neighbor: Improve garbage collection

The existing garbage collection algorithm has a number of problems:

1. The gc algorithm will not evict PERMANENT entries as those entries
   are managed by userspace, yet the existing algorithm walks the entire
   hash table which means it always considers PERMANENT entries when
   looking for entries to evict. In some use cases (e.g., EVPN) there
   can be tens of thousands of PERMANENT entries leading to wasted
   CPU cycles when gc kicks in. As an example, with 32k permanent
   entries, neigh_alloc has been observed taking more than 4 msec per
   invocation.

2. Currently, when the number of neighbor entries hits gc_thresh2 and
   the last flush for the table was more than 5 seconds ago gc kicks in
   walks the entire hash table evicting *all* entries not in PERMANENT
   or REACHABLE state and not marked as externally learned. There is no
   discriminator on when the neigh entry was created or if it just moved
   from REACHABLE to another NUD_VALID state (e.g., NUD_STALE).

   It is possible for entries to be created or for established neighbor
   entries to be moved to STALE (e.g., an external node sends an ARP
   request) right before the 5 second window lapses:

        -----|---------x|----------|-----
            t-5         t         t+5

   If that happens those entries are evicted during gc causing unnecessary
   thrashing on neighbor entries and userspace caches trying to track them.

   Further, this contradicts the description of gc_thresh2 which says
   "Entries older than 5 seconds will be cleared".

   One workaround is to make gc_thresh2 == gc_thresh3 but that negates the
   whole point of having separate thresholds.

3. Clearing *all* neigh non-PERMANENT/REACHABLE/externally learned entries
   when gc_thresh2 is exceeded is over kill and contributes to trashing
   especially during startup.

This patch addresses these problems as follows:

1. Use of a separate list_head to track entries that can be garbage
   collected along with a separate counter. PERMANENT entries are not
   added to this list.

   The gc_thresh parameters are only compared to the new counter, not the
   total entries in the table. The forced_gc function is updated to only
   walk this new gc_list looking for entries to evict.

2. Entries are added to the list head at the tail and removed from the
   front.

3. Entries are only evicted if they were last updated more than 5 seconds
   ago, adhering to the original intent of gc_thresh2.

4. Forced gc is stopped once the number of gc_entries drops below
   gc_thresh2.

5. Since gc checks do not apply to PERMANENT entries, gc levels are skipped
   when allocating a new neighbor for a PERMANENT entry. By extension this
   means there are no explicit limits on the number of PERMANENT entries
   that can be created, but this is no different than FIB entries or FDB
   entries.

Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

3 files changed, 90 insertions, 36 deletions

diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt
index af2a69439b93..acdfb5d2bcaa 100644
--- a/Documentation/networking/ip-sysctl.txt
+++ b/Documentation/networking/ip-sysctl.txt
@@ -108,8 +108,8 @@ neigh/default/gc_thresh2 - INTEGER
 	Default: 512
 
 neigh/default/gc_thresh3 - INTEGER
-	Maximum number of neighbor entries allowed.  Increase this
-	when using large numbers of interfaces and when communicating
+	Maximum number of non-PERMANENT neighbor entries allowed.  Increase
+	this when using large numbers of interfaces and when communicating
 	with large numbers of directly-connected peers.
 	Default: 1024
 
diff --git a/include/net/neighbour.h b/include/net/neighbour.h
index f58b384aa6c9..6c13072910ab 100644
--- a/include/net/neighbour.h
+++ b/include/net/neighbour.h
@@ -154,6 +154,7 @@ struct neighbour {
 	struct hh_cache		hh;
 	int			(*output)(struct neighbour *, struct sk_buff *);
 	const struct neigh_ops	*ops;
+	struct list_head	gc_list;
 	struct rcu_head		rcu;
 	struct net_device	*dev;
 	u8			primary_key[0];
@@ -214,6 +215,8 @@ struct neigh_table {
 	struct timer_list 	proxy_timer;
 	struct sk_buff_head	proxy_queue;
 	atomic_t		entries;
+	atomic_t		gc_entries;
+	struct list_head	gc_list;
 	rwlock_t		lock;
 	unsigned long		last_rand;
 	struct neigh_statistics	__percpu *stats;
diff --git a/net/core/neighbour.c b/net/core/neighbour.c
index 6d479b5562be..c3b58712e98b 100644
--- a/net/core/neighbour.c
+++ b/net/core/neighbour.c
@@ -118,6 +118,34 @@ unsigned long neigh_rand_reach_time(unsigned long base)
 }
 EXPORT_SYMBOL(neigh_rand_reach_time);
 
