Merge branch 'ovs-ipv6-tunnel'

Jiri Benc says:

====================
openvswitch: add IPv6 tunneling support

This builds on the previous work that added IPv6 support to lwtunnels and
adds IPv6 tunneling support to ovs.

To use IPv6 tunneling, there needs to be a metadata based tunnel net_device
created and added to the ovs bridge. Currently, only vxlan is supported by
the kernel, with geneve to follow shortly. There's no need nor intent to add
a support for this into the vport-vxlan (etc.) compat layer.

v3: dropped the last two patches added in v2.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

5 files changed, 97 insertions, 41 deletions

diff --git a/include/uapi/linux/openvswitch.h b/include/uapi/linux/openvswitch.h
index 32e07d8cbaf4..4036e1b1980f 100644
--- a/include/uapi/linux/openvswitch.h
+++ b/include/uapi/linux/openvswitch.h
@@ -349,6 +349,8 @@ enum ovs_tunnel_key_attr {
 	OVS_TUNNEL_KEY_ATTR_TP_SRC,		/* be16 src Transport Port. */
 	OVS_TUNNEL_KEY_ATTR_TP_DST,		/* be16 dst Transport Port. */
 	OVS_TUNNEL_KEY_ATTR_VXLAN_OPTS,		/* Nested OVS_VXLAN_EXT_* */
+	OVS_TUNNEL_KEY_ATTR_IPV6_SRC,		/* struct in6_addr src IPv6 address. */
+	OVS_TUNNEL_KEY_ATTR_IPV6_DST,		/* struct in6_addr dst IPv6 address. */
 	__OVS_TUNNEL_KEY_ATTR_MAX
 };
 
diff --git a/net/openvswitch/flow.c b/net/openvswitch/flow.c
index c8db44ab2ee7..0ea128eeeab2 100644
--- a/net/openvswitch/flow.c
+++ b/net/openvswitch/flow.c
@@ -698,8 +698,7 @@ int ovs_flow_key_extract(const struct ip_tunnel_info *tun_info,
 {
 	/* Extract metadata from packet. */
 	if (tun_info) {
-		if (ip_tunnel_info_af(tun_info) != AF_INET)
-			return -EINVAL;
+		key->tun_proto = ip_tunnel_info_af(tun_info);
 		memcpy(&key->tun_key, &tun_info->key, sizeof(key->tun_key));
 
 		if (tun_info->options_len) {
@@ -714,6 +713,7 @@ int ovs_flow_key_extract(const struct ip_tunnel_info *tun_info,
 			key->tun_opts_len = 0;
 		}
 	} else  {
+		key->tun_proto = 0;
 		key->tun_opts_len = 0;
 		memset(&key->tun_key, 0, sizeof(key->tun_key));
 	}
diff --git a/net/openvswitch/flow.h b/net/openvswitch/flow.h
index fe527d2dd4b7..5688e33e2de6 100644
--- a/net/openvswitch/flow.h
+++ b/net/openvswitch/flow.h
@@ -63,6 +63,7 @@ struct sw_flow_key {
 		u32	skb_mark;	/* SKB mark. */
 		u16	in_port;	/* Input switch port (or DP_MAX_PORTS). */
 	} __packed phy; /* Safe when right after 'tun_key'. */
+	u8 tun_proto;			/* Protocol of encapsulating tunnel. */
 	u32 ovs_flow_hash;		/* Datapath computed hash value.  */
 	u32 recirc_id;			/* Recirculation ID.  */
 	struct {
diff --git a/net/openvswitch/flow_netlink.c b/net/openvswitch/flow_netlink.c
index 5c030a4d7338..77850f177a47 100644
--- a/net/openvswitch/flow_netlink.c
+++ b/net/openvswitch/flow_netlink.c
@@ -262,8 +262,8 @@ size_t ovs_tun_key_attr_size(void)
 	 * updating this function.
 	 */
 	return    nla_total_size(8)    /* OVS_TUNNEL_KEY_ATTR_ID */
-		+ nla_total_size(4)    /* OVS_TUNNEL_KEY_ATTR_IPV4_SRC */
-		+ nla_total_size(4)    /* OVS_TUNNEL_KEY_ATTR_IPV4_DST */
+		+ nla_total_size(16)   /* OVS_TUNNEL_KEY_ATTR_IPV[46]_SRC */
+		+ nla_total_size(16)   /* OVS_TUNNEL_KEY_ATTR_IPV[46]_DST */
 		+ nla_total_size(1)    /* OVS_TUNNEL_KEY_ATTR_TOS */
 		+ nla_total_size(1)    /* OVS_TUNNEL_KEY_ATTR_TTL */
 		+ nla_total_size(0)    /* OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT */
@@ -323,6 +323,8 @@ static const struct ovs_len_tbl ovs_tunnel_key_lens[OVS_TUNNEL_KEY_ATTR_MAX + 1]
 	[OVS_TUNNEL_KEY_ATTR_GENEVE_OPTS]   = { .len = OVS_ATTR_VARIABLE },
 	[OVS_TUNNEL_KEY_ATTR_VXLAN_OPTS]    = { .len = OVS_ATTR_NESTED,
 						.next = ovs_vxlan_ext_key_lens },
+	[OVS_TUNNEL_KEY_ATTR_IPV6_SRC]      = { .len = sizeof(struct in6_addr) },
+	[OVS_TUNNEL_KEY_ATTR_IPV6_DST]      = { .len = sizeof(struct in6_addr) },
 };
 
