From 7a9546ee354ec6f23af403992b8c07baa50a23d2 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <dada1@cosmosbay.com>
Date: Wed, 12 Nov 2008 00:54:20 -0800
Subject: net: ib_net pointer should depends on CONFIG_NET_NS

We can shrink size of "struct inet_bind_bucket" by 50%, using
read_pnet() and write_pnet()

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/inet_hashtables.h | 7 +++++++
 1 file changed, 7 insertions(+)

(limited to 'include/net/inet_hashtables.h')

diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index 5cc182f9eca..cb31fbf8ae2 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -77,13 +77,20 @@ struct inet_ehash_bucket {
  * ports are created in O(1) time?  I thought so. ;-)	-DaveM
  */
 struct inet_bind_bucket {
+#ifdef CONFIG_NET_NS
 	struct net		*ib_net;
+#endif
 	unsigned short		port;
 	signed short		fastreuse;
 	struct hlist_node	node;
 	struct hlist_head	owners;
 };
 
+static inline struct net *ib_net(struct inet_bind_bucket *ib)
+{
+	return read_pnet(&ib->ib_net);
+}
+
 #define inet_bind_bucket_for_each(tb, node, head) \
 	hlist_for_each_entry(tb, node, head, node)
 
-- 
cgit v1.2.3-70-g09d2


From 3ab5aee7fe840b5b1b35a8d1ac11c3de5281e611 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <dada1@cosmosbay.com>
Date: Sun, 16 Nov 2008 19:40:17 -0800
Subject: net: Convert TCP & DCCP hash tables to use RCU / hlist_nulls

RCU was added to UDP lookups, using a fast infrastructure :
- sockets kmem_cache use SLAB_DESTROY_BY_RCU and dont pay the
  price of call_rcu() at freeing time.
- hlist_nulls permits to use few memory barriers.

This patch uses same infrastructure for TCP/DCCP established
and timewait sockets.

Thanks to SLAB_DESTROY_BY_RCU, no slowdown for applications
using short lived TCP connections. A followup patch, converting
rwlocks to spinlocks will even speedup this case.

__inet_lookup_established() is pretty fast now we dont have to
dirty a contended cache line (read_lock/read_unlock)

Only established and timewait hashtable are converted to RCU
(bind table and listen table are still using traditional locking)

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/inet_hashtables.h    |  4 +--
 include/net/inet_timewait_sock.h | 10 +++---
 net/core/sock.c                  |  4 ++-
 net/dccp/ipv4.c                  |  1 +
 net/dccp/ipv6.c                  |  1 +
 net/dccp/proto.c                 |  4 +--
 net/ipv4/inet_diag.c             |  7 ++--
 net/ipv4/inet_hashtables.c       | 78 ++++++++++++++++++++++++++++------------
 net/ipv4/inet_timewait_sock.c    | 26 ++++++++------
 net/ipv4/tcp.c                   |  4 +--
 net/ipv4/tcp_ipv4.c              | 25 ++++++-------
 net/ipv6/inet6_hashtables.c      | 70 ++++++++++++++++++++++++------------
 net/ipv6/tcp_ipv6.c              |  1 +
 13 files changed, 151 insertions(+), 84 deletions(-)

(limited to 'include/net/inet_hashtables.h')

diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index cb31fbf8ae2..48189604511 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -41,8 +41,8 @@
  * I'll experiment with dynamic table growth later.
  */
 struct inet_ehash_bucket {
-	struct hlist_head chain;
-	struct hlist_head twchain;
+	struct hlist_nulls_head chain;
+	struct hlist_nulls_head twchain;
 };
 
 /* There are a few simple rules, which allow for local port reuse by
diff --git a/include/net/inet_timewait_sock.h b/include/net/inet_timewait_sock.h
index 80e4977631b..4b8ece22b8e 100644
--- a/include/net/inet_timewait_sock.h
+++ b/include/net/inet_timewait_sock.h
@@ -110,7 +110,7 @@ struct inet_timewait_sock {
 #define tw_state		__tw_common.skc_state
 #define tw_reuse		__tw_common.skc_reuse
 #define tw_bound_dev_if		__tw_common.skc_bound_dev_if
-#define tw_node			__tw_common.skc_node
+#define tw_node			__tw_common.skc_nulls_node
 #define tw_bind_node		__tw_common.skc_bind_node
 #define tw_refcnt		__tw_common.skc_refcnt
 #define tw_hash			__tw_common.skc_hash
@@ -137,10 +137,10 @@ struct inet_timewait_sock {
 	struct hlist_node	tw_death_node;
 };
 
-static inline void inet_twsk_add_node(struct inet_timewait_sock *tw,
-				      struct hlist_head *list)
+static inline void inet_twsk_add_node_rcu(struct inet_timewait_sock *tw,
+				      struct hlist_nulls_head *list)
 {
-	hlist_add_head(&tw->tw_node, list);
+	hlist_nulls_add_head_rcu(&tw->tw_node, list);
 }
 
 static inline void inet_twsk_add_bind_node(struct inet_timewait_sock *tw,
@@ -175,7 +175,7 @@ static inline int inet_twsk_del_dead_node(struct inet_timewait_sock *tw)
 }
 
 #define inet_twsk_for_each(tw, node, head) \
-	hlist_for_each_entry(tw, node, head, tw_node)
+	hlist_nulls_for_each_entry(tw, node, head, tw_node)
 
 #define inet_twsk_for_each_inmate(tw, node, jail) \
 	hlist_for_each_entry(tw, node, jail, tw_death_node)
diff --git a/net/core/sock.c b/net/core/sock.c
index ded1eb5d2fd..38de9c3f563 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -2082,7 +2082,9 @@ int proto_register(struct proto *prot, int alloc_slab)
 			prot->twsk_prot->twsk_slab =
 				kmem_cache_create(timewait_sock_slab_name,
 						  prot->twsk_prot->twsk_obj_size,
-						  0, SLAB_HWCACHE_ALIGN,
+						  0,
+						  SLAB_HWCACHE_ALIGN |
+							prot->slab_flags,
 						  NULL);
 			if (prot->twsk_prot->twsk_slab == NULL)
 				goto out_free_timewait_sock_slab_name;
diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c
index 528baa2e5be..d1dd95289b8 100644
--- a/net/dccp/ipv4.c
+++ b/net/dccp/ipv4.c
@@ -938,6 +938,7 @@ static struct proto dccp_v4_prot = {
 	.orphan_count		= &dccp_orphan_count,
 	.max_header		= MAX_DCCP_HEADER,
 	.obj_size		= sizeof(struct dccp_sock),
+	.slab_flags		= SLAB_DESTROY_BY_RCU,
 	.rsk_prot		= &dccp_request_sock_ops,
 	.twsk_prot		= &dccp_timewait_sock_ops,
 	.h.hashinfo		= &dccp_hashinfo,
diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c
index 4aa1148cdb2..f033e845bb0 100644
--- a/net/dccp/ipv6.c
+++ b/net/dccp/ipv6.c
@@ -1140,6 +1140,7 @@ static struct proto dccp_v6_prot = {
 	.orphan_count	   = &dccp_orphan_count,
 	.max_header	   = MAX_DCCP_HEADER,
 	.obj_size	   = sizeof(struct dccp6_sock),
+	.slab_flags	   = SLAB_DESTROY_BY_RCU,
 	.rsk_prot	   = &dccp6_request_sock_ops,
 	.twsk_prot	   = &dccp6_timewait_sock_ops,
 	.h.hashinfo	   = &dccp_hashinfo,
diff --git a/net/dccp/proto.c b/net/dccp/proto.c
index 46cb3490d48..1117d4d8c8f 100644
--- a/net/dccp/proto.c
+++ b/net/dccp/proto.c
@@ -1090,8 +1090,8 @@ static int __init dccp_init(void)
 	}
 
 	for (i = 0; i < dccp_hashinfo.ehash_size; i++) {
-		INIT_HLIST_HEAD(&dccp_hashinfo.ehash[i].chain);
-		INIT_HLIST_HEAD(&dccp_hashinfo.ehash[i].twchain);
+		INIT_HLIST_NULLS_HEAD(&dccp_hashinfo.ehash[i].chain, i);
+		INIT_HLIST_NULLS_HEAD(&dccp_hashinfo.ehash[i].twchain, i);
 	}
 
 	if (inet_ehash_locks_alloc(&dccp_hashinfo))
diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c
index 564230dabcb..41b36720e97 100644
--- a/net/ipv4/inet_diag.c
+++ b/net/ipv4/inet_diag.c
@@ -778,18 +778,19 @@ skip_listen_ht:
 		struct inet_ehash_bucket *head = &hashinfo->ehash[i];
 		rwlock_t *lock = inet_ehash_lockp(hashinfo, i);
 		struct sock *sk;
-		struct hlist_node *node;
+		struct hlist_nulls_node *node;
 
 		num = 0;
 
-		if (hlist_empty(&head->chain) && hlist_empty(&head->twchain))
+		if (hlist_nulls_empty(&head->chain) &&
+			hlist_nulls_empty(&head->twchain))
 			continue;
 
 		if (i > s_i)
 			s_num = 0;
 
 		read_lock_bh(lock);
-		sk_for_each(sk, node, &head->chain) {
+		sk_nulls_for_each(sk, node, &head->chain) {
 			struct inet_sock *inet = inet_sk(sk);
 
 			if (num < s_num)
diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index be41ebbec4e..fd269cfef0e 100644
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -223,35 +223,65 @@ struct sock * __inet_lookup_established(struct net *net,
 	INET_ADDR_COOKIE(acookie, saddr, daddr)
 	const __portpair ports = INET_COMBINED_PORTS(sport, hnum);
 	struct sock *sk;
-	const struct hlist_node *node;
+	const struct hlist_nulls_node *node;
 	/* Optimize here for direct hit, only listening connections can
 	 * have wildcards anyways.
 	 */
 	unsigned int hash = inet_ehashfn(net, daddr, hnum, saddr, sport);
-	struct inet_ehash_bucket *head = inet_ehash_bucket(hashinfo, hash);
-	rwlock_t *lock = inet_ehash_lockp(hashinfo, hash);
+	unsigned int slot = hash & (hashinfo->ehash_size - 1);
+	struct inet_ehash_bucket *head = &hashinfo->ehash[slot];
 
