[PATCH 0/3] netfilter: propagate routing and XFRM errors

[PATCH 0/3] netfilter: propagate routing and XFRM errors

[Patrick McHardy]

These patches fix a long standing annoyance: netfilter will return
EPERM for all errors during rerouting or XFRM policy lookups.

These patches change {ip,ip6}_route_me_harder() and nf_xfrm_me_harder()
to return errno codes and use NF_DROP_ERR() to propagate those back
to userspace.

Please apply.

[PATCH 1/3] netfilter: ipv4: propagate routing errors from ip_route_me_harder()

[Patrick McHardy]

Propagate routing errors from ip_route_me_harder() when dropping a packet
using NF_DROP_ERR(). This makes userspace get the proper error instead of
EPERM for everything.

Example:

# ip r a unreachable default table 100
# ip ru add fwmark 0x1 lookup 100
# iptables -t mangle -A OUTPUT -d 8.8.8.8 -j MARK --set-mark 0x1

Current behaviour:

PING 8.8.8.8 (8.8.8.8) 56(84) bytes of data.
ping: sendmsg: Operation not permitted
ping: sendmsg: Operation not permitted
ping: sendmsg: Operation not permitted

New behaviour:

PING 8.8.8.8 (8.8.8.8) 56(84) bytes of data.
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable

Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 net/ipv4/netfilter.c                | 8 ++++----
 net/ipv4/netfilter/iptable_mangle.c | 9 ++++++---
 net/ipv4/netfilter/iptable_nat.c    | 6 ++++--
 3 files changed, 14 insertions(+), 9 deletions(-)

diff --git a/net/ipv4/netfilter.c b/net/ipv4/netfilter.c
index 4c0cf63..8b201e8 100644
--- a/net/ipv4/netfilter.c
+++ b/net/ipv4/netfilter.c
@@ -40,14 +40,14 @@ int ip_route_me_harder(struct sk_buff *skb, unsigned int addr_type)
 	fl4.flowi4_flags = flags;
 	rt = ip_route_output_key(net, &fl4);
 	if (IS_ERR(rt))
-		return -1;
+		return PTR_ERR(rt);
 
 	/* Drop old route. */
 	skb_dst_drop(skb);
 	skb_dst_set(skb, &rt->dst);
 
 	if (skb_dst(skb)->error)
-		return -1;
+		return skb_dst(skb)->error;
 
 #ifdef CONFIG_XFRM
 	if (!(IPCB(skb)->flags & IPSKB_XFRM_TRANSFORMED) &&
@@ -56,7 +56,7 @@ int ip_route_me_harder(struct sk_buff *skb, unsigned int addr_type)
 		skb_dst_set(skb, NULL);
 		dst = xfrm_lookup(net, dst, flowi4_to_flowi(&fl4), skb->sk, 0);
 		if (IS_ERR(dst))
-			return -1;
+			return PTR_ERR(dst);;
 		skb_dst_set(skb, dst);
 	}
 #endif
@@ -66,7 +66,7 @@ int ip_route_me_harder(struct sk_buff *skb, unsigned int addr_type)
 	if (skb_headroom(skb) < hh_len &&
 	    pskb_expand_head(skb, HH_DATA_ALIGN(hh_len - skb_headroom(skb)),
 				0, GFP_ATOMIC))
-		return -1;
+		return -ENOMEM;
 
 	return 0;
 }
diff --git a/net/ipv4/netfilter/iptable_mangle.c b/net/ipv4/netfilter/iptable_mangle.c
index 85d88f2..cba5658 100644
--- a/net/ipv4/netfilter/iptable_mangle.c
+++ b/net/ipv4/netfilter/iptable_mangle.c
@@ -44,6 +44,7 @@ ipt_mangle_out(struct sk_buff *skb, const struct net_device *out)
 	u_int8_t tos;
 	__be32 saddr, daddr;
 	u_int32_t mark;
+	int err;
 
 	/* root is playing with raw sockets. */
 	if (skb->len < sizeof(struct iphdr) ||
@@ -66,9 +67,11 @@ ipt_mangle_out(struct sk_buff *skb, const struct net_device *out)
 		if (iph->saddr != saddr ||
 		    iph->daddr != daddr ||
 		    skb->mark != mark ||
-		    iph->tos != tos)
-			if (ip_route_me_harder(skb, RTN_UNSPEC))
-				ret = NF_DROP;
+		    iph->tos != tos) {
+			err = ip_route_me_harder(skb, RTN_UNSPEC);
+			if (err < 0)
+				ret = NF_DROP_ERR(err);
+		}
 	}
 
 	return ret;
diff --git a/net/ipv4/netfilter/iptable_nat.c b/net/ipv4/netfilter/iptable_nat.c
index eeaff7e..c2937c8 100644
--- a/net/ipv4/netfilter/iptable_nat.c
+++ b/net/ipv4/netfilter/iptable_nat.c
@@ -213,6 +213,7 @@ nf_nat_ipv4_local_fn(unsigned int hooknum,
 	const struct nf_conn *ct;
 	enum ip_conntrack_info ctinfo;
 	unsigned int ret;
+	int err;
 
