netfilter: ipset: Prepare the ipset core to use RCU at set level

Replace rwlock_t with spinlock_t in "struct ip_set" and change the locking
accordingly. Convert the comment extension into an rcu-avare object. Also,
simplify the timeout routines.

Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>

3 files changed, 44 insertions, 28 deletions

diff --git a/include/linux/netfilter/ipset/ip_set.h b/include/linux/netfilter/ipset/ip_set.h
index 5674b6ac6646..19b4969a25fe 100644
--- a/include/linux/netfilter/ipset/ip_set.h
+++ b/include/linux/netfilter/ipset/ip_set.h
@@ -108,8 +108,13 @@ struct ip_set_counter {
 	atomic64_t packets;
 };
 
+struct ip_set_comment_rcu {
+	struct rcu_head rcu;
+	char str[0];
+};
+
 struct ip_set_comment {
-	char *str;
+	struct ip_set_comment_rcu __rcu *c;
 };
 
 struct ip_set_skbinfo {
@@ -226,7 +231,7 @@ struct ip_set {
 	/* The name of the set */
 	char name[IPSET_MAXNAMELEN];
 	/* Lock protecting the set data */
-	rwlock_t lock;
+	spinlock_t lock;
 	/* References to the set */
 	u32 ref;
 	/* The core set type */
diff --git a/include/linux/netfilter/ipset/ip_set_comment.h b/include/linux/netfilter/ipset/ip_set_comment.h
index 21217ea008d7..8d0248525957 100644
--- a/include/linux/netfilter/ipset/ip_set_comment.h
+++ b/include/linux/netfilter/ipset/ip_set_comment.h
@@ -16,41 +16,57 @@ ip_set_comment_uget(struct nlattr *tb)
 	return nla_data(tb);
 }
 
+/* Called from uadd only, protected by the set spinlock.
+ * The kadt functions don't use the comment extensions in any way.
+ */
 static inline void
 ip_set_init_comment(struct ip_set_comment *comment,
 		    const struct ip_set_ext *ext)
 {
+	struct ip_set_comment_rcu *c = rcu_dereference_protected(comment->c, 1);
 	size_t len = ext->comment ? strlen(ext->comment) : 0;
 
-	if (unlikely(comment->str)) {
-		kfree(comment->str);
-		comment->str = NULL;
+	if (unlikely(c)) {
+		kfree_rcu(c, rcu);
+		rcu_assign_pointer(comment->c, NULL);
 	}
 	if (!len)
 		return;
 	if (unlikely(len > IPSET_MAX_COMMENT_SIZE))
 		len = IPSET_MAX_COMMENT_SIZE;
-	comment->str = kzalloc(len + 1, GFP_ATOMIC);
-	if (unlikely(!comment->str))
+	c = kzalloc(sizeof(*c) + len + 1, GFP_ATOMIC);
+	if (unlikely(!c))
 		return;
-	strlcpy(comment->str, ext->comment, len + 1);
+	strlcpy(c->str, ext->comment, len + 1);
+	rcu_assign_pointer(comment->c, c);
 }
 
+/* Used only when dumping a set, protected by rcu_read_lock_bh() */
 static inline int
 ip_set_put_comment(struct sk_buff *skb, struct ip_set_comment *comment)
 {
-	if (!comment->str)
+	struct ip_set_comment_rcu *c = rcu_dereference_bh(comment->c);
+
+	if (!c)
 		return 0;
-	return nla_put_string(skb, IPSET_ATTR_COMMENT, comment->str);
+	return nla_put_string(skb, IPSET_ATTR_COMMENT, c->str);
 }
 
+/* Called from uadd/udel, flush or the garbage collectors protected
+ * by the set spinlock.
+ * Called when the set is destroyed and when there can't be any user
+ * of the set data anymore.
+ */
 static inline void
 ip_set_comment_free(struct ip_set_comment *comment)
 {
-	if (unlikely(!comment->str))
+	struct ip_set_comment_rcu *c;
+
+	c = rcu_dereference_protected(comment->c, 1);
+	if (unlikely(!c))
 		return;
-	kfree(comment->str);
-	comment->str = NULL;
+	kfree_rcu(c, rcu);
+	rcu_assign_pointer(comment->c, NULL);
 }
 
 #endif
diff --git a/include/linux/netfilter/ipset/ip_set_timeout.h b/include/linux/netfilter/ipset/ip_set_timeout.h
index 3c8842bcedaa..1d6a935c1ac5 100644
--- a/include/linux/netfilter/ipset/ip_set_timeout.h
+++ b/include/linux/netfilter/ipset/ip_set_timeout.h
@@ -40,31 +40,26 @@ ip_set_timeout_uget(struct nlattr *tb)
 }
 
 static inline bool
-ip_set_timeout_test(unsigned long timeout)
+ip_set_timeout_expired(unsigned long *t)
 {
-	return timeout == IPSET_ELEM_PERMANENT ||
-	       time_is_after_jiffies(timeout);
-}
-
-static inline bool
-ip_set_timeout_expired(unsigned long *timeout)
-{
-	return *timeout != IPSET_ELEM_PERMANENT &&
-	       time_is_before_jiffies(*timeout);
+	return *t != IPSET_ELEM_PERMANENT && time_is_before_jiffies(*t);
 }
 
 static inline void
-ip_set_timeout_set(unsigned long *timeout, u32 t)
+ip_set_timeout_set(unsigned long *timeout, u32 value)
 {
-	if (!t) {
+	unsigned long t;
+
+	if (!value) {
 		*timeout = IPSET_ELEM_PERMANENT;
 		return;
 	}
 
-	*timeout = msecs_to_jiffies(t * MSEC_PER_SEC) + jiffies;
-	if (*timeout == IPSET_ELEM_PERMANENT)
+	t = msecs_to_jiffies(value * MSEC_PER_SEC) + jiffies;
+	if (t == IPSET_ELEM_PERMANENT)
 		/* Bingo! :-) */
-		(*timeout)--;
+		t--;
+	*timeout = t;
 }
 
 static inline u32