ipv6: route: dissect flow in input path if fib rules need it

Dissect flow in fwd path if fib rules require it. Controlled by
a flag to avoid penatly for the common case. Flag is set when fib
rules with sport, dport and proto match that require flow dissect
are installed. Also passes the dissected hash keys to the multipath
hash function when applicable to avoid dissecting the flow again.
icmp packets will continue to use inner header for hash
calculations (Thanks to Nikolay Aleksandrov for some review here).

Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Acked-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

3 files changed, 38 insertions, 15 deletions

diff --git a/net/ipv4/fib_rules.c b/net/ipv4/fib_rules.c
index 16083b82954b..737d11bc8838 100644
--- a/net/ipv4/fib_rules.c
+++ b/net/ipv4/fib_rules.c
@@ -255,6 +255,9 @@ static int fib4_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
 	}
 #endif
 
+	if (fib_rule_requires_fldissect(rule))
+		net->ipv4.fib_rules_require_fldissect++;
+
 	rule4->src_len = frh->src_len;
 	rule4->srcmask = inet_make_mask(rule4->src_len);
 	rule4->dst_len = frh->dst_len;
@@ -283,6 +286,10 @@ static int fib4_rule_delete(struct fib_rule *rule)
 		net->ipv4.fib_num_tclassid_users--;
 #endif
 	net->ipv4.fib_has_custom_rules = true;
+
+	if (net->ipv4.fib_rules_require_fldissect &&
+	    fib_rule_requires_fldissect(rule))
+		net->ipv4.fib_rules_require_fldissect--;
 errout:
 	return err;
 }
@@ -400,6 +407,7 @@ int __net_init fib4_rules_init(struct net *net)
 		goto fail;
 	net->ipv4.rules_ops = ops;
 	net->ipv4.fib_has_custom_rules = false;
+	net->ipv4.fib_rules_require_fldissect = 0;
 	return 0;
 
 fail:
diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c
index f31e6575ab91..181b0d8d589c 100644
--- a/net/ipv4/fib_semantics.c
+++ b/net/ipv4/fib_semantics.c
@@ -1770,7 +1770,7 @@ void fib_select_path(struct net *net, struct fib_result *res,
 
 #ifdef CONFIG_IP_ROUTE_MULTIPATH
 	if (res->fi->fib_nhs > 1) {
-		int h = fib_multipath_hash(res->fi, fl4, skb);
+		int h = fib_multipath_hash(res->fi, fl4, skb, NULL);
 
 		fib_select_multipath(res, h);
 	}
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 26eefa2eaa44..3bb686dac273 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -1783,7 +1783,7 @@ static void ip_multipath_l3_keys(const struct sk_buff *skb,
 
 /* if skb is set it will be used and fl4 can be NULL */
 int fib_multipath_hash(const struct fib_info *fi, const struct flowi4 *fl4,
-		       const struct sk_buff *skb)
+		       const struct sk_buff *skb, struct flow_keys *flkeys)
 {
 	struct net *net = fi->fib_net;
 	struct flow_keys hash_keys;
@@ -1810,14 +1810,23 @@ int fib_multipath_hash(const struct fib_info *fi, const struct flowi4 *fl4,
 			if (skb->l4_hash)
 				return skb_get_hash_raw(skb) >> 1;
 			memset(&hash_keys, 0, sizeof(hash_keys));
-			skb_flow_dissect_flow_keys(skb, &keys, flag);
 
