net: geneve: support IPv4/IPv6 as inner protocol

This patch adds support for encapsulating IPv4/IPv6 within GENEVE.

In order to use this, a new IFLA_GENEVE_INNER_PROTO_INHERIT flag needs
to be provided at device creation. This property cannot be changed for
the time being.

In case IP traffic is received on a non-tun device the drop count is
increased.

Signed-off-by: Eyal Birger <[e-mail]>
Link: https://lore.kernel.org/r/[e-mail]
Signed-off-by: Paolo Abeni <[e-mail]>

drivers/net/geneve.c

···
 	bool			use_udp6_rx_checksums;
 	bool			ttl_inherit;
 	enum ifla_geneve_df	df;
+	bool			inner_proto_inherit;
 };
 
 /* Pseudo network device */
···
 		}
 	}
 
-	skb_reset_mac_header(skb);
-	skb->protocol = eth_type_trans(skb, geneve->dev);
-	skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN);
-
 	if (tun_dst)
 		skb_dst_set(skb, &tun_dst->dst);
 
-	/* Ignore packet loops (and multicast echo) */
-	if (ether_addr_equal(eth_hdr(skb)->h_source, geneve->dev->dev_addr)) {
-		geneve->dev->stats.rx_errors++;
-		goto drop;
+	if (gnvh->proto_type == htons(ETH_P_TEB)) {
+		skb_reset_mac_header(skb);
+		skb->protocol = eth_type_trans(skb, geneve->dev);
+		skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN);
+
+		/* Ignore packet loops (and multicast echo) */
+		if (ether_addr_equal(eth_hdr(skb)->h_source,
+				     geneve->dev->dev_addr)) {
+			geneve->dev->stats.rx_errors++;
+			goto drop;
+		}
+	} else {
+		skb_reset_mac_header(skb);
+		skb->dev = geneve->dev;
+		skb->pkt_type = PACKET_HOST;
 	}
 
 	oiph = skb_network_header(skb);
···
 	struct genevehdr *geneveh;
 	struct geneve_dev *geneve;
 	struct geneve_sock *gs;
+	__be16 inner_proto;
 	int opts_len;
 
 	/* Need UDP and Geneve header to be present */
···
 	if (unlikely(geneveh->ver != GENEVE_VER))
 		goto drop;
 
-	if (unlikely(geneveh->proto_type != htons(ETH_P_TEB)))
+	inner_proto = geneveh->proto_type;
+
+	if (unlikely((inner_proto != htons(ETH_P_TEB) &&
+		      inner_proto != htons(ETH_P_IP) &&
+		      inner_proto != htons(ETH_P_IPV6))))
 		goto drop;
 
 	gs = rcu_dereference_sk_user_data(sk);
···
 	if (!geneve)
 		goto drop;
 
+	if (unlikely((!geneve->cfg.inner_proto_inherit &&
+		      inner_proto != htons(ETH_P_TEB)))) {
+		geneve->dev->stats.rx_dropped++;
+		goto drop;
+	}
+
 	opts_len = geneveh->opt_len * 4;
-	if (iptunnel_pull_header(skb, GENEVE_BASE_HLEN + opts_len,
-				 htons(ETH_P_TEB),
+	if (iptunnel_pull_header(skb, GENEVE_BASE_HLEN + opts_len, inner_proto,
 				 !net_eq(geneve->net, dev_net(geneve->dev)))) {
 		geneve->dev->stats.rx_dropped++;
 		goto drop;
···
 }
 
 static void geneve_build_header(struct genevehdr *geneveh,
-				const struct ip_tunnel_info *info)
+				const struct ip_tunnel_info *info,
+				__be16 inner_proto)
 {
 	geneveh->ver = GENEVE_VER;
 	geneveh->opt_len = info->options_len / 4;
···
 	geneveh->critical = !!(info->key.tun_flags & TUNNEL_CRIT_OPT);
 	geneveh->rsvd1 = 0;
 	tunnel_id_to_vni(info->key.tun_id, geneveh->vni);
-	geneveh->proto_type = htons(ETH_P_TEB);
+	geneveh->proto_type = inner_proto;
 	geneveh->rsvd2 = 0;
 
