, int mss_now,
 	if (tcp_should_autocork(sk, skb, size_goal)) {
 
 		/* avoid atomic op if TSQ_THROTTLED bit is already set */
-		if (!test_bit(TSQ_THROTTLED, &tp->tsq_flags)) {
+		if (!test_bit(TSQ_THROTTLED, &sk->sk_tsq_flags)) {
 			NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPAUTOCORKING);
-			set_bit(TSQ_THROTTLED, &tp->tsq_flags);
+			set_bit(TSQ_THROTTLED, &sk->sk_tsq_flags);
 		}
 		/* It is possible TX completion already happened
 		 * before we set TSQ_THROTTLED.

net/ipv4/tcp_ipv4.c

@@ -443,7 +443,7 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info)
 			if (!sock_owned_by_user(sk)) {
 				tcp_v4_mtu_reduced(sk);
 			} else {
-				if (!test_and_set_bit(TCP_MTU_REDUCED_DEFERRED, &tp->tsq_flags))
+				if (!test_and_set_bit(TCP_MTU_REDUCED_DEFERRED, &sk->sk_tsq_flags))
 					sock_hold(sk);
 			}
 			goto out;

net/ipv4/tcp_output.c

@@ -767,14 +767,15 @@ static void tcp_tasklet_func(unsigned long data)
 	list_for_each_safe(q, n, &list) {
 		tp = list_entry(q, struct tcp_sock, tsq_node);
 		list_del(&tp->tsq_node);
-		clear_bit(TSQ_QUEUED, &tp->tsq_flags);
 
 		sk = (struct sock *)tp;
+		clear_bit(TSQ_QUEUED, &sk->sk_tsq_flags);
+
 		if (!sk->sk_lock.owned &&
-		    test_bit(TCP_TSQ_DEFERRED, &tp->tsq_flags)) {
+		    test_bit(TCP_TSQ_DEFERRED, &sk->sk_tsq_flags)) {
 			bh_lock_sock(sk);
 			if (!sock_owned_by_user(sk)) {
-				clear_bit(TCP_TSQ_DEFERRED, &tp->tsq_flags);
+				clear_bit(TCP_TSQ_DEFERRED, &sk->sk_tsq_flags);
 				tcp_tsq_handler(sk);
 			}
 			bh_unlock_sock(sk);
@@ -797,16 +798,15 @@ static void tcp_tasklet_func(unsigned long data)
  */
 void tcp_release_cb(struct sock *sk)
 {
-	struct tcp_sock *tp = tcp_sk(sk);
 	unsigned long flags, nflags;
 
 	/* perform an atomic operation only if at least one flag is set */
 	do {
-		flags = tp->tsq_flags;
+		flags = sk->sk_tsq_flags;
 		if (!(flags & TCP_DEFERRED_ALL))
 			return;
 		nflags = flags & ~TCP_DEFERRED_ALL;
-	} while (cmpxchg(&tp->tsq_flags, flags, nflags) != flags);
+	} while (cmpxchg(&sk->sk_tsq_flags, flags, nflags) != flags);
 
 	if (flags & TCPF_TSQ_DEFERRED)
 		tcp_tsq_handler(sk);
@@ -878,7 +878,7 @@ void tcp_wfree(struct sk_buff *skb)
 	if (wmem >= SKB_TRUESIZE(1) && this_cpu_ksoftirqd() == current)
 		goto out;
 
-	for (oval = READ_ONCE(tp->tsq_flags);; oval = nval) {
+	for (oval = READ_ONCE(sk->sk_tsq_flags);; oval = nval) {
 		struct tsq_tasklet *tsq;
 		bool empty;
 
@@ -886,7 +886,7 @@ void tcp_wfree(struct sk_buff *skb)
 			goto out;
 
 		nval = (oval & ~TSQF_THROTTLED) | TSQF_QUEUED | TCPF_TSQ_DEFERRED;
-		nval = cmpxchg(&tp->tsq_flags, oval, nval);
+		nval = cmpxchg(&sk->sk_tsq_flags, oval, nval);
 		if (nval != oval)
 			continue;
 
@@ -2100,7 +2100,7 @@ static bool tcp_small_queue_check(struct sock *sk, const struct sk_buff *skb,
 		    skb->prev == sk->sk_write_queue.next)
 			return false;
 
-		set_bit(TSQ_THROTTLED, &tcp_sk(sk)->tsq_flags);
+		set_bit(TSQ_THROTTLED, &sk->sk_tsq_flags);
 		/* It is possible TX completion already happened
 		 * before we set TSQ_THROTTLED, so we must
 		 * test again the condition.
@@ -2241,8 +2241,8 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle,
 		    unlikely(tso_fragment(sk, skb, limit, mss_now, gfp)))
 			break;
 
-		if (test_bit(TCP_TSQ_DEFERRED, &tp->tsq_flags))
-			clear_bit(TCP_TSQ_DEFERRED, &tp->tsq_flags);
+		if (test_bit(TCP_TSQ_DEFERRED, &sk->sk_tsq_flags))
+			clear_bit(TCP_TSQ_DEFERRED, &sk->sk_tsq_flags);
 		if (tcp_small_queue_check(sk, skb, 0))
 			break;
 
@@ -3545,8 +3545,6 @@ void tcp_send_ack(struct sock *sk)
 	/* We do not want pure acks influencing TCP Small Queues or fq/pacing
 	 * too much.
 	 * SKB_TRUESIZE(max(1 .. 66, MAX_TCP_HEADER)) is unfortunately ~784
-	 * We also avoid tcp_wfree() overhead (cache line miss accessing
-	 * tp->tsq_flags) by using regular sock_wfree()
 	 */
 	skb_set_tcp_pure_ack(buff);
 

net/ipv4/tcp_timer.c

@@ -310,7 +310,7 @@ static void tcp_delack_timer(unsigned long data)
 		inet_csk(sk)->icsk_ack.blocked = 1;
 		__NET_INC_STATS(sock_net(sk), LINUX_MIB_DELAYEDACKLOCKED);
 		/* deleguate our work to tcp_release_cb() */
-		if (!test_and_set_bit(TCP_DELACK_TIMER_DEFERRED, &tcp_sk(sk)->tsq_flags))
+		if (!test_and_set_bit(TCP_DELACK_TIMER_DEFERRED, &sk->sk_tsq_flags))
 			sock_hold(sk);
 	}
 	bh_unlock_sock(sk);
@@ -592,7 +592,7 @@ static void tcp_write_timer(unsigned long data)
 		tcp_write_timer_handler(sk);
 	} else {
 		/* delegate our work to tcp_release_cb() */
-		if (!test_and_set_bit(TCP_WRITE_TIMER_DEFERRED, &tcp_sk(sk)->tsq_flags))
+		if (!test_and_set_bit(TCP_WRITE_TIMER_DEFERRED, &sk->sk_tsq_flags))
 			sock_hold(sk);
 	}
 	bh_unlock_sock(sk);

net/ipv6/tcp_ipv6.c

@@ -399,7 +399,7 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
 		if (!sock_owned_by_user(sk))
 			tcp_v6_mtu_reduced(sk);
 		else if (!test_and_set_bit(TCP_MTU_REDUCED_DEFERRED,
-					   &tp->tsq_flags))
+					   &sk->sk_tsq_flags))
 			sock_hold(sk);
 		goto out;
 	}