net: erspan: refactor existing erspan code

The patch refactors the existing erspan implementation in order
to support erspan version 2, which has additional metadata. So, in
stead of having one 'struct erspanhdr' holding erspan version 1,
breaks it into 'struct erspan_base_hdr' and 'struct erspan_metadata'.

Signed-off-by: William Tu <[e-mail]>
Signed-off-by: David S. Miller <[e-mail]>

include/net/erspan.h

···
  *  s, Recur, Flags, Version fields only S (bit 03) is set to 1. The
  *  other fields are set to zero, so only a sequence number follows.
  *
- *  ERSPAN Type II header (8 octets [42:49])
+ *  ERSPAN Version 1 (Type II) header (8 octets [42:49])
  *  0                   1                   2                   3
  *  0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
  * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
···
  * GRE proto ERSPAN type II = 0x88BE, type III = 0x22EB
  */
 
-#define ERSPAN_VERSION	0x1
+#define ERSPAN_VERSION	0x1	/* ERSPAN type II */
 
 #define VER_MASK	0xf000
 #define VLAN_MASK	0x0fff
···
 	ERSPAN_ENCAP_INFRAME = 0x3,	/* VLAN tag perserved in frame */
 };
 
+#define ERSPAN_V1_MDSIZE	4
+#define ERSPAN_V2_MDSIZE	8
 struct erspan_metadata {
-	__be32 index;   /* type II */
+	union {
+		__be32 index;	/* Version 1 (type II)*/
+	} u;
 };
 
-struct erspanhdr {
+struct erspan_base_hdr {
 	__be16 ver_vlan;
 #define VER_OFFSET  12
 	__be16 session_id;
 #define COS_OFFSET  13
 #define EN_OFFSET   11
 #define T_OFFSET    10
-	struct erspan_metadata md;
 };
+
+static inline int erspan_hdr_len(int version)
+{
+	return sizeof(struct erspan_base_hdr) +
+	       (version == 1 ? ERSPAN_V1_MDSIZE : ERSPAN_V2_MDSIZE);
+}
 
 static inline u8 tos_to_cos(u8 tos)
 {
···
 {
 	struct ethhdr *eth = eth_hdr(skb);
 	enum erspan_encap_type enc_type;
-	struct erspanhdr *ershdr;
+	struct erspan_base_hdr *ershdr;
+	struct erspan_metadata *ersmd;
 	struct qtag_prefix {
 		__be16 eth_type;
 		__be16 tci;
···
 		enc_type = ERSPAN_ENCAP_INFRAME;
 	}
 
-	skb_push(skb, sizeof(*ershdr));
-	ershdr = (struct erspanhdr *)skb->data;
-	memset(ershdr, 0, sizeof(*ershdr));
+	skb_push(skb, sizeof(*ershdr) + ERSPAN_V1_MDSIZE);
+	ershdr = (struct erspan_base_hdr *)skb->data;
+	memset(ershdr, 0, sizeof(*ershdr) + ERSPAN_V1_MDSIZE);
 
+	/* Build base header */
 	ershdr->ver_vlan = htons((vlan_tci & VLAN_MASK) |
 				 (ERSPAN_VERSION << VER_OFFSET));
 	ershdr->session_id = htons((u16)(ntohl(id) & ID_MASK) |
 			   ((tos_to_cos(tos) << COS_OFFSET) & COS_MASK) |
 			   (enc_type << EN_OFFSET & EN_MASK) |
 			   ((truncate << T_OFFSET) & T_MASK));
-	ershdr->md.index = htonl(index & INDEX_MASK);
+
+	/* Build metadata */
+	ersmd = (struct erspan_metadata *)(ershdr + 1);
+	ersmd->u.index = htonl(index & INDEX_MASK);
 }
 