 	/* root is playing with raw sockets. */
 	if (skb->len < sizeof(struct iphdr) ||
@@ -226,8 +227,9 @@ nf_nat_ipv4_local_fn(unsigned int hooknum,
 
 		if (ct->tuplehash[dir].tuple.dst.u3.ip !=
 		    ct->tuplehash[!dir].tuple.src.u3.ip) {
-			if (ip_route_me_harder(skb, RTN_UNSPEC))
-				ret = NF_DROP;
+			err = ip_route_me_harder(skb, RTN_UNSPEC);
+			if (err < 0)
+				ret = NF_DROP_ERR(err);
 		}
 #ifdef CONFIG_XFRM
 		else if (!(IPCB(skb)->flags & IPSKB_XFRM_TRANSFORMED) &&
-- 
1.8.1.4

[PATCH 2/3] netfilter: ipv6: propagate routing errors from ip6_route_me_harder()

[Patrick McHardy]

Propagate routing errors from ip_route_me_harder() when dropping a packet
using NF_DROP_ERR(). This makes userspace get the proper error instead of
EPERM for everything.

# ip -6 r a unreachable default table 100
# ip -6 ru add fwmark 0x1 lookup 100
# ip6tables -t mangle -A OUTPUT -d 2001:4860:4860::8888 -j MARK --set-mark 0x1

Old behaviour:

PING 2001:4860:4860::8888(2001:4860:4860::8888) 56 data bytes
ping: sendmsg: Operation not permitted
ping: sendmsg: Operation not permitted
ping: sendmsg: Operation not permitted

New behaviour:

PING 2001:4860:4860::8888(2001:4860:4860::8888) 56 data bytes
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable

Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 net/ipv6/netfilter.c                 | 6 +++---
 net/ipv6/netfilter/ip6table_mangle.c | 9 ++++++---
 net/ipv6/netfilter/ip6table_nat.c    | 6 ++++--
 3 files changed, 13 insertions(+), 8 deletions(-)

diff --git a/net/ipv6/netfilter.c b/net/ipv6/netfilter.c
index 429089c..fc5fbd7 100644
--- a/net/ipv6/netfilter.c
+++ b/net/ipv6/netfilter.c
@@ -29,7 +29,7 @@ int ip6_route_me_harder(struct sk_buff *skb)
 		IP6_INC_STATS(net, ip6_dst_idev(dst), IPSTATS_MIB_OUTNOROUTES);
 		LIMIT_NETDEBUG(KERN_DEBUG "ip6_route_me_harder: No more route.\n");
 		dst_release(dst);
-		return -EINVAL;
+		return dst->error;
 	}
 
 	/* Drop old route. */
@@ -43,7 +43,7 @@ int ip6_route_me_harder(struct sk_buff *skb)
 		skb_dst_set(skb, NULL);
 		dst = xfrm_lookup(net, dst, flowi6_to_flowi(&fl6), skb->sk, 0);
 		if (IS_ERR(dst))
-			return -1;
+			return PTR_ERR(dst);
 		skb_dst_set(skb, dst);
 	}
 #endif
@@ -53,7 +53,7 @@ int ip6_route_me_harder(struct sk_buff *skb)
 	if (skb_headroom(skb) < hh_len &&
 	    pskb_expand_head(skb, HH_DATA_ALIGN(hh_len - skb_headroom(skb)),
 			     0, GFP_ATOMIC))
-		return -1;
+		return -ENOMEM;
 
 	return 0;
 }
diff --git a/net/ipv6/netfilter/ip6table_mangle.c b/net/ipv6/netfilter/ip6table_mangle.c
index 6134a1e..e075399 100644
--- a/net/ipv6/netfilter/ip6table_mangle.c
+++ b/net/ipv6/netfilter/ip6table_mangle.c
@@ -38,7 +38,7 @@ ip6t_mangle_out(struct sk_buff *skb, const struct net_device *out)
 	struct in6_addr saddr, daddr;
 	u_int8_t hop_limit;
 	u_int32_t flowlabel, mark;
-
+	int err;
 #if 0
 	/* root is playing with raw sockets. */
 	if (skb->len < sizeof(struct iphdr) ||
@@ -65,8 +65,11 @@ ip6t_mangle_out(struct sk_buff *skb, const struct net_device *out)
 	     !ipv6_addr_equal(&ipv6_hdr(skb)->daddr, &daddr) ||
 	     skb->mark != mark ||
 	     ipv6_hdr(skb)->hop_limit != hop_limit ||
-	     flowlabel != *((u_int32_t *)ipv6_hdr(skb))))
-		return ip6_route_me_harder(skb) == 0 ? ret : NF_DROP;
+	     flowlabel != *((u_int32_t *)ipv6_hdr(skb)))) {
+		err = ip6_route_me_harder(skb);
+		if (err < 0)
+			ret = NF_DROP_ERR(err);
+	}
 
