[INET]: Remove per bucket rwlock in tcp/dccp ehash table.

As done two years ago on IP route cache table (commit
22c047ccbc68fa8f3fa57f0e8f906479a062c426) , we can avoid using one
lock per hash bucket for the huge TCP/DCCP hash tables.

On a typical x86_64 platform, this saves about 2MB or 4MB of ram, for
litle performance differences. (we hit a different cache line for the
rwlock, but then the bucket cache line have a better sharing factor
among cpus, since we dirty it less often). For netstat or ss commands
that want a full scan of hash table, we perform fewer memory accesses.

Using a 'small' table of hashed rwlocks should be more than enough to
provide correct SMP concurrency between different buckets, without
using too much memory. Sizing of this table depends on
num_possible_cpus() and various CONFIG settings.

This patch provides some locking abstraction that may ease a future
work using a different model for TCP/DCCP table.

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Acked-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 24 insertions, 20 deletions

diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c
index dc429b6b0ba6..b0170732b5e9 100644
--- a/net/ipv4/inet_diag.c
+++ b/net/ipv4/inet_diag.c
@@ -747,13 +747,14 @@ skip_listen_ht:
 
 	for (i = s_i; i < hashinfo->ehash_size; i++) {
 		struct inet_ehash_bucket *head = &hashinfo->ehash[i];
+		rwlock_t *lock = inet_ehash_lockp(hashinfo, i);
 		struct sock *sk;
 		struct hlist_node *node;
 
 		if (i > s_i)
 			s_num = 0;
 
-		read_lock_bh(&head->lock);
+		read_lock_bh(lock);
 		num = 0;
 		sk_for_each(sk, node, &head->chain) {
 			struct inet_sock *inet = inet_sk(sk);
@@ -769,7 +770,7 @@ skip_listen_ht:
 			    r->id.idiag_dport)
 				goto next_normal;
 			if (inet_csk_diag_dump(sk, skb, cb) < 0) {
-				read_unlock_bh(&head->lock);
+				read_unlock_bh(lock);
 				goto done;
 			}
 next_normal:
@@ -791,14 +792,14 @@ next_normal:
 				    r->id.idiag_dport)
 					goto next_dying;
 				if (inet_twsk_diag_dump(tw, skb, cb) < 0) {
-					read_unlock_bh(&head->lock);
+					read_unlock_bh(lock);
 					goto done;
 				}
 next_dying:
 				++num;
 			}
 		}
-		read_unlock_bh(&head->lock);
+		read_unlock_bh(lock);
 	}
 
 done:
diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index 16eecc7046a3..67704da04fc4 100644
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -204,12 +204,13 @@ static int __inet_check_established(struct inet_timewait_death_row *death_row,
 	const __portpair ports = INET_COMBINED_PORTS(inet->dport, lport);
 	unsigned int hash = inet_ehashfn(daddr, lport, saddr, inet->dport);
 	struct inet_ehash_bucket *head = inet_ehash_bucket(hinfo, hash);
+	rwlock_t *lock = inet_ehash_lockp(hinfo, hash);
 	struct sock *sk2;
 	const struct hlist_node *node;
 	struct inet_timewait_sock *tw;
 
 	prefetch(head->chain.first);
-	write_lock(&head->lock);
+	write_lock(lock);
 
 	/* Check TIME-WAIT sockets first. */
 	sk_for_each(sk2, node, &head->twchain) {
@@ -239,7 +240,7 @@ unique:
 	BUG_TRAP(sk_unhashed(sk));
 	__sk_add_node(sk, &head->chain);
 	sock_prot_inc_use(sk->sk_prot);
-	write_unlock(&head->lock);
+	write_unlock(lock);
 
 	if (twp) {
 		*twp = tw;
@@ -255,7 +256,7 @@ unique:
 	return 0;
 
 not_unique:
-	write_unlock(&head->lock);
+	write_unlock(lock);
 	return -EADDRNOTAVAIL;
 }
 
diff --git a/net/ipv4/inet_timewait_sock.c b/net/ipv4/inet_timewait_sock.c
index 4e189e28f306..a60b99e0ebdc 100644
--- a/net/ipv4/inet_timewait_sock.c
+++ b/net/ipv4/inet_timewait_sock.c
@@ -20,16 +20,16 @@ static void __inet_twsk_kill(struct inet_timewait_sock *tw,
 	struct inet_bind_hashbucket *bhead;
 	struct inet_bind_bucket *tb;
 	/* Unlink from established hashes. */
-	struct inet_ehash_bucket *ehead = inet_ehash_bucket(hashinfo, tw->tw_hash);
+	rwlock_t *lock = inet_ehash_lockp(hashinfo, tw->tw_hash);
 