-	prefetch(head->chain.first);
-	read_lock(lock);
-	sk_for_each(sk, node, &head->chain) {
+	rcu_read_lock();
+begin:
+	sk_nulls_for_each_rcu(sk, node, &head->chain) {
 		if (INET_MATCH(sk, net, hash, acookie,
-					saddr, daddr, ports, dif))
-			goto hit; /* You sunk my battleship! */
+					saddr, daddr, ports, dif)) {
+			if (unlikely(!atomic_inc_not_zero(&sk->sk_refcnt)))
+				goto begintw;
+			if (unlikely(!INET_MATCH(sk, net, hash, acookie,
+				saddr, daddr, ports, dif))) {
+				sock_put(sk);
+				goto begin;
+			}
+			goto out;
+		}
 	}
+	/*
+	 * if the nulls value we got at the end of this lookup is
+	 * not the expected one, we must restart lookup.
+	 * We probably met an item that was moved to another chain.
+	 */
+	if (get_nulls_value(node) != slot)
+		goto begin;
 
+begintw:
 	/* Must check for a TIME_WAIT'er before going to listener hash. */
-	sk_for_each(sk, node, &head->twchain) {
+	sk_nulls_for_each_rcu(sk, node, &head->twchain) {
 		if (INET_TW_MATCH(sk, net, hash, acookie,
-					saddr, daddr, ports, dif))
-			goto hit;
+					saddr, daddr, ports, dif)) {
+			if (unlikely(!atomic_inc_not_zero(&sk->sk_refcnt))) {
+				sk = NULL;
+				goto out;
+			}
+			if (unlikely(!INET_TW_MATCH(sk, net, hash, acookie,
+				 saddr, daddr, ports, dif))) {
+				sock_put(sk);
+				goto begintw;
+			}
+			goto out;
+		}
 	}
+	/*
+	 * if the nulls value we got at the end of this lookup is
+	 * not the expected one, we must restart lookup.
+	 * We probably met an item that was moved to another chain.
+	 */
+	if (get_nulls_value(node) != slot)
+		goto begintw;
 	sk = NULL;
 out:
-	read_unlock(lock);
+	rcu_read_unlock();
 	return sk;
-hit:
-	sock_hold(sk);
-	goto out;
 }
 EXPORT_SYMBOL_GPL(__inet_lookup_established);
 
@@ -272,14 +302,14 @@ static int __inet_check_established(struct inet_timewait_death_row *death_row,
 	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;
+	const struct hlist_nulls_node *node;
 	struct inet_timewait_sock *tw;
 
 	prefetch(head->chain.first);
 	write_lock(lock);
 
 	/* Check TIME-WAIT sockets first. */
-	sk_for_each(sk2, node, &head->twchain) {
+	sk_nulls_for_each(sk2, node, &head->twchain) {
 		tw = inet_twsk(sk2);
 
 		if (INET_TW_MATCH(sk2, net, hash, acookie,
@@ -293,7 +323,7 @@ static int __inet_check_established(struct inet_timewait_death_row *death_row,
 	tw = NULL;
 
 	/* And established part... */
-	sk_for_each(sk2, node, &head->chain) {
+	sk_nulls_for_each(sk2, node, &head->chain) {
 		if (INET_MATCH(sk2, net, hash, acookie,
 					saddr, daddr, ports, dif))
 			goto not_unique;
@@ -306,7 +336,7 @@ unique:
 	inet->sport = htons(lport);
 	sk->sk_hash = hash;
 	WARN_ON(!sk_unhashed(sk));
-	__sk_add_node(sk, &head->chain);
+	__sk_nulls_add_node_rcu(sk, &head->chain);
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
 	write_unlock(lock);
 
@@ -338,7 +368,7 @@ static inline u32 inet_sk_port_offset(const struct sock *sk)
 void __inet_hash_nolisten(struct sock *sk)
 {
 	struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo;
-	struct hlist_head *list;
+	struct hlist_nulls_head *list;
 	rwlock_t *lock;
 	struct inet_ehash_bucket *head;
 
@@ -350,7 +380,7 @@ void __inet_hash_nolisten(struct sock *sk)
 	lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
 
 	write_lock(lock);
-	__sk_add_node(sk, list);
+	__sk_nulls_add_node_rcu(sk, list);
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
 	write_unlock(lock);
 }
@@ -400,13 +430,15 @@ void inet_unhash(struct sock *sk)
 		local_bh_disable();
 		inet_listen_wlock(hashinfo);
 		lock = &hashinfo->lhash_lock;
+		if (__sk_del_node_init(sk))
+			sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
 	} else {
 		lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
 		write_lock_bh(lock);
+		if (__sk_nulls_del_node_init_rcu(sk))
+			sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
 	}
 
-	if (__sk_del_node_init(sk))
-		sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
 	write_unlock_bh(lock);
 out:
 	if (sk->sk_state == TCP_LISTEN)
diff --git a/net/ipv4/inet_timewait_sock.c b/net/ipv4/inet_timewait_sock.c
index 1c5fd38f882..60689951ecd 100644
--- a/net/ipv4/inet_timewait_sock.c
+++ b/net/ipv4/inet_timewait_sock.c
@@ -23,12 +23,12 @@ static void __inet_twsk_kill(struct inet_timewait_sock *tw,
 	rwlock_t *lock = inet_ehash_lockp(hashinfo, tw->tw_hash);
 
 	write_lock(lock);
-	if (hlist_unhashed(&tw->tw_node)) {
+	if (hlist_nulls_unhashed(&tw->tw_node)) {
 		write_unlock(lock);
 		return;
 	}
-	__hlist_del(&tw->tw_node);
-	sk_node_init(&tw->tw_node);
+	hlist_nulls_del_rcu(&tw->tw_node);
+	sk_nulls_node_init(&tw->tw_node);
 	write_unlock(lock);
 
 	/* Disassociate with bind bucket. */
@@ -92,13 +92,17 @@ void __inet_twsk_hashdance(struct inet_timewait_sock *tw, struct sock *sk,
 
 	write_lock(lock);
 
-	/* Step 2: Remove SK from established hash. */
-	if (__sk_del_node_init(sk))
-		sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
-
-	/* Step 3: Hash TW into TIMEWAIT chain. */
-	inet_twsk_add_node(tw, &ehead->twchain);
+	/*
+	 * Step 2: Hash TW into TIMEWAIT chain.
+	 * Should be done before removing sk from established chain
+	 * because readers are lockless and search established first.
+	 */
 	atomic_inc(&tw->tw_refcnt);
+	inet_twsk_add_node_rcu(tw, &ehead->twchain);
+
+	/* Step 3: Remove SK from established hash. */
+	if (__sk_nulls_del_node_init_rcu(sk))
+		sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
 
 	write_unlock(lock);
 }
@@ -416,7 +420,7 @@ void inet_twsk_purge(struct net *net, struct inet_hashinfo *hashinfo,
 {
 	struct inet_timewait_sock *tw;
 	struct sock *sk;
-	struct hlist_node *node;
+	struct hlist_nulls_node *node;
 	int h;
 
 	local_bh_disable();
@@ -426,7 +430,7 @@ void inet_twsk_purge(struct net *net, struct inet_hashinfo *hashinfo,
 		rwlock_t *lock = inet_ehash_lockp(hashinfo, h);
 restart:
 		write_lock(lock);
-		sk_for_each(sk, node, &head->twchain) {
+		sk_nulls_for_each(sk, node, &head->twchain) {
 
 			tw = inet_twsk(sk);
 			if (!net_eq(twsk_net(tw), net) ||
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index f60a5917e54..044224a341e 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -2707,8 +2707,8 @@ 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++) {
-		INIT_HLIST_HEAD(&tcp_hashinfo.ehash[i].chain);
-		INIT_HLIST_HEAD(&tcp_hashinfo.ehash[i].twchain);
+		INIT_HLIST_NULLS_HEAD(&tcp_hashinfo.ehash[i].chain, i);
+		INIT_HLIST_NULLS_HEAD(&tcp_hashinfo.ehash[i].twchain, i);
 	}
 	if (inet_ehash_locks_alloc(&tcp_hashinfo))
 		panic("TCP: failed to alloc ehash_locks");
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index d49233f409b..b2e3ab2287b 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1857,16 +1857,16 @@ EXPORT_SYMBOL(tcp_v4_destroy_sock);
 #ifdef CONFIG_PROC_FS
 /* Proc filesystem TCP sock list dumping. */
 
-static inline struct inet_timewait_sock *tw_head(struct hlist_head *head)
+static inline struct inet_timewait_sock *tw_head(struct hlist_nulls_head *head)
 {
-	return hlist_empty(head) ? NULL :
+	return hlist_nulls_empty(head) ? NULL :
 		list_entry(head->first, struct inet_timewait_sock, tw_node);
 }
 
 static inline struct inet_timewait_sock *tw_next(struct inet_timewait_sock *tw)
 {
-	return tw->tw_node.next ?
-		hlist_entry(tw->tw_node.next, typeof(*tw), tw_node) : NULL;
+	return !is_a_nulls(tw->tw_node.next) ?
+		hlist_nulls_entry(tw->tw_node.next, typeof(*tw), tw_node) : NULL;
 }
 
 static void *listening_get_next(struct seq_file *seq, void *cur)
@@ -1954,8 +1954,8 @@ static void *listening_get_idx(struct seq_file *seq, loff_t *pos)
 
 static inline int empty_bucket(struct tcp_iter_state *st)
 {
-	return hlist_empty(&tcp_hashinfo.ehash[st->bucket].chain) &&
-		hlist_empty(&tcp_hashinfo.ehash[st->bucket].twchain);
+	return hlist_nulls_empty(&tcp_hashinfo.ehash[st->bucket].chain) &&
+		hlist_nulls_empty(&tcp_hashinfo.ehash[st->bucket].twchain);
 }
 
 static void *established_get_first(struct seq_file *seq)
@@ -1966,7 +1966,7 @@ static void *established_get_first(struct seq_file *seq)
 
 	for (st->bucket = 0; st->bucket < tcp_hashinfo.ehash_size; ++st->bucket) {
 		struct sock *sk;
-		struct hlist_node *node;
+		struct hlist_nulls_node *node;
 		struct inet_timewait_sock *tw;
 		rwlock_t *lock = inet_ehash_lockp(&tcp_hashinfo, st->bucket);
 