 	return ret;
 }
diff --git a/net/ipv6/netfilter/ip6table_nat.c b/net/ipv6/netfilter/ip6table_nat.c
index e0e788d..97e2edd 100644
--- a/net/ipv6/netfilter/ip6table_nat.c
+++ b/net/ipv6/netfilter/ip6table_nat.c
@@ -215,6 +215,7 @@ nf_nat_ipv6_local_fn(unsigned int hooknum,
 	const struct nf_conn *ct;
 	enum ip_conntrack_info ctinfo;
 	unsigned int ret;
+	int err;
 
 	/* root is playing with raw sockets. */
 	if (skb->len < sizeof(struct ipv6hdr))
@@ -227,8 +228,9 @@ nf_nat_ipv6_local_fn(unsigned int hooknum,
 
 		if (!nf_inet_addr_cmp(&ct->tuplehash[dir].tuple.dst.u3,
 				      &ct->tuplehash[!dir].tuple.src.u3)) {
-			if (ip6_route_me_harder(skb))
-				ret = NF_DROP;
+			err = ip6_route_me_harder(skb);
+			if (err < 0)
+				ret = NF_DROP_ERR(err);
 		}
 #ifdef CONFIG_XFRM
 		else if (!(IP6CB(skb)->flags & IP6SKB_XFRM_TRANSFORMED) &&
-- 
1.8.1.4

[PATCH 3/3] netfilter: nat: propagate errors from xfrm_me_harder()

[Patrick McHardy]

Propagate errors from ip_xfrm_me_harder() instead of returning EPERM in
all cases.

Signed-off-by: Patrick McHardy <kaber@trash.net>
---
 net/ipv4/netfilter/iptable_nat.c  | 17 +++++++++++------
 net/ipv6/netfilter/ip6table_nat.c | 17 +++++++++++------
 net/netfilter/nf_nat_core.c       |  9 +++++----
 3 files changed, 27 insertions(+), 16 deletions(-)

diff --git a/net/ipv4/netfilter/iptable_nat.c b/net/ipv4/netfilter/iptable_nat.c
index c2937c8..6383273 100644
--- a/net/ipv4/netfilter/iptable_nat.c
+++ b/net/ipv4/netfilter/iptable_nat.c
@@ -176,6 +176,7 @@ nf_nat_ipv4_out(unsigned int hooknum,
 #ifdef CONFIG_XFRM
 	const struct nf_conn *ct;
 	enum ip_conntrack_info ctinfo;
+	int err;
 #endif
 	unsigned int ret;
 
@@ -195,9 +196,11 @@ nf_nat_ipv4_out(unsigned int hooknum,
 		     ct->tuplehash[!dir].tuple.dst.u3.ip) ||
 		    (ct->tuplehash[dir].tuple.dst.protonum != IPPROTO_ICMP &&
 		     ct->tuplehash[dir].tuple.src.u.all !=
-		     ct->tuplehash[!dir].tuple.dst.u.all))
-			if (nf_xfrm_me_harder(skb, AF_INET) < 0)
-				ret = NF_DROP;
+		     ct->tuplehash[!dir].tuple.dst.u.all)) {
+			err = nf_xfrm_me_harder(skb, AF_INET);
+			if (err < 0)
+				ret = NF_DROP_ERR(err);
+		}
 	}
 #endif
 	return ret;
@@ -235,9 +238,11 @@ nf_nat_ipv4_local_fn(unsigned int hooknum,
 		else if (!(IPCB(skb)->flags & IPSKB_XFRM_TRANSFORMED) &&
 			 ct->tuplehash[dir].tuple.dst.protonum != IPPROTO_ICMP &&
 			 ct->tuplehash[dir].tuple.dst.u.all !=
-			 ct->tuplehash[!dir].tuple.src.u.all)
-			if (nf_xfrm_me_harder(skb, AF_INET) < 0)
-				ret = NF_DROP;
+			 ct->tuplehash[!dir].tuple.src.u.all) {
+			err = nf_xfrm_me_harder(skb, AF_INET);
+			if (err < 0)
+				ret = NF_DROP_ERR(err);
+		}
 #endif
 	}
 	return ret;
diff --git a/net/ipv6/netfilter/ip6table_nat.c b/net/ipv6/netfilter/ip6table_nat.c
index 97e2edd..6383f90 100644
--- a/net/ipv6/netfilter/ip6table_nat.c
+++ b/net/ipv6/netfilter/ip6table_nat.c
@@ -179,6 +179,7 @@ nf_nat_ipv6_out(unsigned int hooknum,
 #ifdef CONFIG_XFRM
 	const struct nf_conn *ct;
 	enum ip_conntrack_info ctinfo;
+	int err;
 #endif
 	unsigned int ret;
 