-			hash_keys.control.addr_type = FLOW_DISSECTOR_KEY_IPV4_ADDRS;
-			hash_keys.addrs.v4addrs.src = keys.addrs.v4addrs.src;
-			hash_keys.addrs.v4addrs.dst = keys.addrs.v4addrs.dst;
-			hash_keys.ports.src = keys.ports.src;
-			hash_keys.ports.dst = keys.ports.dst;
-			hash_keys.basic.ip_proto = keys.basic.ip_proto;
+			if (flkeys) {
+				hash_keys.control.addr_type = FLOW_DISSECTOR_KEY_IPV4_ADDRS;
+				hash_keys.addrs.v4addrs.src = flkeys->addrs.v4addrs.src;
+				hash_keys.addrs.v4addrs.dst = flkeys->addrs.v4addrs.dst;
+				hash_keys.ports.src = flkeys->ports.src;
+				hash_keys.ports.dst = flkeys->ports.dst;
+				hash_keys.basic.ip_proto = flkeys->basic.ip_proto;
+			} else {
+				skb_flow_dissect_flow_keys(skb, &keys, flag);
+				hash_keys.control.addr_type = FLOW_DISSECTOR_KEY_IPV4_ADDRS;
+				hash_keys.addrs.v4addrs.src = keys.addrs.v4addrs.src;
+				hash_keys.addrs.v4addrs.dst = keys.addrs.v4addrs.dst;
+				hash_keys.ports.src = keys.ports.src;
+				hash_keys.ports.dst = keys.ports.dst;
+				hash_keys.basic.ip_proto = keys.basic.ip_proto;
+			}
 		} else {
 			memset(&hash_keys, 0, sizeof(hash_keys));
 			hash_keys.control.addr_type = FLOW_DISSECTOR_KEY_IPV4_ADDRS;
@@ -1838,11 +1847,12 @@ int fib_multipath_hash(const struct fib_info *fi, const struct flowi4 *fl4,
 static int ip_mkroute_input(struct sk_buff *skb,
 			    struct fib_result *res,
 			    struct in_device *in_dev,
-			    __be32 daddr, __be32 saddr, u32 tos)
+			    __be32 daddr, __be32 saddr, u32 tos,
+			    struct flow_keys *hkeys)
 {
 #ifdef CONFIG_IP_ROUTE_MULTIPATH
 	if (res->fi && res->fi->fib_nhs > 1) {
-		int h = fib_multipath_hash(res->fi, NULL, skb);
+		int h = fib_multipath_hash(res->fi, NULL, skb, hkeys);
 
 		fib_select_multipath(res, h);
 	}
@@ -1868,13 +1878,14 @@ static int ip_route_input_slow(struct sk_buff *skb, __be32 daddr, __be32 saddr,
 			       struct fib_result *res)
 {
 	struct in_device *in_dev = __in_dev_get_rcu(dev);
+	struct flow_keys *flkeys = NULL, _flkeys;
+	struct net    *net = dev_net(dev);
 	struct ip_tunnel_info *tun_info;
-	struct flowi4	fl4;
+	int		err = -EINVAL;
 	unsigned int	flags = 0;
 	u32		itag = 0;
 	struct rtable	*rth;
-	int		err = -EINVAL;
-	struct net    *net = dev_net(dev);
+	struct flowi4	fl4;
 	bool do_cache;
 
 	/* IP on this device is disabled. */
@@ -1933,6 +1944,10 @@ static int ip_route_input_slow(struct sk_buff *skb, __be32 daddr, __be32 saddr,
 	fl4.daddr = daddr;
 	fl4.saddr = saddr;
 	fl4.flowi4_uid = sock_net_uid(net, NULL);
+
+	if (fib4_rules_early_flow_dissect(net, skb, &fl4, &_flkeys))
+		flkeys = &_flkeys;
+
 	err = fib_lookup(net, &fl4, res, 0);
 	if (err != 0) {
 		if (!IN_DEV_FORWARD(in_dev))
@@ -1958,7 +1973,7 @@ static int ip_route_input_slow(struct sk_buff *skb, __be32 daddr, __be32 saddr,
 	if (res->type != RTN_UNICAST)
 		goto martian_destination;
 
-	err = ip_mkroute_input(skb, res, in_dev, daddr, saddr, tos);
+	err = ip_mkroute_input(skb, res, in_dev, daddr, saddr, tos, flkeys);
 out:	return err;
 
 brd_input: