[IPSEC]: Add ICMP host relookup support

RFC 4301 requires us to relookup ICMP traffic that does not match any
policies using the reverse of its payload.  This patch implements this
for ICMP traffic that originates from or terminates on localhost.

This is activated on outbound with the new policy flag XFRM_POLICY_ICMP,
and on inbound by the new state flag XFRM_STATE_ICMP.

On inbound the policy check is now performed by the ICMP protocol so
that it can repeat the policy check where necessary.

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c
index 7f8b27f..5089a36 100644
--- a/net/ipv4/af_inet.c
+++ b/net/ipv4/af_inet.c
@@ -1291,6 +1291,7 @@
 
 static struct net_protocol icmp_protocol = {
 	.handler =	icmp_rcv,
+	.no_policy =	1,
 };
 
 static int __init init_ipv4_mibs(void)
diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
index 13d7459..3c41a6f 100644
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c
@@ -92,6 +92,7 @@
 #include <asm/system.h>
 #include <asm/uaccess.h>
 #include <net/checksum.h>
+#include <net/xfrm.h>
 
 /*
  *	Build xmit assembly blocks
@@ -563,11 +564,71 @@
 				}
 			}
 		};
+		int err;
+		struct rtable *rt2;
+
 		security_skb_classify_flow(skb_in, &fl);
-		if (ip_route_output_key(&rt, &fl))
+		if (__ip_route_output_key(&rt, &fl))
+			goto out_unlock;
+
+		/* No need to clone since we're just using its address. */
+		rt2 = rt;
+
+		err = xfrm_lookup((struct dst_entry **)&rt, &fl, NULL, 0);
+		switch (err) {
+		case 0:
+			if (rt != rt2)
+				goto route_done;
+			break;
+		case -EPERM:
+			rt = NULL;
+			break;
+		default:
+			goto out_unlock;
+		}
+
+		if (xfrm_decode_session_reverse(skb_in, &fl, AF_INET))
+			goto out_unlock;
+
+		if (inet_addr_type(fl.fl4_src) == RTN_LOCAL)
+			err = __ip_route_output_key(&rt2, &fl);
+		else {
+			struct flowi fl2 = {};
+			struct dst_entry *odst;
+
+			fl2.fl4_dst = fl.fl4_src;
+			if (ip_route_output_key(&rt2, &fl2))
+				goto out_unlock;
+
+			/* Ugh! */
+			odst = skb_in->dst;
+			err = ip_route_input(skb_in, fl.fl4_dst, fl.fl4_src,
+					     RT_TOS(tos), rt2->u.dst.dev);
+
+			dst_release(&rt2->u.dst);
+			rt2 = (struct rtable *)skb_in->dst;
+			skb_in->dst = odst;
+		}
+
+		if (err)
+			goto out_unlock;
+
+		err = xfrm_lookup((struct dst_entry **)&rt2, &fl, NULL,
+				  XFRM_LOOKUP_ICMP);
+		if (err == -ENOENT) {
+			if (!rt)
+				goto out_unlock;
+			goto route_done;
+		}
+
+		dst_release(&rt->u.dst);
+		rt = rt2;
+
+		if (err)
 			goto out_unlock;
 	}
 
+route_done:
 	if (!icmpv4_xrlim_allow(rt, type, code))
 		goto ende;
 
@@ -916,6 +977,22 @@
 	struct icmphdr *icmph;
 	struct rtable *rt = (struct rtable *)skb->dst;
 
+	if (!xfrm4_policy_check(NULL, XFRM_POLICY_IN, skb) &&
+	    skb->sp->xvec[skb->sp->len - 1]->props.flags & XFRM_STATE_ICMP) {
+		int nh;
+
+		if (!pskb_may_pull(skb, sizeof(*icmph) + sizeof(struct iphdr)))
+			goto drop;
+
+		nh = skb_network_offset(skb);
+		skb_set_network_header(skb, sizeof(*icmph));
+
+		if (!xfrm4_policy_check_reverse(NULL, XFRM_POLICY_IN, skb))
+			goto drop;
+
+		skb_set_network_header(skb, nh);
+	}
+
 	ICMP_INC_STATS_BH(ICMP_MIB_INMSGS);
 
 	switch (skb->ip_summed) {
@@ -929,8 +1006,7 @@
 			goto error;
 	}
 
-	if (!pskb_pull(skb, sizeof(struct icmphdr)))
-		goto error;
+	__skb_pull(skb, sizeof(*icmph));
 
 	icmph = icmp_hdr(skb);
 
diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c
index 93c96cf..c0bea7b 100644
--- a/net/ipv6/icmp.c
+++ b/net/ipv6/icmp.c
@@ -63,6 +63,7 @@
 #include <net/ip6_route.h>
 #include <net/addrconf.h>
 #include <net/icmp.h>
+#include <net/xfrm.h>
 
