Introduce RCU in all set types instead of rwlock per set

Performance is tested by Jesper Dangaard Brouer:

Simple drop in FORWARD
~~~~~~~~~~~~~~~~~~~~~~

Dropping via simple iptables net-mask match::

 iptables -t raw -N simple || iptables -t raw -F simple
 iptables -t raw -I simple  -s 198.18.0.0/15 -j DROP
 iptables -t raw -D PREROUTING -j simple
 iptables -t raw -I PREROUTING -j simple

Drop performance in "raw": 11.3Mpps

Generator: sending 12.2Mpps (tx:12264083 pps)

Drop via original ipset in RAW table
~~~~~~~~~~~~~~~~~~~~~~~~~~~

Create a set with lots of elements::

 sudo ./ipset destroy test
 echo "create test hash:ip hashsize 65536" > test.set
 for x in `seq 0 255`; do
    for y in `seq 0 255`; do
        echo "add test 198.18.$x.$y" >> test.set
    done
 done
 sudo ./ipset restore < test.set

Dropping via ipset::

 iptables -t raw -F
 iptables -t raw -N net198 || iptables -t raw -F net198
 iptables -t raw -I net198 -m set --match-set test src -j DROP
 iptables -t raw -I PREROUTING -j net198

Drop performance in "raw" with ipset: 8Mpps

Perf report numbers ipset drop in "raw"::

 +   24.65%  ksoftirqd/1  [ip_set]           [k] ip_set_test
 -   21.42%  ksoftirqd/1  [kernel.kallsyms]  [k] _raw_read_lock_bh
    - _raw_read_lock_bh
       + 99.88% ip_set_test
 -   19.42%  ksoftirqd/1  [kernel.kallsyms]  [k] _raw_read_unlock_bh
    - _raw_read_unlock_bh
       + 99.72% ip_set_test
 +    4.31%  ksoftirqd/1  [ip_set_hash_ip]   [k] hash_ip4_kadt
 +    2.27%  ksoftirqd/1  [ixgbe]            [k] ixgbe_fetch_rx_buffer
 +    2.18%  ksoftirqd/1  [ip_tables]        [k] ipt_do_table
 +    1.81%  ksoftirqd/1  [ip_set_hash_ip]   [k] hash_ip4_test
 +    1.61%  ksoftirqd/1  [kernel.kallsyms]  [k] __netif_receive_skb_core
 +    1.44%  ksoftirqd/1  [kernel.kallsyms]  [k] build_skb
 +    1.42%  ksoftirqd/1  [kernel.kallsyms]  [k] ip_rcv
 +    1.36%  ksoftirqd/1  [kernel.kallsyms]  [k] __local_bh_enable_ip
 +    1.16%  ksoftirqd/1  [kernel.kallsyms]  [k] dev_gro_receive
 +    1.09%  ksoftirqd/1  [kernel.kallsyms]  [k] __rcu_read_unlock
 +    0.96%  ksoftirqd/1  [ixgbe]            [k] ixgbe_clean_rx_irq
 +    0.95%  ksoftirqd/1  [kernel.kallsyms]  [k] __netdev_alloc_frag
 +    0.88%  ksoftirqd/1  [kernel.kallsyms]  [k] kmem_cache_alloc
 +    0.87%  ksoftirqd/1  [xt_set]           [k] set_match_v3
 +    0.85%  ksoftirqd/1  [kernel.kallsyms]  [k] inet_gro_receive
 +    0.83%  ksoftirqd/1  [kernel.kallsyms]  [k] nf_iterate
 +    0.76%  ksoftirqd/1  [kernel.kallsyms]  [k] put_compound_page
 +    0.75%  ksoftirqd/1  [kernel.kallsyms]  [k] __rcu_read_lock

Drop via ipset in RAW table with RCU-locking
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

With RCU locking, the RW-lock is gone.

Drop performance in "raw" with ipset with RCU-locking: 11.3Mpps

Performance-tested-by: Jesper Dangaard Brouer <brouer@redhat.com>

1 files changed, 18 insertions, 17 deletions

diff --git a/kernel/net/netfilter/ipset/ip_set_core.c b/kernel/net/netfilter/ipset/ip_set_core.c
index d37ee69..1f84e64 100644
--- a/kernel/net/netfilter/ipset/ip_set_core.c
+++ b/kernel/net/netfilter/ipset/ip_set_core.c
@@ -223,6 +223,7 @@ ip_set_type_register(struct ip_set_type *type)
 		 type->revision_min, type->revision_max);
 unlock:
 	ip_set_type_unlock();
+	synchronize_rcu();
 	return ret;
 }
 EXPORT_SYMBOL_GPL(ip_set_type_register);
@@ -502,16 +503,16 @@ ip_set_test(ip_set_id_t index, const struct sk_buff *skb,
 	    !(opt->family == set->family || set->family == NFPROTO_UNSPEC))
 		return 0;
 
-	read_lock_bh(&set->lock);
+	rcu_read_lock_bh();
 	ret = set->variant->kadt(set, skb, par, IPSET_TEST, opt);
-	read_unlock_bh(&set->lock);
+	rcu_read_unlock_bh();
 
 	if (ret == -EAGAIN) {
 		/* Type requests element to be completed */
 		pr_debug("element must be completed, ADD is triggered\n");
-		write_lock_bh(&set->lock);
+		spin_lock_bh(&set->lock);
 		set->variant->kadt(set, skb, par, IPSET_ADD, opt);
-		write_unlock_bh(&set->lock);
+		spin_unlock_bh(&set->lock);
 		ret = 1;
 	} else {
 		/* --return-nomatch: invert matched element */
@@ -541,9 +542,9 @@ ip_set_add(ip_set_id_t index, const struct sk_buff *skb,
 	    !(opt->family == set->family || set->family == NFPROTO_UNSPEC))
 		return -IPSET_ERR_TYPE_MISMATCH;
 
-	write_lock_bh(&set->lock);
+	spin_lock_bh(&set->lock);
 	ret = set->variant->kadt(set, skb, par, IPSET_ADD, opt);
-	write_unlock_bh(&set->lock);
+	spin_unlock_bh(&set->lock);
 
 	return ret;
 }
@@ -564,9 +565,9 @@ ip_set_del(ip_set_id_t index, const struct sk_buff *skb,
 	    !(opt->family == set->family || set->family == NFPROTO_UNSPEC))
 		return -IPSET_ERR_TYPE_MISMATCH;
 
-	write_lock_bh(&set->lock);
+	spin_lock_bh(&set->lock);
 	ret = set->variant->kadt(set, skb, par, IPSET_DEL, opt);
-	write_unlock_bh(&set->lock);
+	spin_unlock_bh(&set->lock);
 
 	return ret;
 }
@@ -851,7 +852,7 @@ ip_set_create(struct sock *ctnl, struct sk_buff *skb,
 	set = kzalloc(sizeof(struct ip_set), GFP_KERNEL);
 	if (!set)
 		return -ENOMEM;
-	rwlock_init(&set->lock);
+	spin_lock_init(&set->lock);
 	strlcpy(set->name, name, IPSET_MAXNAMELEN);
 	set->family = family;
 	set->revision = revision;
@@ -1030,9 +1031,9 @@ ip_set_flush_set(struct ip_set *set)
 {
 	pr_debug("set: %s\n",  set->name);
 
-	write_lock_bh(&set->lock);
+	spin_lock_bh(&set->lock);
 	set->variant->flush(set);
-	write_unlock_bh(&set->lock);
+	spin_unlock_bh(&set->lock);
 }
 
 static int
@@ -1333,9 +1334,9 @@ dump_last:
 				goto next_set;
 			/* Fall through and add elements */
 		default:
-			read_lock_bh(&set->lock);
+			rcu_read_lock_bh();
 			ret = set->variant->list(set, skb, cb);
-			read_unlock_bh(&set->lock);
+			rcu_read_unlock_bh();
 			if (!cb->args[IPSET_CB_ARG0])
 				/* Set is done, proceed with next one */
 				goto next_set;
@@ -1423,9 +1424,9 @@ call_ad(struct sock *ctnl, struct sk_buff *skb, struct ip_set *set,
 	bool eexist = flags & IPSET_FLAG_EXIST, retried = false;
 
 	do {
-		write_lock_bh(&set->lock);
+		spin_lock_bh(&set->lock);
 		ret = set->variant->uadt(set, tb, adt, &lineno, flags, retried);
-		write_unlock_bh(&set->lock);
+		spin_unlock_bh(&set->lock);
 		retried = true;
 	} while (ret == -EAGAIN &&
 		 set->variant->resize &&
@@ -1605,9 +1606,9 @@ ip_set_utest(struct sock *ctnl, struct sk_buff *skb,
 			     set->type->adt_policy))
 		return -IPSET_ERR_PROTOCOL;
 
-	read_lock_bh(&set->lock);
+	rcu_read_lock_bh();
 	ret = set->variant->uadt(set, tb, IPSET_TEST, NULL, 0, 0);
-	read_unlock_bh(&set->lock);
+	rcu_read_unlock_bh();
 	/* Userspace can't trigger element to be re-added */
 	if (ret == -EAGAIN)
 		ret = 1;