MFC r260151 (by adrian):

  Use an RLOCK here instead of an RWLOCK - matching all the other calls
  to lla_lookup().

  This drastically reduces the very high lock contention when doing parallel
  TCP throughput tests (> 1024 sockets) with IPv6.

MFC r260187:
  lla_lookup() does modification only when LLE_CREATE is specified.
  Thus we can use IF_AFDATA_RLOCK() instead of IF_AFDATA_LOCK() when doing
  lla_lookup() without LLE_CREATE flag.

MFC r260217:
  Add IF_AFDATA_WLOCK_ASSERT() in case lla_lookup() is called with
  LLE_CREATE flag.

3 files changed, 10 insertions, 9 deletions

diff --git a/sys/netinet6/in6.c b/sys/netinet6/in6.c
index e5c62df..37a41bd 100644
--- a/sys/netinet6/in6.c
+++ b/sys/netinet6/in6.c
@@ -2627,6 +2627,7 @@ in6_lltable_lookup(struct lltable *llt, u_int flags,
 	if (lle == NULL) {
 		if (!(flags & LLE_CREATE))
 			return (NULL);
+		IF_AFDATA_WLOCK_ASSERT(ifp);
 		/*
 		 * A route that covers the given address must have
 		 * been installed 1st because we are doing a resolution,
diff --git a/sys/netinet6/nd6.c b/sys/netinet6/nd6.c
index e76121d..3620623 100644
--- a/sys/netinet6/nd6.c
+++ b/sys/netinet6/nd6.c
@@ -1146,9 +1146,9 @@ nd6_nud_hint(struct rtentry *rt, struct in6_addr *dst6, int force)
 		return;
 
 	ifp = rt->rt_ifp;
-	IF_AFDATA_LOCK(ifp);
+	IF_AFDATA_RLOCK(ifp);
 	ln = nd6_lookup(dst6, ND6_EXCLUSIVE, NULL);
-	IF_AFDATA_UNLOCK(ifp);
+	IF_AFDATA_RUNLOCK(ifp);
 	if (ln == NULL)
 		return;
 
@@ -1574,16 +1574,16 @@ nd6_cache_lladdr(struct ifnet *ifp, struct in6_addr *from, char *lladdr,
 	 * description on it in NS section (RFC 2461 7.2.3).
 	 */
 	flags = lladdr ? ND6_EXCLUSIVE : 0;
-	IF_AFDATA_LOCK(ifp);
+	IF_AFDATA_RLOCK(ifp);
 	ln = nd6_lookup(from, flags, ifp);
-
+	IF_AFDATA_RUNLOCK(ifp);
 	if (ln == NULL) {
 		flags |= ND6_EXCLUSIVE;
+		IF_AFDATA_LOCK(ifp);
 		ln = nd6_lookup(from, flags | ND6_CREATE, ifp);
 		IF_AFDATA_UNLOCK(ifp);
 		is_newentry = 1;
 	} else {
-		IF_AFDATA_UNLOCK(ifp);		
 		/* do nothing if static ndp is set */
 		if (ln->la_flags & LLE_STATIC) {
 			static_route = 1;
@@ -1891,9 +1891,9 @@ nd6_output_lle(struct ifnet *ifp, struct ifnet *origifp, struct mbuf *m0,
 	flags = ((m != NULL) || (lle != NULL)) ? LLE_EXCLUSIVE : 0;
 	if (ln == NULL) {
 	retry:
-		IF_AFDATA_LOCK(ifp);
+		IF_AFDATA_RLOCK(ifp);
 		ln = lla_lookup(LLTABLE6(ifp), flags, (struct sockaddr *)dst);
-		IF_AFDATA_UNLOCK(ifp);
+		IF_AFDATA_RUNLOCK(ifp);
 		if ((ln == NULL) && nd6_is_addr_neighbor(dst, ifp))  {
 			/*
 			 * Since nd6_is_addr_neighbor() internally calls nd6_lookup(),
diff --git a/sys/netinet6/nd6_nbr.c b/sys/netinet6/nd6_nbr.c
index 532c9b8..556f623 100644
--- a/sys/netinet6/nd6_nbr.c
+++ b/sys/netinet6/nd6_nbr.c
@@ -736,9 +736,9 @@ nd6_na_input(struct mbuf *m, int off, int icmp6len)
 	 * If no neighbor cache entry is found, NA SHOULD silently be
 	 * discarded.
 	 */
-	IF_AFDATA_LOCK(ifp);
+	IF_AFDATA_RLOCK(ifp);
 	ln = nd6_lookup(&taddr6, LLE_EXCLUSIVE, ifp);
-	IF_AFDATA_UNLOCK(ifp);
+	IF_AFDATA_RUNLOCK(ifp);
 	if (ln == NULL) {
 		goto freeit;
 	}