-	write_lock(&ehead->lock);
+	write_lock(lock);
 	if (hlist_unhashed(&tw->tw_node)) {
-		write_unlock(&ehead->lock);
+		write_unlock(lock);
 		return;
 	}
 	__hlist_del(&tw->tw_node);
 	sk_node_init(&tw->tw_node);
-	write_unlock(&ehead->lock);
+	write_unlock(lock);
 
 	/* Disassociate with bind bucket. */
 	bhead = &hashinfo->bhash[inet_bhashfn(tw->tw_num, hashinfo->bhash_size)];
@@ -59,6 +59,7 @@ void __inet_twsk_hashdance(struct inet_timewait_sock *tw, struct sock *sk,
 	const struct inet_sock *inet = inet_sk(sk);
 	const struct inet_connection_sock *icsk = inet_csk(sk);
 	struct inet_ehash_bucket *ehead = inet_ehash_bucket(hashinfo, sk->sk_hash);
+	rwlock_t *lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
 	struct inet_bind_hashbucket *bhead;
 	/* Step 1: Put TW into bind hash. Original socket stays there too.
 	   Note, that any socket with inet->num != 0 MUST be bound in
@@ -71,7 +72,7 @@ void __inet_twsk_hashdance(struct inet_timewait_sock *tw, struct sock *sk,
 	inet_twsk_add_bind_node(tw, &tw->tw_tb->owners);
 	spin_unlock(&bhead->lock);
 
-	write_lock(&ehead->lock);
+	write_lock(lock);
 
 	/* Step 2: Remove SK from established hash. */
 	if (__sk_del_node_init(sk))
@@ -81,7 +82,7 @@ void __inet_twsk_hashdance(struct inet_timewait_sock *tw, struct sock *sk,
 	inet_twsk_add_node(tw, &ehead->twchain);
 	atomic_inc(&tw->tw_refcnt);
 
-	write_unlock(&ehead->lock);
+	write_unlock(lock);
 }
 
 EXPORT_SYMBOL_GPL(__inet_twsk_hashdance);
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index c64072bb504b..8e65182f7af1 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2456,11 +2456,11 @@ void __init tcp_init(void)
 					thash_entries ? 0 : 512 * 1024);
 	tcp_hashinfo.ehash_size = 1 << tcp_hashinfo.ehash_size;
 	for (i = 0; i < tcp_hashinfo.ehash_size; i++) {
-		rwlock_init(&tcp_hashinfo.ehash[i].lock);
 		INIT_HLIST_HEAD(&tcp_hashinfo.ehash[i].chain);
 		INIT_HLIST_HEAD(&tcp_hashinfo.ehash[i].twchain);
 	}
-
+	if (inet_ehash_locks_alloc(&tcp_hashinfo))
+		panic("TCP: failed to alloc ehash_locks");
 	tcp_hashinfo.bhash =
 		alloc_large_system_hash("TCP bind",
 					sizeof(struct inet_bind_hashbucket),
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index e9127cdced20..e566f3c67677 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -2049,8 +2049,9 @@ static void *established_get_first(struct seq_file *seq)
 		struct sock *sk;
 		struct hlist_node *node;
 		struct inet_timewait_sock *tw;
+		rwlock_t *lock = inet_ehash_lockp(&tcp_hashinfo, st->bucket);
 
-		read_lock_bh(&tcp_hashinfo.ehash[st->bucket].lock);
+		read_lock_bh(lock);
 		sk_for_each(sk, node, &tcp_hashinfo.ehash[st->bucket].chain) {
 			if (sk->sk_family != st->family) {
 				continue;
@@ -2067,7 +2068,7 @@ static void *established_get_first(struct seq_file *seq)
 			rc = tw;
 			goto out;
 		}
-		read_unlock_bh(&tcp_hashinfo.ehash[st->bucket].lock);
+		read_unlock_bh(lock);
 		st->state = TCP_SEQ_STATE_ESTABLISHED;
 	}
 out:
@@ -2094,11 +2095,11 @@ get_tw:
 			cur = tw;
 			goto out;
 		}
-		read_unlock_bh(&tcp_hashinfo.ehash[st->bucket].lock);
+		read_unlock_bh(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
 		st->state = TCP_SEQ_STATE_ESTABLISHED;
 
 		if (++st->bucket < tcp_hashinfo.ehash_size) {
-			read_lock_bh(&tcp_hashinfo.ehash[st->bucket].lock);
+			read_lock_bh(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
 			sk = sk_head(&tcp_hashinfo.ehash[st->bucket].chain);
 		} else {
 			cur = NULL;
@@ -2206,7 +2207,7 @@ static void tcp_seq_stop(struct seq_file *seq, void *v)
 	case TCP_SEQ_STATE_TIME_WAIT:
 	case TCP_SEQ_STATE_ESTABLISHED:
 		if (v)
-			read_unlock_bh(&tcp_hashinfo.ehash[st->bucket].lock);
+			read_unlock_bh(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
 		break;
 	}
 }