@@ -1975,7 +1975,7 @@ static void *established_get_first(struct seq_file *seq)
 			continue;
 
 		read_lock_bh(lock);
-		sk_for_each(sk, node, &tcp_hashinfo.ehash[st->bucket].chain) {
+		sk_nulls_for_each(sk, node, &tcp_hashinfo.ehash[st->bucket].chain) {
 			if (sk->sk_family != st->family ||
 			    !net_eq(sock_net(sk), net)) {
 				continue;
@@ -2004,7 +2004,7 @@ static void *established_get_next(struct seq_file *seq, void *cur)
 {
 	struct sock *sk = cur;
 	struct inet_timewait_sock *tw;
-	struct hlist_node *node;
+	struct hlist_nulls_node *node;
 	struct tcp_iter_state *st = seq->private;
 	struct net *net = seq_file_net(seq);
 
@@ -2032,11 +2032,11 @@ get_tw:
 			return NULL;
 
 		read_lock_bh(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
-		sk = sk_head(&tcp_hashinfo.ehash[st->bucket].chain);
+		sk = sk_nulls_head(&tcp_hashinfo.ehash[st->bucket].chain);
 	} else
-		sk = sk_next(sk);
+		sk = sk_nulls_next(sk);
 
-	sk_for_each_from(sk, node) {
+	sk_nulls_for_each_from(sk, node) {
 		if (sk->sk_family == st->family && net_eq(sock_net(sk), net))
 			goto found;
 	}
@@ -2375,6 +2375,7 @@ struct proto tcp_prot = {
 	.sysctl_rmem		= sysctl_tcp_rmem,
 	.max_header		= MAX_TCP_HEADER,
 	.obj_size		= sizeof(struct tcp_sock),
+	.slab_flags		= SLAB_DESTROY_BY_RCU,
 	.twsk_prot		= &tcp_timewait_sock_ops,
 	.rsk_prot		= &tcp_request_sock_ops,
 	.h.hashinfo		= &tcp_hashinfo,
diff --git a/net/ipv6/inet6_hashtables.c b/net/ipv6/inet6_hashtables.c
index 1646a565825..c1b4d401fd9 100644
--- a/net/ipv6/inet6_hashtables.c
+++ b/net/ipv6/inet6_hashtables.c
@@ -25,24 +25,28 @@
 void __inet6_hash(struct sock *sk)
 {
 	struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo;
-	struct hlist_head *list;
 	rwlock_t *lock;
 
 	WARN_ON(!sk_unhashed(sk));
 
 	if (sk->sk_state == TCP_LISTEN) {
+		struct hlist_head *list;
+
 		list = &hashinfo->listening_hash[inet_sk_listen_hashfn(sk)];
 		lock = &hashinfo->lhash_lock;
 		inet_listen_wlock(hashinfo);
+		__sk_add_node(sk, list);
 	} else {
 		unsigned int hash;
+		struct hlist_nulls_head *list;
+
 		sk->sk_hash = hash = inet6_sk_ehashfn(sk);
 		list = &inet_ehash_bucket(hashinfo, hash)->chain;
 		lock = inet_ehash_lockp(hashinfo, hash);
 		write_lock(lock);
+		__sk_nulls_add_node_rcu(sk, list);
 	}
 
-	__sk_add_node(sk, list);
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
 	write_unlock(lock);
 }
@@ -63,33 +67,53 @@ struct sock *__inet6_lookup_established(struct net *net,
 					   const int dif)
 {
 	struct sock *sk;
-	const struct hlist_node *node;
+	const struct hlist_nulls_node *node;
 	const __portpair ports = INET_COMBINED_PORTS(sport, hnum);
 	/* Optimize here for direct hit, only listening connections can
 	 * have wildcards anyways.
 	 */
 	unsigned int hash = inet6_ehashfn(net, daddr, hnum, saddr, sport);
-	struct inet_ehash_bucket *head = inet_ehash_bucket(hashinfo, hash);
-	rwlock_t *lock = inet_ehash_lockp(hashinfo, hash);
+	unsigned int slot = hash & (hashinfo->ehash_size - 1);
+	struct inet_ehash_bucket *head = &hashinfo->ehash[slot];
 
-	prefetch(head->chain.first);
-	read_lock(lock);
-	sk_for_each(sk, node, &head->chain) {
+
+	rcu_read_lock();
+begin:
+	sk_nulls_for_each_rcu(sk, node, &head->chain) {
 		/* For IPV6 do the cheaper port and family tests first. */
-		if (INET6_MATCH(sk, net, hash, saddr, daddr, ports, dif))
-			goto hit; /* You sunk my battleship! */
+		if (INET6_MATCH(sk, net, hash, saddr, daddr, ports, dif)) {
+			if (unlikely(!atomic_inc_not_zero(&sk->sk_refcnt)))
+				goto begintw;
+			if (!INET6_MATCH(sk, net, hash, saddr, daddr, ports, dif)) {
+				sock_put(sk);
+				goto begin;
+			}
+		goto out;
+		}
 	}
+	if (get_nulls_value(node) != slot)
+		goto begin;
+
+begintw:
 	/* Must check for a TIME_WAIT'er before going to listener hash. */
-	sk_for_each(sk, node, &head->twchain) {
-		if (INET6_TW_MATCH(sk, net, hash, saddr, daddr, ports, dif))
-			goto hit;
+	sk_nulls_for_each_rcu(sk, node, &head->twchain) {
+		if (INET6_TW_MATCH(sk, net, hash, saddr, daddr, ports, dif)) {
+			if (unlikely(!atomic_inc_not_zero(&sk->sk_refcnt))) {
+				sk = NULL;
+				goto out;
+			}
+			if (!INET6_TW_MATCH(sk, net, hash, saddr, daddr, ports, dif)) {
+				sock_put(sk);
+				goto begintw;
+			}
+			goto out;
+		}
 	}
-	read_unlock(lock);
-	return NULL;
-
-hit:
-	sock_hold(sk);
-	read_unlock(lock);
+	if (get_nulls_value(node) != slot)
+		goto begintw;
+	sk = NULL;
+out:
+	rcu_read_unlock();
 	return sk;
 }
 EXPORT_SYMBOL(__inet6_lookup_established);
@@ -172,14 +196,14 @@ static int __inet6_check_established(struct inet_timewait_death_row *death_row,
 	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;
+	const struct hlist_nulls_node *node;
 	struct inet_timewait_sock *tw;
 
 	prefetch(head->chain.first);
 	write_lock(lock);
 
 	/* Check TIME-WAIT sockets first. */
-	sk_for_each(sk2, node, &head->twchain) {
+	sk_nulls_for_each(sk2, node, &head->twchain) {
 		tw = inet_twsk(sk2);
 
 		if (INET6_TW_MATCH(sk2, net, hash, saddr, daddr, ports, dif)) {
@@ -192,7 +216,7 @@ static int __inet6_check_established(struct inet_timewait_death_row *death_row,
 	tw = NULL;
 
 	/* And established part... */
-	sk_for_each(sk2, node, &head->chain) {
+	sk_nulls_for_each(sk2, node, &head->chain) {
 		if (INET6_MATCH(sk2, net, hash, saddr, daddr, ports, dif))
 			goto not_unique;
 	}
@@ -203,7 +227,7 @@ unique:
 	inet->num = lport;
 	inet->sport = htons(lport);
 	WARN_ON(!sk_unhashed(sk));
-	__sk_add_node(sk, &head->chain);
+	__sk_nulls_add_node_rcu(sk, &head->chain);
 	sk->sk_hash = hash;
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
 	write_unlock(lock);
diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 984276463a8..b3578705631 100644
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -2043,6 +2043,7 @@ struct proto tcpv6_prot = {
 	.sysctl_rmem		= sysctl_tcp_rmem,
 	.max_header		= MAX_TCP_HEADER,
 	.obj_size		= sizeof(struct tcp6_sock),
+	.slab_flags		= SLAB_DESTROY_BY_RCU,
 	.twsk_prot		= &tcp6_timewait_sock_ops,
 	.rsk_prot		= &tcp6_request_sock_ops,
 	.h.hashinfo		= &tcp_hashinfo,
-- 
cgit v1.2.3-70-g09d2


From 5caea4ea7088e80ac5410d04660346094608b909 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <dada1@cosmosbay.com>
Date: Thu, 20 Nov 2008 00:40:07 -0800
Subject: net: listening_hash get a spinlock per bucket

This patch prepares RCU migration of listening_hash table for
TCP/DCCP protocols.

listening_hash table being small (32 slots per protocol), we add
a spinlock for each slot, instead of a single rwlock for whole table.

This should reduce hold time of readers, and writers concurrency.

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/inet_hashtables.h | 45 ++++++++--------------
 net/dccp/proto.c              |  8 +---
 net/ipv4/inet_diag.c          | 12 +++---
 net/ipv4/inet_hashtables.c    | 86 ++++++++++++++++---------------------------
 net/ipv4/tcp_ipv4.c           | 24 ++++++------
 net/ipv6/inet6_hashtables.c   | 23 ++++++------
 6 files changed, 79 insertions(+), 119 deletions(-)

(limited to 'include/net/inet_hashtables.h')

diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index 48189604511..62d2dd0d786 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -99,6 +99,11 @@ struct inet_bind_hashbucket {
 	struct hlist_head	chain;
 };
 
+struct inet_listen_hashbucket {
+	spinlock_t		lock;
+	struct hlist_head	head;
+};
+
 /* This is for listening sockets, thus all sockets which possess wildcards. */
 #define INET_LHTABLE_SIZE	32	/* Yes, really, this is all you need. */
 
@@ -123,22 +128,21 @@ struct inet_hashinfo {
 	unsigned int			bhash_size;
 	/* Note : 4 bytes padding on 64 bit arches */
 
-	/* All sockets in TCP_LISTEN state will be in here.  This is the only
-	 * table where wildcard'd TCP sockets can exist.  Hash function here
-	 * is just local port number.
-	 */
-	struct hlist_head		listening_hash[INET_LHTABLE_SIZE];
+	struct kmem_cache		*bind_bucket_cachep;
 