 #endif

net/ipv4/ip_gre.c

···
 {
 	struct net *net = dev_net(skb->dev);
 	struct metadata_dst *tun_dst = NULL;
+	struct erspan_base_hdr *ershdr;
+	struct erspan_metadata *pkt_md;
 	struct ip_tunnel_net *itn;
 	struct ip_tunnel *tunnel;
-	struct erspanhdr *ershdr;
 	const struct iphdr *iph;
-	__be32 index;
+	int ver;
 	int len;
 
 	itn = net_generic(net, erspan_net_id);
 	len = gre_hdr_len + sizeof(*ershdr);
 
+	/* Check based hdr len */
 	if (unlikely(!pskb_may_pull(skb, len)))
 		return -ENOMEM;
 
 	iph = ip_hdr(skb);
-	ershdr = (struct erspanhdr *)(skb->data + gre_hdr_len);
+	ershdr = (struct erspan_base_hdr *)(skb->data + gre_hdr_len);
+	ver = (ntohs(ershdr->ver_vlan) & VER_MASK) >> VER_OFFSET;
 
 	/* The original GRE header does not have key field,
 	 * Use ERSPAN 10-bit session ID as key.
 	 */
 	tpi->key = cpu_to_be32(ntohs(ershdr->session_id) & ID_MASK);
-	index = ershdr->md.index;
+	pkt_md = (struct erspan_metadata *)(ershdr + 1);
 	tunnel = ip_tunnel_lookup(itn, skb->dev->ifindex,
 				  tpi->flags | TUNNEL_KEY,
 				  iph->saddr, iph->daddr, tpi->key);
 
 	if (tunnel) {
+		len = gre_hdr_len + erspan_hdr_len(ver);
+		if (unlikely(!pskb_may_pull(skb, len)))
+			return -ENOMEM;
+
 		if (__iptunnel_pull_header(skb,
-					   gre_hdr_len + sizeof(*ershdr),
+					   len,
 					   htons(ETH_P_TEB),
 					   false, false) < 0)
 			goto drop;
···
 			if (!md)
 				return PACKET_REJECT;
 
-			md->index = index;
+			memcpy(md, pkt_md, sizeof(*md));
 			info = &tun_dst->u.tun_info;
 			info->key.tun_flags |= TUNNEL_ERSPAN_OPT;
 			info->options_len = sizeof(*md);
 		} else {
-			tunnel->index = ntohl(index);
+			tunnel->index = ntohl(pkt_md->u.index);
 		}
 
 		skb_reset_mac_header(skb);
···
 	key = &tun_info->key;
 
 	/* ERSPAN has fixed 8 byte GRE header */
-	tunnel_hlen = 8 + sizeof(struct erspanhdr);
+	tunnel_hlen = 8 + sizeof(struct erspan_base_hdr) + ERSPAN_V1_MDSIZE;
 
 	rt = prepare_fb_xmit(skb, dev, &fl, tunnel_hlen);
 	if (!rt)
···
 		goto err_free_rt;
 
 	erspan_build_header(skb, tunnel_id_to_key32(key->tun_id),
-			    ntohl(md->index), truncate, true);
+			    ntohl(md->u.index), truncate, true);
 
 	gre_build_header(skb, 8, TUNNEL_SEQ,
 			 htons(ETH_P_ERSPAN), 0, htonl(tunnel->o_seqno++));
···
 	tunnel->tun_hlen = 8;
 	tunnel->parms.iph.protocol = IPPROTO_GRE;
 	tunnel->hlen = tunnel->tun_hlen + tunnel->encap_hlen +
-		       sizeof(struct erspanhdr);
+		       sizeof(struct erspan_base_hdr) + ERSPAN_V1_MDSIZE;
 	t_hlen = tunnel->hlen + sizeof(struct iphdr);
 
 	dev->needed_headroom = LL_MAX_HEADER + t_hlen + 4;

net/ipv6/ip6_gre.c

···
 static int ip6erspan_rcv(struct sk_buff *skb, int gre_hdr_len,
 			 struct tnl_ptk_info *tpi)
 {
+	struct erspan_base_hdr *ershdr;
+	struct erspan_metadata *pkt_md;
 	const struct ipv6hdr *ipv6h;
-	struct erspanhdr *ershdr;
 	struct ip6_tnl *tunnel;
-	__be32 index;
+	u8 ver;
 
 	ipv6h = ipv6_hdr(skb);
-	ershdr = (struct erspanhdr *)skb->data;
+	ershdr = (struct erspan_base_hdr *)skb->data;
 
 	if (unlikely(!pskb_may_pull(skb, sizeof(*ershdr))))
 		return PACKET_REJECT;
 