 #include <asm/uaccess.h>
 #include <asm/system.h>
@@ -86,7 +87,7 @@
 
 static struct inet6_protocol icmpv6_protocol = {
 	.handler	=	icmpv6_rcv,
-	.flags		=	INET6_PROTO_FINAL,
+	.flags		=	INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
 };
 
 static __inline__ int icmpv6_xmit_lock(void)
@@ -310,8 +311,10 @@
 	struct ipv6_pinfo *np;
 	struct in6_addr *saddr = NULL;
 	struct dst_entry *dst;
+	struct dst_entry *dst2;
 	struct icmp6hdr tmp_hdr;
 	struct flowi fl;
+	struct flowi fl2;
 	struct icmpv6_msg msg;
 	int iif = 0;
 	int addr_type = 0;
@@ -418,9 +421,42 @@
 		goto out_dst_release;
 	}
 
-	if ((err = xfrm_lookup(&dst, &fl, sk, 0)) < 0)
+	/* No need to clone since we're just using its address. */
+	dst2 = dst;
+
+	err = xfrm_lookup(&dst, &fl, sk, 0);
+	switch (err) {
+	case 0:
+		if (dst != dst2)
+			goto route_done;
+		break;
+	case -EPERM:
+		dst = NULL;
+		break;
+	default:
+		goto out;
+	}
+
+	if (xfrm_decode_session_reverse(skb, &fl2, AF_INET6))
 		goto out;
 
+	if (ip6_dst_lookup(sk, &dst2, &fl))
+		goto out;
+
+	err = xfrm_lookup(&dst2, &fl, sk, XFRM_LOOKUP_ICMP);
+	if (err == -ENOENT) {
+		if (!dst)
+			goto out;
+		goto route_done;
+	}
+
+	dst_release(dst);
+	dst = dst2;
+
+	if (err)
+		goto out;
+
+route_done:
 	if (ipv6_addr_is_multicast(&fl.fl6_dst))
 		hlimit = np->mcast_hops;
 	else
@@ -608,6 +644,22 @@
 	struct icmp6hdr *hdr;
 	int type;
 
+	if (xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb) &&
+	    skb->sp->xvec[skb->sp->len - 1]->props.flags & XFRM_STATE_ICMP) {
+		int nh;
+
+		if (!pskb_may_pull(skb, sizeof(*hdr) + sizeof(*orig_hdr)))
+			goto drop_no_count;
+
+		nh = skb_network_offset(skb);
+		skb_set_network_header(skb, sizeof(*hdr));
+
+		if (!xfrm6_policy_check_reverse(NULL, XFRM_POLICY_IN, skb))
+			goto drop_no_count;
+
+		skb_set_network_header(skb, nh);
+	}
+
 	ICMP6_INC_STATS_BH(idev, ICMP6_MIB_INMSGS);
 
 	saddr = &ipv6_hdr(skb)->saddr;
@@ -630,8 +682,7 @@
 		}
 	}
 
-	if (!pskb_pull(skb, sizeof(struct icmp6hdr)))
-		goto discard_it;
+	__skb_pull(skb, sizeof(*hdr));
 
 	hdr = icmp6_hdr(skb);
 
@@ -717,6 +768,7 @@
 
 discard_it:
 	ICMP6_INC_STATS_BH(idev, ICMP6_MIB_INERRORS);
+drop_no_count:
 	kfree_skb(skb);
 	return 0;
 }
diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c
index 2e10d46..a83b5e1 100644
--- a/net/xfrm/xfrm_policy.c
+++ b/net/xfrm/xfrm_policy.c
@@ -1469,11 +1469,13 @@
 			goto dropdst;
 	}
 
+	err = -ENOENT;
+
 	if (!policy) {
 		/* To accelerate a bit...  */
 		if ((dst_orig->flags & DST_NOXFRM) ||
 		    !xfrm_policy_count[XFRM_POLICY_OUT])
-			return 0;
+			goto nopol;
 
 		policy = flow_cache_lookup(fl, dst_orig->ops->family,
 					   dir, xfrm_policy_lookup);
@@ -1483,14 +1485,18 @@
 	}
 
 	if (!policy)
-		return 0;
+		goto nopol;
 
 	family = dst_orig->ops->family;
-	policy->curlft.use_time = get_seconds();
 	pols[0] = policy;
 	npols ++;
 	xfrm_nr += pols[0]->xfrm_nr;
 
+	if ((flags & XFRM_LOOKUP_ICMP) && !(policy->flags & XFRM_POLICY_ICMP))
+		goto error;
+
+	policy->curlft.use_time = get_seconds();
+
 	switch (policy->action) {
 	default:
 	case XFRM_POLICY_BLOCK:
@@ -1649,6 +1655,11 @@
 	dst_release(dst_orig);
 	*dst_p = NULL;
 	return err;
+
+nopol:
+	if (flags & XFRM_LOOKUP_ICMP)
+		goto dropdst;
+	return 0;
 }
 EXPORT_SYMBOL(__xfrm_lookup);