 	/* All the above members are written once at bootup and
 	 * never written again _or_ are predominantly read-access.
 	 *
 	 * Now align to a new cache line as all the following members
-	 * are often dirty.
+	 * might be often dirty.
+	 */
+	/* All sockets in TCP_LISTEN state will be in here.  This is the only
+	 * table where wildcard'd TCP sockets can exist.  Hash function here
+	 * is just local port number.
 	 */
-	rwlock_t			lhash_lock ____cacheline_aligned;
-	atomic_t			lhash_users;
-	wait_queue_head_t		lhash_wait;
-	struct kmem_cache			*bind_bucket_cachep;
+	struct inet_listen_hashbucket	listening_hash[INET_LHTABLE_SIZE]
+					____cacheline_aligned_in_smp;
+
 };
 
 static inline struct inet_ehash_bucket *inet_ehash_bucket(
@@ -236,26 +240,7 @@ extern void __inet_inherit_port(struct sock *sk, struct sock *child);
 
 extern void inet_put_port(struct sock *sk);
 
-extern void inet_listen_wlock(struct inet_hashinfo *hashinfo);
-
-/*
- * - We may sleep inside this lock.
- * - If sleeping is not required (or called from BH),
- *   use plain read_(un)lock(&inet_hashinfo.lhash_lock).
- */
-static inline void inet_listen_lock(struct inet_hashinfo *hashinfo)
-{
-	/* read_lock synchronizes to candidates to writers */
-	read_lock(&hashinfo->lhash_lock);
-	atomic_inc(&hashinfo->lhash_users);
-	read_unlock(&hashinfo->lhash_lock);
-}
-
-static inline void inet_listen_unlock(struct inet_hashinfo *hashinfo)
-{
-	if (atomic_dec_and_test(&hashinfo->lhash_users))
-		wake_up(&hashinfo->lhash_wait);
-}
+void inet_hashinfo_init(struct inet_hashinfo *h);
 
 extern void __inet_hash_nolisten(struct sock *sk);
 extern void inet_hash(struct sock *sk);
diff --git a/net/dccp/proto.c b/net/dccp/proto.c
index bdf784c422b..8b63394ec24 100644
--- a/net/dccp/proto.c
+++ b/net/dccp/proto.c
@@ -44,12 +44,7 @@ atomic_t dccp_orphan_count = ATOMIC_INIT(0);
 
 EXPORT_SYMBOL_GPL(dccp_orphan_count);
 
-struct inet_hashinfo __cacheline_aligned dccp_hashinfo = {
-	.lhash_lock	= RW_LOCK_UNLOCKED,
-	.lhash_users	= ATOMIC_INIT(0),
-	.lhash_wait = __WAIT_QUEUE_HEAD_INITIALIZER(dccp_hashinfo.lhash_wait),
-};
-
+struct inet_hashinfo dccp_hashinfo;
 EXPORT_SYMBOL_GPL(dccp_hashinfo);
 
 /* the maximum queue length for tx in packets. 0 is no limit */
@@ -1030,6 +1025,7 @@ static int __init dccp_init(void)
 	BUILD_BUG_ON(sizeof(struct dccp_skb_cb) >
 		     FIELD_SIZEOF(struct sk_buff, cb));
 
+	inet_hashinfo_init(&dccp_hashinfo);
 	dccp_hashinfo.bind_bucket_cachep =
 		kmem_cache_create("dccp_bind_bucket",
 				  sizeof(struct inet_bind_bucket), 0,
diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c
index 41b36720e97..1cb154ed75a 100644
--- a/net/ipv4/inet_diag.c
+++ b/net/ipv4/inet_diag.c
@@ -718,13 +718,15 @@ static int inet_diag_dump(struct sk_buff *skb, struct netlink_callback *cb)
 		if (!(r->idiag_states & (TCPF_LISTEN | TCPF_SYN_RECV)))
 			goto skip_listen_ht;
 
-		inet_listen_lock(hashinfo);
 		for (i = s_i; i < INET_LHTABLE_SIZE; i++) {
 			struct sock *sk;
 			struct hlist_node *node;
+			struct inet_listen_hashbucket *ilb;
 
 			num = 0;
-			sk_for_each(sk, node, &hashinfo->listening_hash[i]) {
+			ilb = &hashinfo->listening_hash[i];
+			spin_lock_bh(&ilb->lock);
+			sk_for_each(sk, node, &ilb->head) {
 				struct inet_sock *inet = inet_sk(sk);
 
 				if (num < s_num) {
@@ -742,7 +744,7 @@ static int inet_diag_dump(struct sk_buff *skb, struct netlink_callback *cb)
 					goto syn_recv;
 
 				if (inet_csk_diag_dump(sk, skb, cb) < 0) {
-					inet_listen_unlock(hashinfo);
+					spin_unlock_bh(&ilb->lock);
 					goto done;
 				}
 
@@ -751,7 +753,7 @@ syn_recv:
 					goto next_listen;
 
 				if (inet_diag_dump_reqs(skb, sk, cb) < 0) {
-					inet_listen_unlock(hashinfo);
+					spin_unlock_bh(&ilb->lock);
 					goto done;
 				}
 
@@ -760,12 +762,12 @@ next_listen:
 				cb->args[4] = 0;
 				++num;
 			}
+			spin_unlock_bh(&ilb->lock);
 
 			s_num = 0;
 			cb->args[3] = 0;
 			cb->args[4] = 0;
 		}
-		inet_listen_unlock(hashinfo);
 skip_listen_ht:
 		cb->args[0] = 1;
 		s_i = num = s_num = 0;
diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index fd269cfef0e..377d004e572 100644
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -110,35 +110,6 @@ void __inet_inherit_port(struct sock *sk, struct sock *child)
 
 EXPORT_SYMBOL_GPL(__inet_inherit_port);
 
-/*
- * This lock without WQ_FLAG_EXCLUSIVE is good on UP and it can be very bad on SMP.
- * Look, when several writers sleep and reader wakes them up, all but one
- * immediately hit write lock and grab all the cpus. Exclusive sleep solves
- * this, _but_ remember, it adds useless work on UP machines (wake up each
- * exclusive lock release). It should be ifdefed really.
- */
-void inet_listen_wlock(struct inet_hashinfo *hashinfo)
-	__acquires(hashinfo->lhash_lock)
-{
-	write_lock(&hashinfo->lhash_lock);
-
-	if (atomic_read(&hashinfo->lhash_users)) {
-		DEFINE_WAIT(wait);
-
-		for (;;) {
-			prepare_to_wait_exclusive(&hashinfo->lhash_wait,
-						  &wait, TASK_UNINTERRUPTIBLE);
-			if (!atomic_read(&hashinfo->lhash_users))
-				break;
-			write_unlock_bh(&hashinfo->lhash_lock);
-			schedule();
-			write_lock_bh(&hashinfo->lhash_lock);
-		}
-
-		finish_wait(&hashinfo->lhash_wait, &wait);
-	}
-}
-
 /*
  * Don't inline this cruft. Here are some nice properties to exploit here. The
  * BSD API does not allow a listening sock to specify the remote port nor the
@@ -191,25 +162,25 @@ struct sock *__inet_lookup_listener(struct net *net,
 				    const int dif)
 {
 	struct sock *sk = NULL;
-	const struct hlist_head *head;
+	struct inet_listen_hashbucket *ilb;
 
-	read_lock(&hashinfo->lhash_lock);
-	head = &hashinfo->listening_hash[inet_lhashfn(net, hnum)];
-	if (!hlist_empty(head)) {
-		const struct inet_sock *inet = inet_sk((sk = __sk_head(head)));
+	ilb = &hashinfo->listening_hash[inet_lhashfn(net, hnum)];
+	spin_lock(&ilb->lock);
+	if (!hlist_empty(&ilb->head)) {
+		const struct inet_sock *inet = inet_sk((sk = __sk_head(&ilb->head)));
 
 		if (inet->num == hnum && !sk->sk_node.next &&
 		    (!inet->rcv_saddr || inet->rcv_saddr == daddr) &&
 		    (sk->sk_family == PF_INET || !ipv6_only_sock(sk)) &&
 		    !sk->sk_bound_dev_if && net_eq(sock_net(sk), net))
 			goto sherry_cache;
-		sk = inet_lookup_listener_slow(net, head, daddr, hnum, dif);
+		sk = inet_lookup_listener_slow(net, &ilb->head, daddr, hnum, dif);
 	}
 	if (sk) {
 sherry_cache:
 		sock_hold(sk);
 	}
-	read_unlock(&hashinfo->lhash_lock);
+	spin_unlock(&ilb->lock);
 	return sk;
 }
 EXPORT_SYMBOL_GPL(__inet_lookup_listener);
@@ -389,8 +360,7 @@ EXPORT_SYMBOL_GPL(__inet_hash_nolisten);
 static void __inet_hash(struct sock *sk)
 {
 	struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo;
-	struct hlist_head *list;
-	rwlock_t *lock;
+	struct inet_listen_hashbucket *ilb;
 
 	if (sk->sk_state != TCP_LISTEN) {
 		__inet_hash_nolisten(sk);
@@ -398,14 +368,12 @@ static void __inet_hash(struct sock *sk)
 	}
 
 	WARN_ON(!sk_unhashed(sk));
-	list = &hashinfo->listening_hash[inet_sk_listen_hashfn(sk)];
-	lock = &hashinfo->lhash_lock;
+	ilb = &hashinfo->listening_hash[inet_sk_listen_hashfn(sk)];
 
-	inet_listen_wlock(hashinfo);
-	__sk_add_node(sk, list);
+	spin_lock(&ilb->lock);
+	__sk_add_node(sk, &ilb->head);
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
-	write_unlock(lock);
-	wake_up(&hashinfo->lhash_wait);
+	spin_unlock(&ilb->lock);
 }
 
 void inet_hash(struct sock *sk)
@@ -420,29 +388,27 @@ EXPORT_SYMBOL_GPL(inet_hash);
 
 void inet_unhash(struct sock *sk)
 {
-	rwlock_t *lock;
 	struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo;
 
 	if (sk_unhashed(sk))
-		goto out;
+		return;
 
 	if (sk->sk_state == TCP_LISTEN) {
-		local_bh_disable();
-		inet_listen_wlock(hashinfo);
-		lock = &hashinfo->lhash_lock;
+		struct inet_listen_hashbucket *ilb;
+
+		ilb = &hashinfo->listening_hash[inet_sk_listen_hashfn(sk)];
+		spin_lock_bh(&ilb->lock);
 		if (__sk_del_node_init(sk))
 			sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
+		spin_unlock_bh(&ilb->lock);
 	} else {
-		lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
+		rwlock_t *lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
+
 		write_lock_bh(lock);
 		if (__sk_nulls_del_node_init_rcu(sk))
 			sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
+		write_unlock_bh(lock);
 	}
-
-	write_unlock_bh(lock);
-out:
-	if (sk->sk_state == TCP_LISTEN)
-		wake_up(&hashinfo->lhash_wait);
 }
 EXPORT_SYMBOL_GPL(inet_unhash);
 
