netfilter: nft_meta: add double-tagged vlan and pppoe support

Currently:

  add rule netdev x y ip saddr 1.1.1.1

does not work with neither double-tagged vlan nor pppoe packets. This is
because the network and transport header offset are not pointing to the
IP and transport protocol headers in the stack.

This patch expands NFT_META_PROTOCOL and NFT_META_L4PROTO to parse
double-tagged vlan and pppoe packets so matching network and transport
header fields becomes possible with the existing userspace generated
bytecode. Note that this parser only supports double-tagged vlan which
is composed of vlan offload + vlan header in the skb payload area for
simplicity.

NFT_META_PROTOCOL is used by bridge and netdev family as an implicit
dependency in the bytecode to match on network header fields.
Similarly, there is also NFT_META_L4PROTO, which is also used as an
implicit dependency when matching on the transport protocol header
fields.

Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Florian Westphal <fw@strlen.de>

include/net/netfilter/nf_tables.h

@@ -31,7 +31,9 @@ struct nft_pktinfo {
 	const struct nf_hook_state	*state;
 	u8				flags;
 	u8				tprot;
+	__be16				ethertype;
 	u16				fragoff;
+	u16				nhoff;
 	u16				thoff;
 	u16				inneroff;
 };
@@ -83,6 +85,8 @@ static inline void nft_set_pktinfo_unspec(struct nft_pktinfo *pkt)
 {
 	pkt->flags = 0;
 	pkt->tprot = 0;
+	pkt->ethertype = pkt->skb->protocol;
+	pkt->nhoff = 0;
 	pkt->thoff = 0;
 	pkt->fragoff = 0;
 }

include/net/netfilter/nf_tables_ipv4.h

@@ -12,16 +12,19 @@ static inline void nft_set_pktinfo_ipv4(struct nft_pktinfo *pkt)
 	ip = ip_hdr(pkt->skb);
 	pkt->flags = NFT_PKTINFO_L4PROTO;
 	pkt->tprot = ip->protocol;
+	pkt->ethertype = pkt->skb->protocol;
+	pkt->nhoff = 0;
 	pkt->thoff = ip_hdrlen(pkt->skb);
 	pkt->fragoff = ntohs(ip->frag_off) & IP_OFFSET;
 }
 
-static inline int __nft_set_pktinfo_ipv4_validate(struct nft_pktinfo *pkt)
+static inline int __nft_set_pktinfo_ipv4_validate(struct nft_pktinfo *pkt,
+						  int nhoff)
 {
 	struct iphdr *iph, _iph;
 	u32 len, thoff, skb_len;
 
-	iph = skb_header_pointer(pkt->skb, skb_network_offset(pkt->skb),
+	iph = skb_header_pointer(pkt->skb, skb_network_offset(pkt->skb) + nhoff,
 				 sizeof(*iph), &_iph);
 	if (!iph)
 		return -1;
@@ -31,7 +34,7 @@ static inline int __nft_set_pktinfo_ipv4_validate(struct nft_pktinfo *pkt)
 
 	len = iph_totlen(pkt->skb, iph);
 	thoff = iph->ihl * 4;
-	skb_len = pkt->skb->len - skb_network_offset(pkt->skb);
+	skb_len = pkt->skb->len - skb_network_offset(pkt->skb) - nhoff;
 
 	if (skb_len < len)
 		return -1;
@@ -42,7 +45,9 @@ static inline int __nft_set_pktinfo_ipv4_validate(struct nft_pktinfo *pkt)
 
 	pkt->flags = NFT_PKTINFO_L4PROTO;
 	pkt->tprot = iph->protocol;
-	pkt->thoff = skb_network_offset(pkt->skb) + thoff;
+	pkt->ethertype = pkt->skb->protocol;
+	pkt->nhoff = nhoff;
+	pkt->thoff = skb_network_offset(pkt->skb) + nhoff + thoff;
 	pkt->fragoff = ntohs(iph->frag_off) & IP_OFFSET;
 
 	return 0;
@@ -50,7 +55,7 @@ static inline int __nft_set_pktinfo_ipv4_validate(struct nft_pktinfo *pkt)
 
 static inline void nft_set_pktinfo_ipv4_validate(struct nft_pktinfo *pkt)
 {
-	if (__nft_set_pktinfo_ipv4_validate(pkt) < 0)
+	if (__nft_set_pktinfo_ipv4_validate(pkt, 0) < 0)
 		nft_set_pktinfo_unspec(pkt);
 }
 
