1 files changed, 82 insertions, 77 deletions

diff --git a/net/core/neighbour.c b/net/core/neighbour.c
index 8b6140e67e7f..e571007d083c 100644
--- a/net/core/neighbour.c
+++ b/net/core/neighbour.c
@@ -224,10 +224,13 @@ 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;
+	int max_clean = atomic_read(&tbl->gc_entries) -
+			READ_ONCE(tbl->gc_thresh2);
+	u64 tmax = ktime_get_ns() + NSEC_PER_MSEC;
 	unsigned long tref = jiffies - 5 * HZ;
 	struct neighbour *n, *tmp;
 	int shrunk = 0;
+	int loop = 0;
 
 	NEIGH_CACHE_STAT_INC(tbl, forced_gc_runs);
 
@@ -242,7 +245,7 @@ static int neigh_forced_gc(struct neigh_table *tbl)
 			    (n->nud_state == NUD_NOARP) ||
 			    (tbl->is_multicast &&
 			     tbl->is_multicast(n->primary_key)) ||
-			    time_after(tref, n->updated))
+			    !time_in_range(n->updated, tref, jiffies))
 				remove = true;
 			write_unlock(&n->lock);
 
@@ -250,11 +253,16 @@ static int neigh_forced_gc(struct neigh_table *tbl)
 				shrunk++;
 			if (shrunk >= max_clean)
 				break;
+			if (++loop == 16) {
+				if (ktime_get_ns() > tmax)
+					goto unlock;
+				loop = 0;
+			}
 		}
 	}
 
-	tbl->last_flush = jiffies;
-
+	WRITE_ONCE(tbl->last_flush, jiffies);
+unlock:
 	write_unlock_bh(&tbl->lock);
 
 	return shrunk;
@@ -262,7 +270,17 @@ static int neigh_forced_gc(struct neigh_table *tbl)
 
 static void neigh_add_timer(struct neighbour *n, unsigned long when)
 {
+	/* Use safe distance from the jiffies - LONG_MAX point while timer
+	 * is running in DELAY/PROBE state but still show to user space
+	 * large times in the past.
+	 */
+	unsigned long mint = jiffies - (LONG_MAX - 86400 * HZ);
+
 	neigh_hold(n);
+	if (!time_in_range(n->confirmed, mint, jiffies))
+		n->confirmed = mint;
+	if (time_before(n->used, n->confirmed))
+		n->used = n->confirmed;
 	if (unlikely(mod_timer(&n->timer, when))) {
 		printk("NEIGH: BUG, double timer add, state is %x\n",
 		       n->nud_state);
@@ -280,11 +298,26 @@ static int neigh_del_timer(struct neighbour *n)
 	return 0;
 }
 
-static void pneigh_queue_purge(struct sk_buff_head *list)
+static void pneigh_queue_purge(struct sk_buff_head *list, struct net *net)
 {
+	struct sk_buff_head tmp;
+	unsigned long flags;
 	struct sk_buff *skb;
 
-	while ((skb = skb_dequeue(list)) != NULL) {
+	skb_queue_head_init(&tmp);
+	spin_lock_irqsave(&list->lock, flags);
+	skb = skb_peek(list);
+	while (skb != NULL) {
+		struct sk_buff *skb_next = skb_peek_next(skb, list);
+		if (net == NULL || net_eq(dev_net(skb->dev), net)) {
+			__skb_unlink(skb, list);
+			__skb_queue_tail(&tmp, skb);
+		}
+		skb = skb_next;
+	}
+	spin_unlock_irqrestore(&list->lock, flags);
+
+	while ((skb = __skb_dequeue(&tmp))) {
 		dev_put(skb->dev);
 		kfree_skb(skb);
 	}
@@ -358,9 +391,9 @@ static int __neigh_ifdown(struct neigh_table *tbl, struct net_device *dev,
 	write_lock_bh(&tbl->lock);
 	neigh_flush_dev(tbl, dev, skip_perm);
 	pneigh_ifdown_and_unlock(tbl, dev);
-
-	del_timer_sync(&tbl->proxy_timer);
-	pneigh_queue_purge(&tbl->proxy_queue);
+	pneigh_queue_purge(&tbl->proxy_queue, dev ? dev_net(dev) : NULL);
+	if (skb_queue_empty_lockless(&tbl->proxy_queue))
+		del_timer_sync(&tbl->proxy_timer);
 	return 0;
 }
 