@@ -556,3 +522,13 @@ int inet_hash_connect(struct inet_timewait_death_row *death_row,
 }
 
 EXPORT_SYMBOL_GPL(inet_hash_connect);
+
+void inet_hashinfo_init(struct inet_hashinfo *h)
+{
+	int i;
+
+	for (i = 0; i < INET_LHTABLE_SIZE; i++)
+		spin_lock_init(&h->listening_hash[i].lock);
+}
+
+EXPORT_SYMBOL_GPL(inet_hashinfo_init);
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 5559fea61e8..330b08a1227 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -97,11 +97,7 @@ struct tcp_md5sig_key *tcp_v4_md5_do_lookup(struct sock *sk, __be32 addr)
 }
 #endif
 
-struct inet_hashinfo __cacheline_aligned tcp_hashinfo = {
-	.lhash_lock  = __RW_LOCK_UNLOCKED(tcp_hashinfo.lhash_lock),
-	.lhash_users = ATOMIC_INIT(0),
-	.lhash_wait  = __WAIT_QUEUE_HEAD_INITIALIZER(tcp_hashinfo.lhash_wait),
-};
+struct inet_hashinfo tcp_hashinfo;
 
 static inline __u32 tcp_v4_init_sequence(struct sk_buff *skb)
 {
@@ -1874,15 +1870,18 @@ static void *listening_get_next(struct seq_file *seq, void *cur)
 	struct inet_connection_sock *icsk;
 	struct hlist_node *node;
 	struct sock *sk = cur;
+	struct inet_listen_hashbucket *ilb;
 	struct tcp_iter_state *st = seq->private;
 	struct net *net = seq_file_net(seq);
 
 	if (!sk) {
 		st->bucket = 0;
-		sk = sk_head(&tcp_hashinfo.listening_hash[0]);
+		ilb = &tcp_hashinfo.listening_hash[0];
+		spin_lock_bh(&ilb->lock);
+		sk = sk_head(&ilb->head);
 		goto get_sk;
 	}
-
+	ilb = &tcp_hashinfo.listening_hash[st->bucket];
 	++st->num;
 
 	if (st->state == TCP_SEQ_STATE_OPENREQ) {
@@ -1932,8 +1931,11 @@ start_req:
 		}
 		read_unlock_bh(&icsk->icsk_accept_queue.syn_wait_lock);
 	}
+	spin_unlock_bh(&ilb->lock);
 	if (++st->bucket < INET_LHTABLE_SIZE) {
-		sk = sk_head(&tcp_hashinfo.listening_hash[st->bucket]);
+		ilb = &tcp_hashinfo.listening_hash[st->bucket];
+		spin_lock_bh(&ilb->lock);
+		sk = sk_head(&ilb->head);
 		goto get_sk;
 	}
 	cur = NULL;
@@ -2066,12 +2068,10 @@ static void *tcp_get_idx(struct seq_file *seq, loff_t pos)
 	void *rc;
 	struct tcp_iter_state *st = seq->private;
 
-	inet_listen_lock(&tcp_hashinfo);
 	st->state = TCP_SEQ_STATE_LISTENING;
 	rc	  = listening_get_idx(seq, &pos);
 
 	if (!rc) {
-		inet_listen_unlock(&tcp_hashinfo);
 		st->state = TCP_SEQ_STATE_ESTABLISHED;
 		rc	  = established_get_idx(seq, pos);
 	}
@@ -2103,7 +2103,6 @@ static void *tcp_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 	case TCP_SEQ_STATE_LISTENING:
 		rc = listening_get_next(seq, v);
 		if (!rc) {
-			inet_listen_unlock(&tcp_hashinfo);
 			st->state = TCP_SEQ_STATE_ESTABLISHED;
 			rc	  = established_get_first(seq);
 		}
@@ -2130,7 +2129,7 @@ static void tcp_seq_stop(struct seq_file *seq, void *v)
 		}
 	case TCP_SEQ_STATE_LISTENING:
 		if (v != SEQ_START_TOKEN)
-			inet_listen_unlock(&tcp_hashinfo);
+			spin_unlock_bh(&tcp_hashinfo.listening_hash[st->bucket].lock);
 		break;
 	case TCP_SEQ_STATE_TIME_WAIT:
 	case TCP_SEQ_STATE_ESTABLISHED:
@@ -2405,6 +2404,7 @@ static struct pernet_operations __net_initdata tcp_sk_ops = {
 
 void __init tcp_v4_init(void)
 {
+	inet_hashinfo_init(&tcp_hashinfo);
 	if (register_pernet_device(&tcp_sk_ops))
 		panic("Failed to create the TCP control socket.\n");
 }
diff --git a/net/ipv6/inet6_hashtables.c b/net/ipv6/inet6_hashtables.c
index c1b4d401fd9..21544b9be25 100644
--- a/net/ipv6/inet6_hashtables.c
+++ b/net/ipv6/inet6_hashtables.c
@@ -25,30 +25,30 @@
 void __inet6_hash(struct sock *sk)
 {
 	struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo;
-	rwlock_t *lock;
 
 	WARN_ON(!sk_unhashed(sk));
 
 	if (sk->sk_state == TCP_LISTEN) {
-		struct hlist_head *list;
+		struct inet_listen_hashbucket *ilb;
 
-		list = &hashinfo->listening_hash[inet_sk_listen_hashfn(sk)];
-		lock = &hashinfo->lhash_lock;
-		inet_listen_wlock(hashinfo);
-		__sk_add_node(sk, list);
+		ilb = &hashinfo->listening_hash[inet_sk_listen_hashfn(sk)];
+		spin_lock(&ilb->lock);
+		__sk_add_node(sk, &ilb->head);
+		spin_unlock(&ilb->lock);
 	} else {
 		unsigned int hash;
 		struct hlist_nulls_head *list;
+		rwlock_t *lock;
 
 		sk->sk_hash = hash = inet6_sk_ehashfn(sk);
 		list = &inet_ehash_bucket(hashinfo, hash)->chain;
 		lock = inet_ehash_lockp(hashinfo, hash);
 		write_lock(lock);
 		__sk_nulls_add_node_rcu(sk, list);
+		write_unlock(lock);
 	}
 
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
-	write_unlock(lock);
 }
 EXPORT_SYMBOL(__inet6_hash);
 
@@ -126,10 +126,11 @@ struct sock *inet6_lookup_listener(struct net *net,
 	const struct hlist_node *node;
 	struct sock *result = NULL;
 	int score, hiscore = 0;
+	struct inet_listen_hashbucket *ilb;
 
-	read_lock(&hashinfo->lhash_lock);
-	sk_for_each(sk, node,
-			&hashinfo->listening_hash[inet_lhashfn(net, hnum)]) {
+	ilb = &hashinfo->listening_hash[inet_lhashfn(net, hnum)];
+	spin_lock(&ilb->lock);
+	sk_for_each(sk, node, &ilb->head) {
 		if (net_eq(sock_net(sk), net) && inet_sk(sk)->num == hnum &&
 				sk->sk_family == PF_INET6) {
 			const struct ipv6_pinfo *np = inet6_sk(sk);
@@ -157,7 +158,7 @@ struct sock *inet6_lookup_listener(struct net *net,
 	}
 	if (result)
 		sock_hold(result);
-	read_unlock(&hashinfo->lhash_lock);
+	spin_unlock(&ilb->lock);
 	return result;
 }
 
-- 
cgit v1.2.3-70-g09d2


From 9db66bdcc83749affe61c61eb8ff3cf08f42afec Mon Sep 17 00:00:00 2001
From: Eric Dumazet <dada1@cosmosbay.com>
Date: Thu, 20 Nov 2008 20:39:09 -0800
Subject: net: convert TCP/DCCP ehash rwlocks to spinlocks

Now TCP & DCCP use RCU lookups, we can convert ehash rwlocks to spinlocks.

/proc/net/tcp and other seq_file 'readers' can safely be converted to 'writers'.

This should speedup writers, since spin_lock()/spin_unlock()
only use one atomic operation instead of two for write_lock()/write_unlock()

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/inet_hashtables.h | 14 +++++++-------
 net/ipv4/inet_hashtables.c    | 21 ++++++++++-----------
 net/ipv4/inet_timewait_sock.c | 22 +++++++++++-----------
 net/ipv4/tcp_ipv4.c           | 12 ++++++------
 net/ipv6/inet6_hashtables.c   | 15 +++++++--------
 5 files changed, 41 insertions(+), 43 deletions(-)

(limited to 'include/net/inet_hashtables.h')

diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index 62d2dd0d786..28b3ee3e8d6 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -116,7 +116,7 @@ struct inet_hashinfo {
 	 * TIME_WAIT sockets use a separate chain (twchain).
 	 */
 	struct inet_ehash_bucket	*ehash;
-	rwlock_t			*ehash_locks;
+	spinlock_t			*ehash_locks;
 	unsigned int			ehash_size;
 	unsigned int			ehash_locks_mask;
 
@@ -152,7 +152,7 @@ static inline struct inet_ehash_bucket *inet_ehash_bucket(
 	return &hashinfo->ehash[hash & (hashinfo->ehash_size - 1)];
 }
 
