Merge branch 'fib_trie-next'

Alexander Duyck says:

====================
fib_trie: Reduce time spent in fib_table_lookup by 35 to 75%

These patches are meant to address several performance issues I have seen
in the fib_trie implementation, and fib_table_lookup specifically.  With
these changes in place I have seen a reduction of up to 35 to 75% for the
total time spent in fib_table_lookup depending on the type of search being
performed.

On a VM running in my Corei7-4930K system with a trie of maximum depth of 7
this resulted in a reduction of over 370ns per packet in the total time to
process packets received from an ixgbe interface and route them to a dummy
interface.  This represents a failed lookup in the local trie followed by
a successful search in the main trie.

				Baseline	Refactor
  ixgbe->dummy routing		1.20Mpps	2.21Mpps
  ------------------------------------------------------------
  processing time per packet		835ns		453ns
  fib_table_lookup		50.1%	418ns	25.0%	113ns
  check_leaf.isra.9		 7.9%	 66ns	   --	 --
  ixgbe_clean_rx_irq		 5.3%	 44ns	 9.8%	 44ns
  ip_route_input_noref		 2.9%	 25ns	 4.6%	 21ns
  pvclock_clocksource_read	 2.6%	 21ns	 4.6%	 21ns
  ip_rcv			 2.6%	 22ns	 4.0%	 18ns

In the simple case of receiving a frame and dropping it before it can reach
the socket layer I saw a reduction of 40ns per packet.  This represents a
trip through the local trie with the correct leaf found with no need for
any backtracing.

				Baseline	Refactor
  ixgbe->local receive		2.65Mpps	2.96Mpps
  ------------------------------------------------------------
  processing time per packet		377ns		337ns
  fib_table_lookup		25.1%	 95ns	25.8%	 87ns
  ixgbe_clean_rx_irq		 8.7%	 33ns	 9.0%	 30ns
  check_leaf.isra.9		 7.2%	 27ns	   --	 --
  ip_rcv			 5.7%	 21ns	 6.5%	 22ns

These changes have resulted in several functions being inlined such as
check_leaf and fib_find_node, but due to the code simplification the
overall size of the code has been reduced.

   text	   data	    bss	    dec	    hex	filename
  16932	    376	     16	  17324	   43ac	net/ipv4/fib_trie.o - before
  15259	    376	      8	  15643	   3d1b	net/ipv4/fib_trie.o - after

Changes since RFC:
  Replaced this_cpu_ptr with correct call to this_cpu_inc in patch 1
  Changed test for leaf_info mismatch to (key ^ n->key) & li->mask_plen in patch 10
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

include/net/ip_fib.h

@@ -222,16 +222,19 @@ static inline struct fib_table *fib_new_table(struct net *net, u32 id)
 static inline int fib_lookup(struct net *net, const struct flowi4 *flp,
 			     struct fib_result *res)
 {
-	struct fib_table *table;
+	int err = -ENETUNREACH;
+
+	rcu_read_lock();
+
+	if (!fib_table_lookup(fib_get_table(net, RT_TABLE_LOCAL), flp, res,
+			      FIB_LOOKUP_NOREF) ||
+	    !fib_table_lookup(fib_get_table(net, RT_TABLE_MAIN), flp, res,
+			      FIB_LOOKUP_NOREF))
+		err = 0;
 
-	table = fib_get_table(net, RT_TABLE_LOCAL);
-	if (!fib_table_lookup(table, flp, res, FIB_LOOKUP_NOREF))
-		return 0;
+	rcu_read_unlock();
 
-	table = fib_get_table(net, RT_TABLE_MAIN);
-	if (!fib_table_lookup(table, flp, res, FIB_LOOKUP_NOREF))
-		return 0;
-	return -ENETUNREACH;
+	return err;
 }
 
 #else /* CONFIG_IP_MULTIPLE_TABLES */
@@ -247,20 +250,25 @@ static inline int fib_lookup(struct net *net, struct flowi4 *flp,
 			     struct fib_result *res)
 {
 	if (!net->ipv4.fib_has_custom_rules) {
+		int err = -ENETUNREACH;
+
+		rcu_read_lock();
+
 		res->tclassid = 0;
-		if (net->ipv4.fib_local &&
-		    !fib_table_lookup(net->ipv4.fib_local, flp, res,
-				      FIB_LOOKUP_NOREF))
-			return 0;
-		if (net->ipv4.fib_main &&
-		    !fib_table_lookup(net->ipv4.fib_main, flp, res,
-				      FIB_LOOKUP_NOREF))
-			return 0;
-		if (net->ipv4.fib_default &&
-		    !fib_table_lookup(net->ipv4.fib_default, flp, res,
-				      FIB_LOOKUP_NOREF))
-			return 0;
-		return -ENETUNREACH;
+		if ((net->ipv4.fib_local &&
+		     !fib_table_lookup(net->ipv4.fib_local, flp, res,
+				       FIB_LOOKUP_NOREF)) ||
+		    (net->ipv4.fib_main &&
+		     !fib_table_lookup(net->ipv4.fib_main, flp, res,
+				       FIB_LOOKUP_NOREF)) ||
+		    (net->ipv4.fib_default &&
+		     !fib_table_lookup(net->ipv4.fib_default, flp, res,
+				       FIB_LOOKUP_NOREF)))
+			err = 0;
+
+		rcu_read_unlock();
+
+		return err;
 	}
 	return __fib_lookup(net, flp, res);
 }

net/ipv4/fib_frontend.c

@@ -67,7 +67,7 @@ static int __net_init fib4_rules_init(struct net *net)
 	return 0;
 
 fail:
-	kfree(local_table);
+	fib_free_table(local_table);
 	return -ENOMEM;
 }
 #else
@@ -109,6 +109,7 @@ struct fib_table *fib_new_table(struct net *net, u32 id)
 	return tb;
 }
 