@@ -384,17 +417,17 @@ static struct neighbour *neigh_alloc(struct neigh_table *tbl,
 {
 	struct neighbour *n = NULL;
 	unsigned long now = jiffies;
-	int entries;
+	int entries, gc_thresh3;
 
 	if (exempt_from_gc)
 		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))) {
-		if (!neigh_forced_gc(tbl) &&
-		    entries >= tbl->gc_thresh3) {
+	gc_thresh3 = READ_ONCE(tbl->gc_thresh3);
+	if (entries >= gc_thresh3 ||
+	    (entries >= READ_ONCE(tbl->gc_thresh2) &&
+	     time_after(now, READ_ONCE(tbl->last_flush) + 5 * HZ))) {
+		if (!neigh_forced_gc(tbl) && entries >= gc_thresh3) {
 			net_info_ratelimited("%s: neighbor table overflow!\n",
 					     tbl->id);
 			NEIGH_CACHE_STAT_INC(tbl, table_fulls);
@@ -546,37 +579,6 @@ struct neighbour *neigh_lookup(struct neigh_table *tbl, const void *pkey,
 }
 EXPORT_SYMBOL(neigh_lookup);
 
-struct neighbour *neigh_lookup_nodev(struct neigh_table *tbl, struct net *net,
-				     const void *pkey)
-{
-	struct neighbour *n;
-	unsigned int key_len = tbl->key_len;
-	u32 hash_val;
-	struct neigh_hash_table *nht;
-
-	NEIGH_CACHE_STAT_INC(tbl, lookups);
-
-	rcu_read_lock_bh();
-	nht = rcu_dereference_bh(tbl->nht);
-	hash_val = tbl->hash(pkey, NULL, nht->hash_rnd) >> (32 - nht->hash_shift);
-
-	for (n = rcu_dereference_bh(nht->hash_buckets[hash_val]);
-	     n != NULL;
-	     n = rcu_dereference_bh(n->next)) {
-		if (!memcmp(n->primary_key, pkey, key_len) &&
-		    net_eq(dev_net(n->dev), net)) {
-			if (!refcount_inc_not_zero(&n->refcnt))
-				n = NULL;
-			NEIGH_CACHE_STAT_INC(tbl, hits);
-			break;
-		}
-	}
-
-	rcu_read_unlock_bh();
-	return n;
-}
-EXPORT_SYMBOL(neigh_lookup_nodev);
-
 static struct neighbour *
 ___neigh_create(struct neigh_table *tbl, const void *pkey,
 		struct net_device *dev, u8 flags,
@@ -908,13 +910,14 @@ static void neigh_periodic_work(struct work_struct *work)
 
 	if (time_after(jiffies, tbl->last_rand + 300 * HZ)) {
 		struct neigh_parms *p;
-		tbl->last_rand = jiffies;
+
+		WRITE_ONCE(tbl->last_rand, jiffies);
 		list_for_each_entry(p, &tbl->parms_list, list)
 			p->reachable_time =
 				neigh_rand_reach_time(NEIGH_VAR(p, BASE_REACHABLE_TIME));
 	}
 
-	if (atomic_read(&tbl->entries) < tbl->gc_thresh1)
+	if (atomic_read(&tbl->entries) < READ_ONCE(tbl->gc_thresh1))
 		goto out;
 
 	for (i = 0 ; i < (1 << nht->hash_shift); i++) {
@@ -933,13 +936,17 @@ static void neigh_periodic_work(struct work_struct *work)
 				goto next_elt;
 			}
 