-static inline rwlock_t *inet_ehash_lockp(
+static inline spinlock_t *inet_ehash_lockp(
 	struct inet_hashinfo *hashinfo,
 	unsigned int hash)
 {
@@ -177,16 +177,16 @@ static inline int inet_ehash_locks_alloc(struct inet_hashinfo *hashinfo)
 		size = 4096;
 	if (sizeof(rwlock_t) != 0) {
 #ifdef CONFIG_NUMA
-		if (size * sizeof(rwlock_t) > PAGE_SIZE)
-			hashinfo->ehash_locks = vmalloc(size * sizeof(rwlock_t));
+		if (size * sizeof(spinlock_t) > PAGE_SIZE)
+			hashinfo->ehash_locks = vmalloc(size * sizeof(spinlock_t));
 		else
 #endif
-		hashinfo->ehash_locks =	kmalloc(size * sizeof(rwlock_t),
+		hashinfo->ehash_locks =	kmalloc(size * sizeof(spinlock_t),
 						GFP_KERNEL);
 		if (!hashinfo->ehash_locks)
 			return ENOMEM;
 		for (i = 0; i < size; i++)
-			rwlock_init(&hashinfo->ehash_locks[i]);
+			spin_lock_init(&hashinfo->ehash_locks[i]);
 	}
 	hashinfo->ehash_locks_mask = size - 1;
 	return 0;
@@ -197,7 +197,7 @@ static inline void inet_ehash_locks_free(struct inet_hashinfo *hashinfo)
 	if (hashinfo->ehash_locks) {
 #ifdef CONFIG_NUMA
 		unsigned int size = (hashinfo->ehash_locks_mask + 1) *
-							sizeof(rwlock_t);
+							sizeof(spinlock_t);
 		if (size > PAGE_SIZE)
 			vfree(hashinfo->ehash_locks);
 		else
diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index 377d004e572..4c273a9981a 100644
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -271,13 +271,12 @@ static int __inet_check_established(struct inet_timewait_death_row *death_row,
 	struct net *net = sock_net(sk);
 	unsigned int hash = inet_ehashfn(net, daddr, lport, saddr, inet->dport);
 	struct inet_ehash_bucket *head = inet_ehash_bucket(hinfo, hash);
-	rwlock_t *lock = inet_ehash_lockp(hinfo, hash);
+	spinlock_t *lock = inet_ehash_lockp(hinfo, hash);
 	struct sock *sk2;
 	const struct hlist_nulls_node *node;
 	struct inet_timewait_sock *tw;
 
-	prefetch(head->chain.first);
-	write_lock(lock);
+	spin_lock(lock);
 
 	/* Check TIME-WAIT sockets first. */
 	sk_nulls_for_each(sk2, node, &head->twchain) {
@@ -308,8 +307,8 @@ unique:
 	sk->sk_hash = hash;
 	WARN_ON(!sk_unhashed(sk));
 	__sk_nulls_add_node_rcu(sk, &head->chain);
+	spin_unlock(lock);
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
-	write_unlock(lock);
 
 	if (twp) {
 		*twp = tw;
@@ -325,7 +324,7 @@ unique:
 	return 0;
 
 not_unique:
-	write_unlock(lock);
+	spin_unlock(lock);
 	return -EADDRNOTAVAIL;
 }
 
@@ -340,7 +339,7 @@ void __inet_hash_nolisten(struct sock *sk)
 {
 	struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo;
 	struct hlist_nulls_head *list;
-	rwlock_t *lock;
+	spinlock_t *lock;
 	struct inet_ehash_bucket *head;
 
 	WARN_ON(!sk_unhashed(sk));
@@ -350,10 +349,10 @@ void __inet_hash_nolisten(struct sock *sk)
 	list = &head->chain;
 	lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
 
-	write_lock(lock);
+	spin_lock(lock);
 	__sk_nulls_add_node_rcu(sk, list);
+	spin_unlock(lock);
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
-	write_unlock(lock);
 }
 EXPORT_SYMBOL_GPL(__inet_hash_nolisten);
 
@@ -402,12 +401,12 @@ void inet_unhash(struct sock *sk)
 			sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
 		spin_unlock_bh(&ilb->lock);
 	} else {
-		rwlock_t *lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
+		spinlock_t *lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
 
-		write_lock_bh(lock);
+		spin_lock_bh(lock);
 		if (__sk_nulls_del_node_init_rcu(sk))
 			sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
-		write_unlock_bh(lock);
+		spin_unlock_bh(lock);
 	}
 }
 EXPORT_SYMBOL_GPL(inet_unhash);
diff --git a/net/ipv4/inet_timewait_sock.c b/net/ipv4/inet_timewait_sock.c
index 60689951ecd..8554d0ea171 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. */
-	rwlock_t *lock = inet_ehash_lockp(hashinfo, tw->tw_hash);
+	spinlock_t *lock = inet_ehash_lockp(hashinfo, tw->tw_hash);
 
-	write_lock(lock);
+	spin_lock(lock);
 	if (hlist_nulls_unhashed(&tw->tw_node)) {
-		write_unlock(lock);
+		spin_unlock(lock);
 		return;
 	}
 	hlist_nulls_del_rcu(&tw->tw_node);
 	sk_nulls_node_init(&tw->tw_node);
-	write_unlock(lock);
+	spin_unlock(lock);
 
 	/* Disassociate with bind bucket. */
 	bhead = &hashinfo->bhash[inet_bhashfn(twsk_net(tw), tw->tw_num,
@@ -76,7 +76,7 @@ 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);
+	spinlock_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
@@ -90,7 +90,7 @@ 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(lock);
+	spin_lock(lock);
 
 	/*
 	 * Step 2: Hash TW into TIMEWAIT chain.
@@ -104,7 +104,7 @@ void __inet_twsk_hashdance(struct inet_timewait_sock *tw, struct sock *sk,
 	if (__sk_nulls_del_node_init_rcu(sk))
 		sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
 
-	write_unlock(lock);
+	spin_unlock(lock);
 }
 
 EXPORT_SYMBOL_GPL(__inet_twsk_hashdance);
@@ -427,9 +427,9 @@ void inet_twsk_purge(struct net *net, struct inet_hashinfo *hashinfo,
 	for (h = 0; h < (hashinfo->ehash_size); h++) {
 		struct inet_ehash_bucket *head =
 			inet_ehash_bucket(hashinfo, h);
-		rwlock_t *lock = inet_ehash_lockp(hashinfo, h);
+		spinlock_t *lock = inet_ehash_lockp(hashinfo, h);
 restart:
-		write_lock(lock);
+		spin_lock(lock);
 		sk_nulls_for_each(sk, node, &head->twchain) {
 
 			tw = inet_twsk(sk);
@@ -438,13 +438,13 @@ restart:
 				continue;
 
 			atomic_inc(&tw->tw_refcnt);
-			write_unlock(lock);
+			spin_unlock(lock);
 			inet_twsk_deschedule(tw, twdr);
 			inet_twsk_put(tw);
 
 			goto restart;
 		}
-		write_unlock(lock);
+		spin_unlock(lock);
 	}
 	local_bh_enable();
 }
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 330b08a1227..a81caa1be0c 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1970,13 +1970,13 @@ static void *established_get_first(struct seq_file *seq)
 		struct sock *sk;
 		struct hlist_nulls_node *node;
 		struct inet_timewait_sock *tw;
-		rwlock_t *lock = inet_ehash_lockp(&tcp_hashinfo, st->bucket);
+		spinlock_t *lock = inet_ehash_lockp(&tcp_hashinfo, st->bucket);
 
 		/* Lockless fast path for the common case of empty buckets */
 		if (empty_bucket(st))
 			continue;
 
-		read_lock_bh(lock);
+		spin_lock_bh(lock);
 		sk_nulls_for_each(sk, node, &tcp_hashinfo.ehash[st->bucket].chain) {
 			if (sk->sk_family != st->family ||
 			    !net_eq(sock_net(sk), net)) {
@@ -1995,7 +1995,7 @@ static void *established_get_first(struct seq_file *seq)
 			rc = tw;
 			goto out;
 		}
-		read_unlock_bh(lock);
+		spin_unlock_bh(lock);
 		st->state = TCP_SEQ_STATE_ESTABLISHED;
 	}
 out:
@@ -2023,7 +2023,7 @@ get_tw:
 			cur = tw;
 			goto out;
 		}
-		read_unlock_bh(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
+		spin_unlock_bh(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
 		st->state = TCP_SEQ_STATE_ESTABLISHED;
 
 		/* Look for next non empty bucket */
@@ -2033,7 +2033,7 @@ get_tw:
 		if (st->bucket >= tcp_hashinfo.ehash_size)
 			return NULL;
 
-		read_lock_bh(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
+		spin_lock_bh(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
 		sk = sk_nulls_head(&tcp_hashinfo.ehash[st->bucket].chain);
 	} else
 		sk = sk_nulls_next(sk);
@@ -2134,7 +2134,7 @@ 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(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
+			spin_unlock_bh(inet_ehash_lockp(&tcp_hashinfo, st->bucket));
 		break;
 	}
 }
diff --git a/net/ipv6/inet6_hashtables.c b/net/ipv6/inet6_hashtables.c
index 21544b9be25..e0fd68187f8 100644
--- a/net/ipv6/inet6_hashtables.c
+++ b/net/ipv6/inet6_hashtables.c
@@ -38,14 +38,14 @@ void __inet6_hash(struct sock *sk)
 	} else {
 		unsigned int hash;
 		struct hlist_nulls_head *list;
-		rwlock_t *lock;
+		spinlock_t *lock;
 
 		sk->sk_hash = hash = inet6_sk_ehashfn(sk);
 		list = &inet_ehash_bucket(hashinfo, hash)->chain;
 		lock = inet_ehash_lockp(hashinfo, hash);
-		write_lock(lock);
+		spin_lock(lock);
 		__sk_nulls_add_node_rcu(sk, list);
-		write_unlock(lock);
+		spin_unlock(lock);
 	}
 
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
@@ -195,13 +195,12 @@ static int __inet6_check_established(struct inet_timewait_death_row *death_row,
 	const unsigned int hash = inet6_ehashfn(net, daddr, lport, saddr,
 						inet->dport);
 	struct inet_ehash_bucket *head = inet_ehash_bucket(hinfo, hash);
-	rwlock_t *lock = inet_ehash_lockp(hinfo, hash);
+	spinlock_t *lock = inet_ehash_lockp(hinfo, hash);
 	struct sock *sk2;
 	const struct hlist_nulls_node *node;
 	struct inet_timewait_sock *tw;
 
-	prefetch(head->chain.first);
-	write_lock(lock);
+	spin_lock(lock);
 
 	/* Check TIME-WAIT sockets first. */
 	sk_nulls_for_each(sk2, node, &head->twchain) {
@@ -230,8 +229,8 @@ unique:
 	WARN_ON(!sk_unhashed(sk));
 	__sk_nulls_add_node_rcu(sk, &head->chain);
 	sk->sk_hash = hash;
+	spin_unlock(lock);
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
-	write_unlock(lock);
 
 	if (twp != NULL) {
 		*twp = tw;
@@ -246,7 +245,7 @@ unique:
 	return 0;
 
 not_unique:
-	write_unlock(lock);
+	spin_unlock(lock);
 	return -EADDRNOTAVAIL;
 }
 
-- 
cgit v1.2.3-70-g09d2


From f757fec4b0d45dfcb52f9a914a12225a6a0a3e05 Mon Sep 17 00:00:00 2001
From: Eric Dumazet <dada1@cosmosbay.com>
Date: Fri, 21 Nov 2008 15:49:19 -0800
Subject: net: use net_eq() in INET_MATCH and INET_TW_MATCH

We can avoid some useless instructions if !CONFIG_NET_NS

Because of RCU, we use INET_MATCH or INET_TW_MATCH twice for the found
socket, so thats six instructions less per incoming TCP packet.