+static void neigh_mark_dead(struct neighbour *n)
+{
+	n->dead = 1;
+	if (!list_empty(&n->gc_list)) {
+		list_del_init(&n->gc_list);
+		atomic_dec(&n->tbl->gc_entries);
+	}
+}
+
+static void neigh_change_state(struct neighbour *n, u8 new)
+{
+	bool on_gc_list = !list_empty(&n->gc_list);
+	bool new_is_perm = new & NUD_PERMANENT;
+
+	n->nud_state = new;
+
+	/* remove from the gc list if new state is permanent;
+	 * add to the gc list if new state is not permanent
+	 */
+	if (new_is_perm && on_gc_list) {
+		list_del_init(&n->gc_list);
+		atomic_dec(&n->tbl->gc_entries);
+	} else if (!new_is_perm && !on_gc_list) {
+		/* add entries to the tail; cleaning removes from the front */
+		list_add_tail(&n->gc_list, &n->tbl->gc_list);
+		atomic_inc(&n->tbl->gc_entries);
+	}
+}
 
 static bool neigh_del(struct neighbour *n, __u8 state, __u8 flags,
 		      struct neighbour __rcu **np, struct neigh_table *tbl)
@@ -132,7 +160,7 @@ static bool neigh_del(struct neighbour *n, __u8 state, __u8 flags,
 		neigh = rcu_dereference_protected(n->next,
 						  lockdep_is_held(&tbl->lock));
 		rcu_assign_pointer(*np, neigh);
-		n->dead = 1;
+		neigh_mark_dead(n);
 		retval = true;
 	}
 	write_unlock(&n->lock);
@@ -166,32 +194,31 @@ bool neigh_remove_one(struct neighbour *ndel, struct neigh_table *tbl)
 
 static int neigh_forced_gc(struct neigh_table *tbl)
 {
+	int max_clean = atomic_read(&tbl->gc_entries) - tbl->gc_thresh2;
+	unsigned long tref = jiffies - 5 * HZ;
+	u8 flags = NTF_EXT_LEARNED;
+	struct neighbour *n, *tmp;
+	u8 state = NUD_PERMANENT;
 	int shrunk = 0;
-	int i;
-	struct neigh_hash_table *nht;
 
 	NEIGH_CACHE_STAT_INC(tbl, forced_gc_runs);
 
 	write_lock_bh(&tbl->lock);
-	nht = rcu_dereference_protected(tbl->nht,
-					lockdep_is_held(&tbl->lock));
-	for (i = 0; i < (1 << nht->hash_shift); i++) {
-		struct neighbour *n;
-		struct neighbour __rcu **np;
 
-		np = &nht->hash_buckets[i];
-		while ((n = rcu_dereference_protected(*np,
-					lockdep_is_held(&tbl->lock))) != NULL) {
-			/* Neighbour record may be discarded if:
-			 * - nobody refers to it.
-			 * - it is not permanent
-			 */
-			if (neigh_del(n, NUD_PERMANENT, NTF_EXT_LEARNED, np,
-				      tbl)) {
-				shrunk = 1;
-				continue;
-			}
-			np = &n->next;
+	list_for_each_entry_safe(n, tmp, &tbl->gc_list, gc_list) {
+		if (refcount_read(&n->refcnt) == 1) {
+			bool remove = false;
+
+			write_lock(&n->lock);
+			if (!(n->nud_state & state) && !(n->flags & flags) &&
+			    time_after(tref, n->updated))
+				remove = true;
+			write_unlock(&n->lock);
+
+			if (remove && neigh_remove_one(n, tbl))
+				shrunk++;
+			if (shrunk >= max_clean)
+				break;
 		}
 	}
 
@@ -260,8 +287,7 @@ static void neigh_flush_dev(struct neigh_table *tbl, struct net_device *dev,
 						lockdep_is_held(&tbl->lock)));
 			write_lock(&n->lock);
 			neigh_del_timer(n);
-			n->dead = 1;
-
+			neigh_mark_dead(n);
 			if (refcount_read(&n->refcnt) != 1) {
 				/* The most unpleasant situation.
 				   We must destroy neighbour entry,
@@ -321,13 +347,18 @@ int neigh_ifdown(struct neigh_table *tbl, struct net_device *dev)
 }
 EXPORT_SYMBOL(neigh_ifdown);
 
-static struct neighbour *neigh_alloc(struct neigh_table *tbl, struct net_device *dev)
+static struct neighbour *neigh_alloc(struct neigh_table *tbl,
+				     struct net_device *dev,
+				     bool permanent)
 {
 	struct neighbour *n = NULL;
 	unsigned long now = jiffies;
 	int entries;
 
-	entries = atomic_inc_return(&tbl->entries) - 1;
+	if (permanent)
+		goto do_alloc;
+
+	entries = atomic_inc_return(&tbl->gc_entries) - 1;
 	if (entries >= tbl->gc_thresh3 ||
 	    (entries >= tbl->gc_thresh2 &&
 	     time_after(now, tbl->last_flush + 5 * HZ))) {
@@ -340,6 +371,7 @@ static struct neighbour *neigh_alloc(struct neigh_table *tbl, struct net_device
 		}
 	}
 