@@ -197,9 +198,11 @@ nf_nat_ipv6_out(unsigned int hooknum,
 				      &ct->tuplehash[!dir].tuple.dst.u3) ||
 		    (ct->tuplehash[dir].tuple.dst.protonum != IPPROTO_ICMPV6 &&
 		     ct->tuplehash[dir].tuple.src.u.all !=
-		     ct->tuplehash[!dir].tuple.dst.u.all))
-			if (nf_xfrm_me_harder(skb, AF_INET6) < 0)
-				ret = NF_DROP;
+		     ct->tuplehash[!dir].tuple.dst.u.all)) {
+			err = nf_xfrm_me_harder(skb, AF_INET6);
+			if (err < 0)
+				ret = NF_DROP_ERR(err);
+		}
 	}
 #endif
 	return ret;
@@ -236,9 +239,11 @@ nf_nat_ipv6_local_fn(unsigned int hooknum,
 		else if (!(IP6CB(skb)->flags & IP6SKB_XFRM_TRANSFORMED) &&
 			 ct->tuplehash[dir].tuple.dst.protonum != IPPROTO_ICMPV6 &&
 			 ct->tuplehash[dir].tuple.dst.u.all !=
-			 ct->tuplehash[!dir].tuple.src.u.all)
-			if (nf_xfrm_me_harder(skb, AF_INET6))
-				ret = NF_DROP;
+			 ct->tuplehash[!dir].tuple.src.u.all) {
+			err = nf_xfrm_me_harder(skb, AF_INET6);
+			if (err < 0)
+				ret = NF_DROP_ERR(err);
+		}
 #endif
 	}
 	return ret;
diff --git a/net/netfilter/nf_nat_core.c b/net/netfilter/nf_nat_core.c
index 8d5769c..346f871 100644
--- a/net/netfilter/nf_nat_core.c
+++ b/net/netfilter/nf_nat_core.c
@@ -87,9 +87,10 @@ int nf_xfrm_me_harder(struct sk_buff *skb, unsigned int family)
 	struct flowi fl;
 	unsigned int hh_len;
 	struct dst_entry *dst;
+	int err;
 
-	if (xfrm_decode_session(skb, &fl, family) < 0)
-		return -1;
+	err = xfrm_decode_session(skb, &fl, family);
+		return err;
 
 	dst = skb_dst(skb);
 	if (dst->xfrm)
@@ -98,7 +99,7 @@ int nf_xfrm_me_harder(struct sk_buff *skb, unsigned int family)
 
 	dst = xfrm_lookup(dev_net(dst->dev), dst, &fl, skb->sk, 0);
 	if (IS_ERR(dst))
-		return -1;
+		return PTR_ERR(dst);
 
 	skb_dst_drop(skb);
 	skb_dst_set(skb, dst);
@@ -107,7 +108,7 @@ int nf_xfrm_me_harder(struct sk_buff *skb, unsigned int family)
 	hh_len = skb_dst(skb)->dev->hard_header_len;
 	if (skb_headroom(skb) < hh_len &&
 	    pskb_expand_head(skb, hh_len - skb_headroom(skb), 0, GFP_ATOMIC))
-		return -1;
+		return -ENOMEM;
 	return 0;
 }
 EXPORT_SYMBOL(nf_xfrm_me_harder);
-- 
1.8.1.4

Re: [PATCH 0/3] netfilter: propagate routing and XFRM errors

[Pablo Neira Ayuso]

On Fri, Apr 05, 2013 at 06:41:09PM +0200, Patrick McHardy wrote:
> These patches fix a long standing annoyance: netfilter will return
> EPERM for all errors during rerouting or XFRM policy lookups.
> 
> These patches change {ip,ip6}_route_me_harder() and nf_xfrm_me_harder()
> to return errno codes and use NF_DROP_ERR() to propagate those back
> to userspace.

Applied, thanks Patrick!