 	if (info->key.tun_flags & TUNNEL_GENEVE_OPT)
···
 
 static int geneve_build_skb(struct dst_entry *dst, struct sk_buff *skb,
 			    const struct ip_tunnel_info *info,
-			    bool xnet, int ip_hdr_len)
+			    bool xnet, int ip_hdr_len,
+			    bool inner_proto_inherit)
 {
 	bool udp_sum = !!(info->key.tun_flags & TUNNEL_CSUM);
 	struct genevehdr *gnvh;
+	__be16 inner_proto;
 	int min_headroom;
 	int err;
 
···
 		goto free_dst;
 
 	gnvh = __skb_push(skb, sizeof(*gnvh) + info->options_len);
-	geneve_build_header(gnvh, info);
-	skb_set_inner_protocol(skb, htons(ETH_P_TEB));
+	inner_proto = inner_proto_inherit ? skb->protocol : htons(ETH_P_TEB);
+	geneve_build_header(gnvh, info, inner_proto);
+	skb_set_inner_protocol(skb, inner_proto);
 	return 0;
 
 free_dst:
···
 		}
 	}
 
-	err = geneve_build_skb(&rt->dst, skb, info, xnet, sizeof(struct iphdr));
+	err = geneve_build_skb(&rt->dst, skb, info, xnet, sizeof(struct iphdr),
+			       geneve->cfg.inner_proto_inherit);
 	if (unlikely(err))
 		return err;
 
···
 			ttl = key->ttl;
 		ttl = ttl ? : ip6_dst_hoplimit(dst);
 	}
-	err = geneve_build_skb(dst, skb, info, xnet, sizeof(struct ipv6hdr));
+	err = geneve_build_skb(dst, skb, info, xnet, sizeof(struct ipv6hdr),
+			       geneve->cfg.inner_proto_inherit);
 	if (unlikely(err))
 		return err;
 
···
 	dst_cache_reset(&geneve->cfg.info.dst_cache);
 	memcpy(&geneve->cfg, cfg, sizeof(*cfg));
 
+	if (geneve->cfg.inner_proto_inherit) {
+		dev->header_ops = NULL;
+		dev->type = ARPHRD_NONE;
+		dev->hard_header_len = 0;
+		dev->addr_len = 0;
+		dev->flags = IFF_NOARP;
+	}
+
 	err = register_netdevice(dev);
 	if (err)
 		return err;
···
 #endif
 	}
 
+	if (data[IFLA_GENEVE_INNER_PROTO_INHERIT]) {
+		if (changelink) {
+			attrtype = IFLA_GENEVE_INNER_PROTO_INHERIT;
+			goto change_notsup;
+		}
+		cfg->inner_proto_inherit = true;
+	}
+
 	return 0;
 change_notsup:
 	NL_SET_ERR_MSG_ATTR(extack, data[attrtype],
-			    "Changing VNI, Port, endpoint IP address family, external, and UDP checksum attributes are not supported");
+			    "Changing VNI, Port, endpoint IP address family, external, inner_proto_inherit, and UDP checksum attributes are not supported");
 	return -EOPNOTSUPP;
 }
 
···
 #endif
 
 	if (nla_put_u8(skb, IFLA_GENEVE_TTL_INHERIT, ttl_inherit))
+		goto nla_put_failure;
+
+	if (geneve->cfg.inner_proto_inherit &&
+	    nla_put_flag(skb, IFLA_GENEVE_INNER_PROTO_INHERIT))
 		goto nla_put_failure;
 
 	return 0;

include/uapi/linux/if_link.h

···
 	IFLA_GENEVE_LABEL,
 	IFLA_GENEVE_TTL_INHERIT,
 	IFLA_GENEVE_DF,
+	IFLA_GENEVE_INNER_PROTO_INHERIT,
 	__IFLA_GENEVE_MAX
 };
 #define IFLA_GENEVE_MAX	(__IFLA_GENEVE_MAX - 1)