@@ -78,6 +83,8 @@ static inline int nft_set_pktinfo_ipv4_ingress(struct nft_pktinfo *pkt)
 	}
 
 	pkt->flags = NFT_PKTINFO_L4PROTO;
+	pkt->ethertype = pkt->skb->protocol;
+	pkt->nhoff = 0;
 	pkt->tprot = iph->protocol;
 	pkt->thoff = thoff;
 	pkt->fragoff = ntohs(iph->frag_off) & IP_OFFSET;

include/net/netfilter/nf_tables_ipv6.h

@@ -20,21 +20,23 @@ static inline void nft_set_pktinfo_ipv6(struct nft_pktinfo *pkt)
 
 	pkt->flags = NFT_PKTINFO_L4PROTO;
 	pkt->tprot = protohdr;
+	pkt->ethertype = pkt->skb->protocol;
+	pkt->nhoff = 0;
 	pkt->thoff = thoff;
 	pkt->fragoff = frag_off;
 }
 
-static inline int __nft_set_pktinfo_ipv6_validate(struct nft_pktinfo *pkt)
+static inline int __nft_set_pktinfo_ipv6_validate(struct nft_pktinfo *pkt, int nhoff)
 {
 #if IS_ENABLED(CONFIG_IPV6)
 	unsigned int flags = IP6_FH_F_AUTH;
 	struct ipv6hdr *ip6h, _ip6h;
-	unsigned int thoff = 0;
+	unsigned int thoff = nhoff;
 	unsigned short frag_off;
 	u32 pkt_len, skb_len;
 	int protohdr;
 
-	ip6h = skb_header_pointer(pkt->skb, skb_network_offset(pkt->skb),
+	ip6h = skb_header_pointer(pkt->skb, skb_network_offset(pkt->skb) + nhoff,
 				  sizeof(*ip6h), &_ip6h);
 	if (!ip6h)
 		return -1;
@@ -43,7 +45,7 @@ static inline int __nft_set_pktinfo_ipv6_validate(struct nft_pktinfo *pkt)
 		return -1;
 
 	pkt_len = ipv6_payload_len(pkt->skb, ip6h);
-	skb_len = pkt->skb->len - skb_network_offset(pkt->skb);
+	skb_len = pkt->skb->len - skb_network_offset(pkt->skb) - nhoff;
 	if (pkt_len + sizeof(*ip6h) > skb_len)
 		return -1;
 
@@ -53,6 +55,8 @@ static inline int __nft_set_pktinfo_ipv6_validate(struct nft_pktinfo *pkt)
 
 	pkt->flags = NFT_PKTINFO_L4PROTO;
 	pkt->tprot = protohdr;
+	pkt->ethertype = pkt->skb->protocol;
+	pkt->nhoff = nhoff;
 	pkt->thoff = thoff;
 	pkt->fragoff = frag_off;
 
@@ -64,7 +68,7 @@ static inline int __nft_set_pktinfo_ipv6_validate(struct nft_pktinfo *pkt)
 
 static inline void nft_set_pktinfo_ipv6_validate(struct nft_pktinfo *pkt)
 {
-	if (__nft_set_pktinfo_ipv6_validate(pkt) < 0)
+	if (__nft_set_pktinfo_ipv6_validate(pkt, 0) < 0)
 		nft_set_pktinfo_unspec(pkt);
 }
 