+	ver = (ntohs(ershdr->ver_vlan) & VER_MASK) >> VER_OFFSET;
 	tpi->key = cpu_to_be32(ntohs(ershdr->session_id) & ID_MASK);
-	index = ershdr->md.index;
+	pkt_md = (struct erspan_metadata *)(ershdr + 1);
 
 	tunnel = ip6gre_tunnel_lookup(skb->dev,
 				      &ipv6h->saddr, &ipv6h->daddr, tpi->key,
 				      tpi->proto);
 	if (tunnel) {
-		if (__iptunnel_pull_header(skb, sizeof(*ershdr),
+		int len = erspan_hdr_len(ver);
+
+		if (unlikely(!pskb_may_pull(skb, len)))
+			return -ENOMEM;
+
+		if (__iptunnel_pull_header(skb, len,
 					   htons(ETH_P_TEB),
 					   false, false) < 0)
 			return PACKET_REJECT;
···
 			if (!md)
 				return PACKET_REJECT;
 
-			md->index = index;
+			memcpy(md, pkt_md, sizeof(*md));
 			info->key.tun_flags |= TUNNEL_ERSPAN_OPT;
 			info->options_len = sizeof(*md);
 
 			ip6_tnl_rcv(tunnel, skb, tpi, tun_dst, log_ecn_error);
 
 		} else {
-			tunnel->parms.index = ntohl(index);
+			tunnel->parms.index = ntohl(pkt_md->u.index);
 			ip6_tnl_rcv(tunnel, skb, tpi, NULL, log_ecn_error);
 		}
 
···
 			goto tx_err;
 
 		erspan_build_header(skb, tunnel_id_to_key32(key->tun_id),
-				    ntohl(md->index), truncate, false);
+				    ntohl(md->u.index), truncate, false);
 
 	} else {
 		switch (skb->protocol) {
···
 
 	tunnel->tun_hlen = 8;
 	tunnel->hlen = tunnel->tun_hlen + tunnel->encap_hlen +
-		       sizeof(struct erspanhdr);
+		       sizeof(struct erspan_base_hdr) + ERSPAN_V1_MDSIZE;
 	t_hlen = tunnel->hlen + sizeof(struct ipv6hdr);
 
 	dev->hard_header_len = LL_MAX_HEADER + t_hlen;

net/openvswitch/flow_netlink.c

···
 	BUILD_BUG_ON(sizeof(opts) > sizeof(match->key->tun_opts));
 
 	memset(&opts, 0, sizeof(opts));
-	opts.index = nla_get_be32(attr);
+	opts.u.index = nla_get_be32(attr);
 
 	/* Index has only 20-bit */
-	if (ntohl(opts.index) & ~INDEX_MASK) {
+	if (ntohl(opts.u.index) & ~INDEX_MASK) {
 		OVS_NLERR(log, "ERSPAN index number %x too large.",
-			  ntohl(opts.index));
+			  ntohl(opts.u.index));
 		return -EINVAL;
 	}
 
···
 			return -EMSGSIZE;
 		else if (output->tun_flags & TUNNEL_ERSPAN_OPT &&
 			 nla_put_be32(skb, OVS_TUNNEL_KEY_ATTR_ERSPAN_OPTS,
-				      ((struct erspan_metadata *)tun_opts)->index))
+				      ((struct erspan_metadata *)tun_opts)->u.index))
 			return -EMSGSIZE;
 	}