-			if (time_before(n->used, n->confirmed))
+			if (time_before(n->used, n->confirmed) &&
+			    time_is_before_eq_jiffies(n->confirmed))
 				n->used = n->confirmed;
 
 			if (refcount_read(&n->refcnt) == 1 &&
 			    (state == NUD_FAILED ||
-			     time_after(jiffies, n->used + NEIGH_VAR(n->parms, GC_STALETIME)))) {
-				*np = n->next;
+			     !time_in_range_open(jiffies, n->used,
+						 n->used + NEIGH_VAR(n->parms, GC_STALETIME)))) {
+				rcu_assign_pointer(*np,
+					rcu_dereference_protected(n->next,
+						lockdep_is_held(&tbl->lock)));
 				neigh_mark_dead(n);
 				write_unlock(&n->lock);
 				neigh_cleanup_and_release(n);
@@ -1741,7 +1748,7 @@ int neigh_table_clear(int index, struct neigh_table *tbl)
 	/* It is not clean... Fix it to unload IPv6 module safely */
 	cancel_delayed_work_sync(&tbl->gc_work);
 	del_timer_sync(&tbl->proxy_timer);
-	pneigh_queue_purge(&tbl->proxy_queue);
+	pneigh_queue_purge(&tbl->proxy_queue, NULL);
 	neigh_ifdown(tbl, NULL);
 	if (atomic_read(&tbl->entries))
 		pr_crit("neighbour leakage\n");
@@ -1773,9 +1780,6 @@ static struct neigh_table *neigh_find_table(int family)
 	case AF_INET6:
 		tbl = neigh_tables[NEIGH_ND_TABLE];
 		break;
-	case AF_DECnet:
-		tbl = neigh_tables[NEIGH_DN_TABLE];
-		break;
 	}
 
 	return tbl;
@@ -2055,15 +2059,16 @@ static int neightbl_fill_info(struct sk_buff *skb, struct neigh_table *tbl,
 	ndtmsg->ndtm_pad2   = 0;
 
 	if (nla_put_string(skb, NDTA_NAME, tbl->id) ||
-	    nla_put_msecs(skb, NDTA_GC_INTERVAL, tbl->gc_interval, NDTA_PAD) ||
-	    nla_put_u32(skb, NDTA_THRESH1, tbl->gc_thresh1) ||
-	    nla_put_u32(skb, NDTA_THRESH2, tbl->gc_thresh2) ||
-	    nla_put_u32(skb, NDTA_THRESH3, tbl->gc_thresh3))
+	    nla_put_msecs(skb, NDTA_GC_INTERVAL, READ_ONCE(tbl->gc_interval),
+			  NDTA_PAD) ||
+	    nla_put_u32(skb, NDTA_THRESH1, READ_ONCE(tbl->gc_thresh1)) ||
+	    nla_put_u32(skb, NDTA_THRESH2, READ_ONCE(tbl->gc_thresh2)) ||
+	    nla_put_u32(skb, NDTA_THRESH3, READ_ONCE(tbl->gc_thresh3)))
 		goto nla_put_failure;
 	{
 		unsigned long now = jiffies;
-		long flush_delta = now - tbl->last_flush;
-		long rand_delta = now - tbl->last_rand;
+		long flush_delta = now - READ_ONCE(tbl->last_flush);
+		long rand_delta = now - READ_ONCE(tbl->last_rand);
 		struct neigh_hash_table *nht;
 		struct ndt_config ndc = {
 			.ndtc_key_len		= tbl->key_len,
@@ -2071,7 +2076,7 @@ static int neightbl_fill_info(struct sk_buff *skb, struct neigh_table *tbl,
 			.ndtc_entries		= atomic_read(&tbl->entries),
 			.ndtc_last_flush	= jiffies_to_msecs(flush_delta),
 			.ndtc_last_rand		= jiffies_to_msecs(rand_delta),
-			.ndtc_proxy_qlen	= tbl->proxy_queue.qlen,
+			.ndtc_proxy_qlen	= READ_ONCE(tbl->proxy_queue.qlen),
 		};
 