Yet another tbench speedup :)

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/inet_hashtables.h | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

(limited to 'include/net/inet_hashtables.h')

diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index 28b3ee3e8d6..ec7ee2e46d8 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -291,25 +291,25 @@ typedef __u64 __bitwise __addrpair;
 				   ((__force __u64)(__be32)(__saddr)));
 #endif /* __BIG_ENDIAN */
 #define INET_MATCH(__sk, __net, __hash, __cookie, __saddr, __daddr, __ports, __dif)\
-	(((__sk)->sk_hash == (__hash)) && sock_net((__sk)) == (__net)	&&	\
+	(((__sk)->sk_hash == (__hash)) && net_eq(sock_net(__sk), (__net)) &&	\
 	 ((*((__addrpair *)&(inet_sk(__sk)->daddr))) == (__cookie))	&&	\
 	 ((*((__portpair *)&(inet_sk(__sk)->dport))) == (__ports))	&&	\
 	 (!((__sk)->sk_bound_dev_if) || ((__sk)->sk_bound_dev_if == (__dif))))
 #define INET_TW_MATCH(__sk, __net, __hash, __cookie, __saddr, __daddr, __ports, __dif)\
-	(((__sk)->sk_hash == (__hash)) && sock_net((__sk)) == (__net)	&&	\
+	(((__sk)->sk_hash == (__hash)) && net_eq(sock_net(__sk), (__net)) &&	\
 	 ((*((__addrpair *)&(inet_twsk(__sk)->tw_daddr))) == (__cookie)) &&	\
 	 ((*((__portpair *)&(inet_twsk(__sk)->tw_dport))) == (__ports)) &&	\
 	 (!((__sk)->sk_bound_dev_if) || ((__sk)->sk_bound_dev_if == (__dif))))
 #else /* 32-bit arch */
 #define INET_ADDR_COOKIE(__name, __saddr, __daddr)
 #define INET_MATCH(__sk, __net, __hash, __cookie, __saddr, __daddr, __ports, __dif)	\
-	(((__sk)->sk_hash == (__hash)) && sock_net((__sk)) == (__net)	&&	\
+	(((__sk)->sk_hash == (__hash)) && net_eq(sock_net(__sk), (__net))	&&	\
 	 (inet_sk(__sk)->daddr		== (__saddr))		&&	\
 	 (inet_sk(__sk)->rcv_saddr	== (__daddr))		&&	\
 	 ((*((__portpair *)&(inet_sk(__sk)->dport))) == (__ports))	&&	\
 	 (!((__sk)->sk_bound_dev_if) || ((__sk)->sk_bound_dev_if == (__dif))))
 #define INET_TW_MATCH(__sk, __net, __hash,__cookie, __saddr, __daddr, __ports, __dif)	\
-	(((__sk)->sk_hash == (__hash)) && sock_net((__sk)) == (__net)	&&	\
+	(((__sk)->sk_hash == (__hash)) && net_eq(sock_net(__sk), (__net))	&&	\
 	 (inet_twsk(__sk)->tw_daddr	== (__saddr))		&&	\
 	 (inet_twsk(__sk)->tw_rcv_saddr	== (__daddr))		&&	\
 	 ((*((__portpair *)&(inet_twsk(__sk)->tw_dport))) == (__ports)) &&	\
-- 
cgit v1.2.3-70-g09d2


From c25eb3bfb97294d0543a81230fbc237046b4b84c Mon Sep 17 00:00:00 2001
From: Eric Dumazet <dada1@cosmosbay.com>
Date: Sun, 23 Nov 2008 17:22:55 -0800
Subject: net: Convert TCP/DCCP listening hash tables to use RCU

This is the last step to be able to perform full RCU lookups
in __inet_lookup() : After established/timewait tables, we
add RCU lookups to listening hash table.

The only trick here is that a socket of a given type (TCP ipv4,
TCP ipv6, ...) can now flight between two different tables
(established and listening) during a RCU grace period, so we
must use different 'nulls' end-of-chain values for two tables.

We define a large value :

#define LISTENING_NULLS_BASE (1U << 29)

So that slots in listening table are guaranteed to have different
end-of-chain values than slots in established table. A reader can
still detect it finished its lookup in the right chain.

Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
 include/net/inet_hashtables.h |   9 ++-
 net/ipv4/inet_diag.c          |   4 +-
 net/ipv4/inet_hashtables.c    | 148 +++++++++++++++++++++---------------------
 net/ipv4/tcp_ipv4.c           |   8 +--
 net/ipv6/inet6_hashtables.c   |  94 +++++++++++++++++----------
 5 files changed, 147 insertions(+), 116 deletions(-)

(limited to 'include/net/inet_hashtables.h')

diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index ec7ee2e46d8..f44bb5c77a7 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -99,9 +99,16 @@ struct inet_bind_hashbucket {
 	struct hlist_head	chain;
 };
 
+/*
+ * Sockets can be hashed in established or listening table
+ * We must use different 'nulls' end-of-chain value for listening
+ * hash table, or we might find a socket that was closed and
+ * reallocated/inserted into established hash table
+ */
+#define LISTENING_NULLS_BASE (1U << 29)
 struct inet_listen_hashbucket {
 	spinlock_t		lock;
-	struct hlist_head	head;
+	struct hlist_nulls_head	head;
 };
 
 /* This is for listening sockets, thus all sockets which possess wildcards. */
diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c
index 998a78f169f..588a7796e3e 100644
--- a/net/ipv4/inet_diag.c
+++ b/net/ipv4/inet_diag.c
@@ -720,13 +720,13 @@ static int inet_diag_dump(struct sk_buff *skb, struct netlink_callback *cb)
 
 		for (i = s_i; i < INET_LHTABLE_SIZE; i++) {
 			struct sock *sk;
-			struct hlist_node *node;
+			struct hlist_nulls_node *node;
 			struct inet_listen_hashbucket *ilb;
 
 			num = 0;
 			ilb = &hashinfo->listening_hash[i];
 			spin_lock_bh(&ilb->lock);
-			sk_for_each(sk, node, &ilb->head) {
+			sk_nulls_for_each(sk, node, &ilb->head) {
 				struct inet_sock *inet = inet_sk(sk);
 
 				if (num < s_num) {
diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index 4c273a9981a..11fcb87a1fd 100644
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -110,78 +110,79 @@ void __inet_inherit_port(struct sock *sk, struct sock *child)
 
 EXPORT_SYMBOL_GPL(__inet_inherit_port);
 
+static inline int compute_score(struct sock *sk, struct net *net,
+				const unsigned short hnum, const __be32 daddr,
+				const int dif)
+{
+	int score = -1;
+	struct inet_sock *inet = inet_sk(sk);
+
+	if (net_eq(sock_net(sk), net) && inet->num == hnum &&
+			!ipv6_only_sock(sk)) {
+		__be32 rcv_saddr = inet->rcv_saddr;
+		score = sk->sk_family == PF_INET ? 1 : 0;
+		if (rcv_saddr) {
+			if (rcv_saddr != daddr)
+				return -1;
+			score += 2;
+		}
+		if (sk->sk_bound_dev_if) {
+			if (sk->sk_bound_dev_if != dif)
+				return -1;
+			score += 2;
+		}
+	}
+	return score;
+}
+
 /*
  * Don't inline this cruft. Here are some nice properties to exploit here. The
  * BSD API does not allow a listening sock to specify the remote port nor the
  * remote address for the connection. So always assume those are both
  * wildcarded during the search since they can never be otherwise.
  */
-static struct sock *inet_lookup_listener_slow(struct net *net,
-					      const struct hlist_head *head,
-					      const __be32 daddr,
-					      const unsigned short hnum,
-					      const int dif)
-{
-	struct sock *result = NULL, *sk;
-	const struct hlist_node *node;
-	int hiscore = -1;
-
-	sk_for_each(sk, node, head) {
-		const struct inet_sock *inet = inet_sk(sk);
-
-		if (net_eq(sock_net(sk), net) && inet->num == hnum &&
-				!ipv6_only_sock(sk)) {
-			const __be32 rcv_saddr = inet->rcv_saddr;
-			int score = sk->sk_family == PF_INET ? 1 : 0;
-
-			if (rcv_saddr) {
-				if (rcv_saddr != daddr)
-					continue;
-				score += 2;
-			}
-			if (sk->sk_bound_dev_if) {
-				if (sk->sk_bound_dev_if != dif)
-					continue;
-				score += 2;
-			}
-			if (score == 5)
-				return sk;
-			if (score > hiscore) {
-				hiscore	= score;
-				result	= sk;
-			}
-		}
-	}
-	return result;
-}
 
-/* Optimize the common listener case. */
+
 struct sock *__inet_lookup_listener(struct net *net,
 				    struct inet_hashinfo *hashinfo,
 				    const __be32 daddr, const unsigned short hnum,
 				    const int dif)
 {
-	struct sock *sk = NULL;
-	struct inet_listen_hashbucket *ilb;
+	struct sock *sk, *result;
+	struct hlist_nulls_node *node;
+	unsigned int hash = inet_lhashfn(net, hnum);
+	struct inet_listen_hashbucket *ilb = &hashinfo->listening_hash[hash];
+	int score, hiscore;
 