@@ -99,6 +103,8 @@ static inline int nft_set_pktinfo_ipv6_ingress(struct nft_pktinfo *pkt)
 
 	pkt->flags = NFT_PKTINFO_L4PROTO;
 	pkt->tprot = protohdr;
+	pkt->ethertype = pkt->skb->protocol;
+	pkt->nhoff = 0;
 	pkt->thoff = thoff;
 	pkt->fragoff = frag_off;
 

net/netfilter/nf_tables_core.c

@@ -151,7 +151,7 @@ static bool nft_payload_fast_eval(const struct nft_expr *expr,
 	unsigned char *ptr;
 
 	if (priv->base == NFT_PAYLOAD_NETWORK_HEADER)
-		ptr = skb_network_header(skb);
+		ptr = skb_network_header(skb) + pkt->nhoff;
 	else {
 		if (!(pkt->flags & NFT_PKTINFO_L4PROTO))
 			return false;

net/netfilter/nft_meta.c

@@ -23,6 +23,8 @@
 #include <net/tcp_states.h> /* for TCP_TIME_WAIT */
 #include <net/netfilter/nf_tables.h>
 #include <net/netfilter/nf_tables_core.h>
+#include <net/netfilter/nf_tables_ipv4.h>
+#include <net/netfilter/nf_tables_ipv6.h>
 #include <net/netfilter/nft_meta.h>
 #include <net/netfilter/nf_tables_offload.h>
 
@@ -309,6 +311,54 @@ nft_meta_get_eval_sdifname(u32 *dest, const struct nft_pktinfo *pkt)
 	nft_meta_store_ifname(dest, dev);
 }
 
+static void nft_meta_pktinfo_may_update(struct nft_pktinfo *pkt)
+{
+	struct sk_buff *skb = pkt->skb;
+	struct vlan_ethhdr *veth;
+	__be16 ethertype;
+	int nhoff;
+
+	/* Is this an IP packet? Then, skip. */
+	if (pkt->flags)
+		return;
+
+	/* ... else maybe an IP packet over PPPoE or Q-in-Q? */
+	switch (skb->protocol) {
+	case htons(ETH_P_8021Q):
+		if (!pskb_may_pull(skb, skb_mac_offset(skb) + sizeof(*veth)))
+			return;
+
+		veth = (struct vlan_ethhdr *)skb_mac_header(skb);
+		nhoff = VLAN_HLEN;
+		ethertype = veth->h_vlan_encapsulated_proto;
+		break;
+	case htons(ETH_P_PPP_SES):
+		if (!nf_flow_pppoe_proto(skb, &ethertype))
+			return;
+
+		nhoff = PPPOE_SES_HLEN;
+		break;
+	default:
+		return;
+	}
+
+	nhoff += skb_network_offset(skb);
+	switch (ethertype) {
+	case htons(ETH_P_IP):
+		if (__nft_set_pktinfo_ipv4_validate(pkt, nhoff))
+			nft_set_pktinfo_unspec(pkt);
+		break;
+	case htons(ETH_P_IPV6):
+		if (__nft_set_pktinfo_ipv6_validate(pkt, nhoff))
+			nft_set_pktinfo_unspec(pkt);
+		break;
+	default:
+		break;
+	}
+
+	pkt->ethertype = ethertype;
+}
+
 void nft_meta_get_eval(const struct nft_expr *expr,
 		       struct nft_regs *regs,
 		       const struct nft_pktinfo *pkt)
@@ -322,12 +372,14 @@ void nft_meta_get_eval(const struct nft_expr *expr,
 		*dest = skb->len;
 		break;
 	case NFT_META_PROTOCOL:
-		nft_reg_store16(dest, (__force u16)skb->protocol);
+		nft_meta_pktinfo_may_update((struct nft_pktinfo *)pkt);
+		nft_reg_store16(dest, (__force u16)pkt->ethertype);
 		break;
 	case NFT_META_NFPROTO:
 		nft_reg_store8(dest, nft_pf(pkt));
 		break;
 	case NFT_META_L4PROTO:
+		nft_meta_pktinfo_may_update((struct nft_pktinfo *)pkt);
 		if (!(pkt->flags & NFT_PKTINFO_L4PROTO))
 			goto err;
 		nft_reg_store8(dest, pkt->tprot);

net/netfilter/nft_payload.c

@@ -183,7 +183,7 @@ void nft_payload_eval(const struct nft_expr *expr,
 		offset = skb_mac_header(skb) - skb->data;
 		break;
 	case NFT_PAYLOAD_NETWORK_HEADER:
-		offset = skb_network_offset(skb);
+		offset = skb_network_offset(skb) + pkt->nhoff;
 		break;
 	case NFT_PAYLOAD_TRANSPORT_HEADER:
 		if (!(pkt->flags & NFT_PKTINFO_L4PROTO) || pkt->fragoff)