 		rcu_read_lock_bh();
@@ -2094,17 +2099,17 @@ static int neightbl_fill_info(struct sk_buff *skb, struct neigh_table *tbl,
 			struct neigh_statistics	*st;
 
 			st = per_cpu_ptr(tbl->stats, cpu);
-			ndst.ndts_allocs		+= st->allocs;
-			ndst.ndts_destroys		+= st->destroys;
-			ndst.ndts_hash_grows		+= st->hash_grows;
-			ndst.ndts_res_failed		+= st->res_failed;
-			ndst.ndts_lookups		+= st->lookups;
-			ndst.ndts_hits			+= st->hits;
-			ndst.ndts_rcv_probes_mcast	+= st->rcv_probes_mcast;
-			ndst.ndts_rcv_probes_ucast	+= st->rcv_probes_ucast;
-			ndst.ndts_periodic_gc_runs	+= st->periodic_gc_runs;
-			ndst.ndts_forced_gc_runs	+= st->forced_gc_runs;
-			ndst.ndts_table_fulls		+= st->table_fulls;
+			ndst.ndts_allocs		+= READ_ONCE(st->allocs);
+			ndst.ndts_destroys		+= READ_ONCE(st->destroys);
+			ndst.ndts_hash_grows		+= READ_ONCE(st->hash_grows);
+			ndst.ndts_res_failed		+= READ_ONCE(st->res_failed);
+			ndst.ndts_lookups		+= READ_ONCE(st->lookups);
+			ndst.ndts_hits			+= READ_ONCE(st->hits);
+			ndst.ndts_rcv_probes_mcast	+= READ_ONCE(st->rcv_probes_mcast);
+			ndst.ndts_rcv_probes_ucast	+= READ_ONCE(st->rcv_probes_ucast);
+			ndst.ndts_periodic_gc_runs	+= READ_ONCE(st->periodic_gc_runs);
+			ndst.ndts_forced_gc_runs	+= READ_ONCE(st->forced_gc_runs);
+			ndst.ndts_table_fulls		+= READ_ONCE(st->table_fulls);
 		}
 
 		if (nla_put_64bit(skb, NDTA_STATS, sizeof(ndst), &ndst,
@@ -2328,16 +2333,16 @@ static int neightbl_set(struct sk_buff *skb, struct nlmsghdr *nlh,
 		goto errout_tbl_lock;
 
 	if (tb[NDTA_THRESH1])
-		tbl->gc_thresh1 = nla_get_u32(tb[NDTA_THRESH1]);
+		WRITE_ONCE(tbl->gc_thresh1, nla_get_u32(tb[NDTA_THRESH1]));
 
 	if (tb[NDTA_THRESH2])
-		tbl->gc_thresh2 = nla_get_u32(tb[NDTA_THRESH2]);
+		WRITE_ONCE(tbl->gc_thresh2, nla_get_u32(tb[NDTA_THRESH2]));
 
 	if (tb[NDTA_THRESH3])
-		tbl->gc_thresh3 = nla_get_u32(tb[NDTA_THRESH3]);
+		WRITE_ONCE(tbl->gc_thresh3, nla_get_u32(tb[NDTA_THRESH3]));
 
 	if (tb[NDTA_GC_INTERVAL])
-		tbl->gc_interval = nla_get_msecs(tb[NDTA_GC_INTERVAL]);
+		WRITE_ONCE(tbl->gc_interval, nla_get_msecs(tb[NDTA_GC_INTERVAL]));
 
 	err = 0;