-	ilb = &hashinfo->listening_hash[inet_lhashfn(net, hnum)];
-	spin_lock(&ilb->lock);
-	if (!hlist_empty(&ilb->head)) {
-		const struct inet_sock *inet = inet_sk((sk = __sk_head(&ilb->head)));
-
-		if (inet->num == hnum && !sk->sk_node.next &&
-		    (!inet->rcv_saddr || inet->rcv_saddr == daddr) &&
-		    (sk->sk_family == PF_INET || !ipv6_only_sock(sk)) &&
-		    !sk->sk_bound_dev_if && net_eq(sock_net(sk), net))
-			goto sherry_cache;
-		sk = inet_lookup_listener_slow(net, &ilb->head, daddr, hnum, dif);
+	rcu_read_lock();
+begin:
+	result = NULL;
+	hiscore = -1;
+	sk_nulls_for_each_rcu(sk, node, &ilb->head) {
+		score = compute_score(sk, net, hnum, daddr, dif);
+		if (score > hiscore) {
+			result = sk;
+			hiscore = score;
+		}
 	}
-	if (sk) {
-sherry_cache:
-		sock_hold(sk);
+	/*
+	 * if the nulls value we got at the end of this lookup is
+	 * not the expected one, we must restart lookup.
+	 * We probably met an item that was moved to another chain.
+	 */
+	if (get_nulls_value(node) != hash + LISTENING_NULLS_BASE)
+		goto begin;
+	if (result) {
+		if (unlikely(!atomic_inc_not_zero(&result->sk_refcnt)))
+			result = NULL;
+		else if (unlikely(compute_score(result, net, hnum, daddr,
+				  dif) < hiscore)) {
+			sock_put(result);
+			goto begin;
+		}
 	}
-	spin_unlock(&ilb->lock);
-	return sk;
+	rcu_read_unlock();
+	return result;
 }
 EXPORT_SYMBOL_GPL(__inet_lookup_listener);
 
@@ -370,7 +371,7 @@ static void __inet_hash(struct sock *sk)
 	ilb = &hashinfo->listening_hash[inet_sk_listen_hashfn(sk)];
 
 	spin_lock(&ilb->lock);
-	__sk_add_node(sk, &ilb->head);
+	__sk_nulls_add_node_rcu(sk, &ilb->head);
 	sock_prot_inuse_add(sock_net(sk), sk->sk_prot, 1);
 	spin_unlock(&ilb->lock);
 }
@@ -388,26 +389,22 @@ EXPORT_SYMBOL_GPL(inet_hash);
 void inet_unhash(struct sock *sk)
 {
 	struct inet_hashinfo *hashinfo = sk->sk_prot->h.hashinfo;
+	spinlock_t *lock;
+	int done;
 
 	if (sk_unhashed(sk))
 		return;
 
-	if (sk->sk_state == TCP_LISTEN) {
-		struct inet_listen_hashbucket *ilb;
+	if (sk->sk_state == TCP_LISTEN)
+		lock = &hashinfo->listening_hash[inet_sk_listen_hashfn(sk)].lock;
+	else
+		lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
 
-		ilb = &hashinfo->listening_hash[inet_sk_listen_hashfn(sk)];
-		spin_lock_bh(&ilb->lock);
-		if (__sk_del_node_init(sk))
-			sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
-		spin_unlock_bh(&ilb->lock);
-	} else {
-		spinlock_t *lock = inet_ehash_lockp(hashinfo, sk->sk_hash);
-
-		spin_lock_bh(lock);
-		if (__sk_nulls_del_node_init_rcu(sk))
-			sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
-		spin_unlock_bh(lock);
-	}
+	spin_lock_bh(lock);
+	done =__sk_nulls_del_node_init_rcu(sk);
+	spin_unlock_bh(lock);
+	if (done)
+		sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
 }
 EXPORT_SYMBOL_GPL(inet_unhash);
 
@@ -526,8 +523,11 @@ void inet_hashinfo_init(struct inet_hashinfo *h)
 {
 	int i;
 
-	for (i = 0; i < INET_LHTABLE_SIZE; i++)
+	for (i = 0; i < INET_LHTABLE_SIZE; i++) {
 		spin_lock_init(&h->listening_hash[i].lock);
+		INIT_HLIST_NULLS_HEAD(&h->listening_hash[i].head,
+				      i + LISTENING_NULLS_BASE);
+		}
 }
 
 EXPORT_SYMBOL_GPL(inet_hashinfo_init);
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index a81caa1be0c..cab2458f86f 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1868,7 +1868,7 @@ static inline struct inet_timewait_sock *tw_next(struct inet_timewait_sock *tw)
 static void *listening_get_next(struct seq_file *seq, void *cur)
 {
 	struct inet_connection_sock *icsk;
-	struct hlist_node *node;
+	struct hlist_nulls_node *node;
 	struct sock *sk = cur;
 	struct inet_listen_hashbucket *ilb;
 	struct tcp_iter_state *st = seq->private;
@@ -1878,7 +1878,7 @@ static void *listening_get_next(struct seq_file *seq, void *cur)
 		st->bucket = 0;
 		ilb = &tcp_hashinfo.listening_hash[0];
 		spin_lock_bh(&ilb->lock);
-		sk = sk_head(&ilb->head);
+		sk = sk_nulls_head(&ilb->head);
 		goto get_sk;
 	}
 	ilb = &tcp_hashinfo.listening_hash[st->bucket];
@@ -1914,7 +1914,7 @@ get_req:
 		sk = sk_next(sk);
 	}
 get_sk:
-	sk_for_each_from(sk, node) {
+	sk_nulls_for_each_from(sk, node) {
 		if (sk->sk_family == st->family && net_eq(sock_net(sk), net)) {
 			cur = sk;
 			goto out;
@@ -1935,7 +1935,7 @@ start_req:
 	if (++st->bucket < INET_LHTABLE_SIZE) {
 		ilb = &tcp_hashinfo.listening_hash[st->bucket];
 		spin_lock_bh(&ilb->lock);
-		sk = sk_head(&ilb->head);
+		sk = sk_nulls_head(&ilb->head);
 		goto get_sk;
 	}
 	cur = NULL;
diff --git a/net/ipv6/inet6_hashtables.c b/net/ipv6/inet6_hashtables.c
index e0fd68187f8..8fe267feb81 100644
--- a/net/ipv6/inet6_hashtables.c
+++ b/net/ipv6/inet6_hashtables.c
@@ -33,7 +33,7 @@ void __inet6_hash(struct sock *sk)
 
 		ilb = &hashinfo->listening_hash[inet_sk_listen_hashfn(sk)];
 		spin_lock(&ilb->lock);
-		__sk_add_node(sk, &ilb->head);
+		__sk_nulls_add_node_rcu(sk, &ilb->head);
 		spin_unlock(&ilb->lock);
 	} else {
 		unsigned int hash;
@@ -118,47 +118,71 @@ out:
 }
 EXPORT_SYMBOL(__inet6_lookup_established);
 
+static int inline compute_score(struct sock *sk, struct net *net,
+				const unsigned short hnum,
+				const struct in6_addr *daddr,
+				const int dif)
+{
+	int score = -1;
+
+	if (net_eq(sock_net(sk), net) && inet_sk(sk)->num == hnum &&
+	    sk->sk_family == PF_INET6) {
+		const struct ipv6_pinfo *np = inet6_sk(sk);
+
+		score = 1;
+		if (!ipv6_addr_any(&np->rcv_saddr)) {
+			if (!ipv6_addr_equal(&np->rcv_saddr, daddr))
+				return -1;
+			score++;
+		}
+		if (sk->sk_bound_dev_if) {
+			if (sk->sk_bound_dev_if != dif)
+				return -1;
+			score++;
+		}
+	}
+	return score;
+}
+
 struct sock *inet6_lookup_listener(struct net *net,
 		struct inet_hashinfo *hashinfo, const struct in6_addr *daddr,
 		const unsigned short hnum, const int dif)
 {
 	struct sock *sk;
-	const struct hlist_node *node;
-	struct sock *result = NULL;
-	int score, hiscore = 0;
-	struct inet_listen_hashbucket *ilb;
-
-	ilb = &hashinfo->listening_hash[inet_lhashfn(net, hnum)];
-	spin_lock(&ilb->lock);
-	sk_for_each(sk, node, &ilb->head) {
-		if (net_eq(sock_net(sk), net) && inet_sk(sk)->num == hnum &&
-				sk->sk_family == PF_INET6) {
-			const struct ipv6_pinfo *np = inet6_sk(sk);
-
-			score = 1;
-			if (!ipv6_addr_any(&np->rcv_saddr)) {
-				if (!ipv6_addr_equal(&np->rcv_saddr, daddr))
-					continue;
-				score++;
-			}
-			if (sk->sk_bound_dev_if) {
-				if (sk->sk_bound_dev_if != dif)
-					continue;
-				score++;
-			}
-			if (score == 3) {
-				result = sk;
-				break;
-			}
-			if (score > hiscore) {
-				hiscore = score;
-				result = sk;
-			}
+	const struct hlist_nulls_node *node;
+	struct sock *result;
+	int score, hiscore;
+	unsigned int hash = inet_lhashfn(net, hnum);
+	struct inet_listen_hashbucket *ilb = &hashinfo->listening_hash[hash];
+
+	rcu_read_lock();
+begin:
+	result = NULL;
+	hiscore = -1;
+	sk_nulls_for_each(sk, node, &ilb->head) {
+		score = compute_score(sk, net, hnum, daddr, dif);
+		if (score > hiscore) {
+			hiscore = score;
+			result = sk;
 		}
 	}
-	if (result)
-		sock_hold(result);
-	spin_unlock(&ilb->lock);
+	/*
+	 * if the nulls value we got at the end of this lookup is
+	 * not the expected one, we must restart lookup.
+	 * We probably met an item that was moved to another chain.
+	 */
+	if (get_nulls_value(node) != hash + LISTENING_NULLS_BASE)
+		goto begin;
+	if (result) {
+		if (unlikely(!atomic_inc_not_zero(&result->sk_refcnt)))
+			result = NULL;
+		else if (unlikely(compute_score(result, net, hnum, daddr,
+				  dif) < hiscore)) {
+			sock_put(result);
+			goto begin;
+		}
+	}
+	rcu_read_unlock();
 	return result;
 }
 
-- 
cgit v1.2.3-70-g09d2