+/* caller must hold either rtnl or rcu read lock */
 struct fib_table *fib_get_table(struct net *net, u32 id)
 {
 	struct fib_table *tb;
@@ -119,15 +120,11 @@ struct fib_table *fib_get_table(struct net *net, u32 id)
 		id = RT_TABLE_MAIN;
 	h = id & (FIB_TABLE_HASHSZ - 1);
 
-	rcu_read_lock();
 	head = &net->ipv4.fib_table_hash[h];
 	hlist_for_each_entry_rcu(tb, head, tb_hlist) {
-		if (tb->tb_id == id) {
-			rcu_read_unlock();
+		if (tb->tb_id == id)
 			return tb;
-		}
 	}
-	rcu_read_unlock();
 	return NULL;
 }
 #endif /* CONFIG_IP_MULTIPLE_TABLES */
@@ -167,16 +164,18 @@ static inline unsigned int __inet_dev_addr_type(struct net *net,
 	if (ipv4_is_multicast(addr))
 		return RTN_MULTICAST;
 
+	rcu_read_lock();
+
 	local_table = fib_get_table(net, RT_TABLE_LOCAL);
 	if (local_table) {
 		ret = RTN_UNICAST;
-		rcu_read_lock();
 		if (!fib_table_lookup(local_table, &fl4, &res, FIB_LOOKUP_NOREF)) {
 			if (!dev || dev == res.fi->fib_dev)
 				ret = res.type;
 		}
-		rcu_read_unlock();
 	}
+
+	rcu_read_unlock();
 	return ret;
 }
 
@@ -919,7 +918,7 @@ void fib_del_ifaddr(struct in_ifaddr *ifa, struct in_ifaddr *iprim)
 #undef BRD1_OK
 }
 
-static void nl_fib_lookup(struct fib_result_nl *frn, struct fib_table *tb)
+static void nl_fib_lookup(struct net *net, struct fib_result_nl *frn)
 {
 
 	struct fib_result       res;
@@ -929,6 +928,11 @@ static void nl_fib_lookup(struct fib_result_nl *frn, struct fib_table *tb)
 		.flowi4_tos = frn->fl_tos,
 		.flowi4_scope = frn->fl_scope,
 	};
+	struct fib_table *tb;
+
+	rcu_read_lock();
+
+	tb = fib_get_table(net, frn->tb_id_in);
 
 	frn->err = -ENOENT;
 	if (tb) {
@@ -945,6 +949,8 @@ static void nl_fib_lookup(struct fib_result_nl *frn, struct fib_table *tb)
 		}
 		local_bh_enable();
 	}
+
+	rcu_read_unlock();
 }
 
 static void nl_fib_input(struct sk_buff *skb)
@@ -952,7 +958,6 @@ static void nl_fib_input(struct sk_buff *skb)
 	struct net *net;
 	struct fib_result_nl *frn;
 	struct nlmsghdr *nlh;
-	struct fib_table *tb;
 	u32 portid;
 
 	net = sock_net(skb->sk);
@@ -967,9 +972,7 @@ static void nl_fib_input(struct sk_buff *skb)
 	nlh = nlmsg_hdr(skb);
 
 	frn = (struct fib_result_nl *) nlmsg_data(nlh);
-	tb = fib_get_table(net, frn->tb_id_in);
-
-	nl_fib_lookup(frn, tb);
+	nl_fib_lookup(net, frn);
 
 	portid = NETLINK_CB(skb).portid;      /* netlink portid */
 	NETLINK_CB(skb).portid = 0;        /* from kernel */

net/ipv4/fib_rules.c

@@ -81,27 +81,25 @@ static int fib4_rule_action(struct fib_rule *rule, struct flowi *flp,
 		break;
 
 	case FR_ACT_UNREACHABLE:
-		err = -ENETUNREACH;
-		goto errout;
+		return -ENETUNREACH;
 
 	case FR_ACT_PROHIBIT:
-		err = -EACCES;
-		goto errout;
+		return -EACCES;
 
 	case FR_ACT_BLACKHOLE:
 	default:
-		err = -EINVAL;
-		goto errout;
+		return -EINVAL;
 	}
 
+	rcu_read_lock();
+
 	tbl = fib_get_table(rule->fr_net, rule->table);
-	if (!tbl)
-		goto errout;
+	if (tbl)
+		err = fib_table_lookup(tbl, &flp->u.ip4,
+				       (struct fib_result *)arg->result,
+				       arg->flags);
 
-	err = fib_table_lookup(tbl, &flp->u.ip4, (struct fib_result *) arg->result, arg->flags);
-	if (err > 0)
-		err = -EAGAIN;
-errout:
+	rcu_read_unlock();
 	return err;
 }