+do_alloc:
 	n = kzalloc(tbl->entry_size + dev->neigh_priv_len, GFP_ATOMIC);
 	if (!n)
 		goto out_entries;
@@ -358,11 +390,19 @@ static struct neighbour *neigh_alloc(struct neigh_table *tbl, struct net_device
 	n->tbl		  = tbl;
 	refcount_set(&n->refcnt, 1);
 	n->dead		  = 1;
+
+	if (!permanent)
+		list_add_tail(&n->gc_list, &n->tbl->gc_list);
+	else
+		INIT_LIST_HEAD(&n->gc_list);
+
+	atomic_inc(&tbl->entries);
 out:
 	return n;
 
 out_entries:
-	atomic_dec(&tbl->entries);
+	if (!permanent)
+		atomic_dec(&tbl->gc_entries);
 	goto out;
 }
 
@@ -505,13 +545,15 @@ struct neighbour *neigh_lookup_nodev(struct neigh_table *tbl, struct net *net,
 }
 EXPORT_SYMBOL(neigh_lookup_nodev);
 
-struct neighbour *__neigh_create(struct neigh_table *tbl, const void *pkey,
-				 struct net_device *dev, bool want_ref)
+static struct neighbour *___neigh_create(struct neigh_table *tbl,
+					 const void *pkey,
+					 struct net_device *dev,
+					 bool permanent, bool want_ref)
 {
+	struct neighbour *n1, *rc, *n = neigh_alloc(tbl, dev, permanent);
 	u32 hash_val;
 	unsigned int key_len = tbl->key_len;
 	int error;
-	struct neighbour *n1, *rc, *n = neigh_alloc(tbl, dev);
 	struct neigh_hash_table *nht;
 
 	if (!n) {
@@ -591,6 +633,12 @@ out_neigh_release:
 	neigh_release(n);
 	goto out;
 }
+
+struct neighbour *__neigh_create(struct neigh_table *tbl, const void *pkey,
+				 struct net_device *dev, bool want_ref)
+{
+	return ___neigh_create(tbl, pkey, dev, false, want_ref);
+}
 EXPORT_SYMBOL(__neigh_create);
 
 static u32 pneigh_hash(const void *pkey, unsigned int key_len)
@@ -854,7 +902,7 @@ static void neigh_periodic_work(struct work_struct *work)
 			    (state == NUD_FAILED ||
 			     time_after(jiffies, n->used + NEIGH_VAR(n->parms, GC_STALETIME)))) {
 				*np = n->next;
-				n->dead = 1;
+				neigh_mark_dead(n);
 				write_unlock(&n->lock);
 				neigh_cleanup_and_release(n);
 				continue;
@@ -1167,7 +1215,7 @@ static int __neigh_update(struct neighbour *neigh, const u8 *lladdr,
 		neigh_del_timer(neigh);
 		if (old & NUD_CONNECTED)
 			neigh_suspect(neigh);
-		neigh->nud_state = new;
+		neigh_change_state(neigh, new);
 		err = 0;
 		notify = old & NUD_VALID;
 		if ((old & (NUD_INCOMPLETE | NUD_PROBE)) &&
@@ -1246,7 +1294,7 @@ static int __neigh_update(struct neighbour *neigh, const u8 *lladdr,
 						((new & NUD_REACHABLE) ?
 						 neigh->parms->reachable_time :
 						 0)));
-		neigh->nud_state = new;
+		neigh_change_state(neigh, new);
 		notify = 1;
 	}
 
@@ -1582,6 +1630,7 @@ void neigh_table_init(int index, struct neigh_table *tbl)
 	unsigned long phsize;
 
 	INIT_LIST_HEAD(&tbl->parms_list);
+	INIT_LIST_HEAD(&tbl->gc_list);
 	list_add(&tbl->parms.list, &tbl->parms_list);
 	write_pnet(&tbl->parms.net, &init_net);
 	refcount_set(&tbl->parms.refcnt, 1);
@@ -1813,7 +1862,9 @@ static int neigh_add(struct sk_buff *skb, struct nlmsghdr *nlh,
 			goto out;
 		}
 
-		neigh = __neigh_lookup_errno(tbl, dst, dev);
+		neigh = ___neigh_create(tbl, dst, dev,
+					ndm->ndm_state & NUD_PERMANENT,
+					true);
 		if (IS_ERR(neigh)) {
 			err = PTR_ERR(neigh);
 			goto out;
@@ -2654,7 +2705,7 @@ void __neigh_for_each_release(struct neigh_table *tbl,
 				rcu_assign_pointer(*np,
 					rcu_dereference_protected(n->next,
 						lockdep_is_held(&tbl->lock)));
-				n->dead = 1;
+				neigh_mark_dead(n);
 			} else
 				np = &n->next;
 			write_unlock(&n->lock);