 /* The size of the argument for each %OVS_KEY_ATTR_* Netlink attribute.  */
@@ -542,14 +544,14 @@ static int vxlan_tun_opt_from_nlattr(const struct nlattr *attr,
 	return 0;
 }
 
-static int ipv4_tun_from_nlattr(const struct nlattr *attr,
-				struct sw_flow_match *match, bool is_mask,
-				bool log)
+static int ip_tun_from_nlattr(const struct nlattr *attr,
+			      struct sw_flow_match *match, bool is_mask,
+			      bool log)
 {
 	struct nlattr *a;
 	int rem;
 	bool ttl = false;
-	__be16 tun_flags = 0;
+	__be16 tun_flags = 0, ipv4 = false, ipv6 = false;
 	int opts_type = 0;
 
 	nla_for_each_nested(a, attr, rem) {
@@ -578,10 +580,22 @@ static int ipv4_tun_from_nlattr(const struct nlattr *attr,
 		case OVS_TUNNEL_KEY_ATTR_IPV4_SRC:
 			SW_FLOW_KEY_PUT(match, tun_key.u.ipv4.src,
 					nla_get_in_addr(a), is_mask);
+			ipv4 = true;
 			break;
 		case OVS_TUNNEL_KEY_ATTR_IPV4_DST:
 			SW_FLOW_KEY_PUT(match, tun_key.u.ipv4.dst,
 					nla_get_in_addr(a), is_mask);
+			ipv4 = true;
+			break;
+		case OVS_TUNNEL_KEY_ATTR_IPV6_SRC:
+			SW_FLOW_KEY_PUT(match, tun_key.u.ipv6.dst,
+					nla_get_in6_addr(a), is_mask);
+			ipv6 = true;
+			break;
+		case OVS_TUNNEL_KEY_ATTR_IPV6_DST:
+			SW_FLOW_KEY_PUT(match, tun_key.u.ipv6.dst,
+					nla_get_in6_addr(a), is_mask);
+			ipv6 = true;
 			break;
 		case OVS_TUNNEL_KEY_ATTR_TOS:
 			SW_FLOW_KEY_PUT(match, tun_key.tos,
@@ -636,28 +650,46 @@ static int ipv4_tun_from_nlattr(const struct nlattr *attr,
 			opts_type = type;
 			break;
 		default:
-			OVS_NLERR(log, "Unknown IPv4 tunnel attribute %d",
+			OVS_NLERR(log, "Unknown IP tunnel attribute %d",
 				  type);
 			return -EINVAL;
 		}
 	}
 
 	SW_FLOW_KEY_PUT(match, tun_key.tun_flags, tun_flags, is_mask);
+	if (is_mask)
+		SW_FLOW_KEY_MEMSET_FIELD(match, tun_proto, 0xff, true);
+	else
+		SW_FLOW_KEY_PUT(match, tun_proto, ipv6 ? AF_INET6 : AF_INET,
+				false);
 
 	if (rem > 0) {
-		OVS_NLERR(log, "IPv4 tunnel attribute has %d unknown bytes.",
+		OVS_NLERR(log, "IP tunnel attribute has %d unknown bytes.",
 			  rem);
 		return -EINVAL;
 	}
 
+	if (ipv4 && ipv6) {
+		OVS_NLERR(log, "Mixed IPv4 and IPv6 tunnel attributes");
+		return -EINVAL;
+	}
+
 	if (!is_mask) {
-		if (!match->key->tun_key.u.ipv4.dst) {
+		if (!ipv4 && !ipv6) {
+			OVS_NLERR(log, "IP tunnel dst address not specified");
+			return -EINVAL;
+		}
+		if (ipv4 && !match->key->tun_key.u.ipv4.dst) {
 			OVS_NLERR(log, "IPv4 tunnel dst address is zero");
 			return -EINVAL;
 		}
+		if (ipv6 && ipv6_addr_any(&match->key->tun_key.u.ipv6.dst)) {
+			OVS_NLERR(log, "IPv6 tunnel dst address is zero");
+			return -EINVAL;
+		}
 
 		if (!ttl) {
-			OVS_NLERR(log, "IPv4 tunnel TTL not specified.");
+			OVS_NLERR(log, "IP tunnel TTL not specified.");
 			return -EINVAL;
 		}
 	}
@@ -682,21 +714,36 @@ static int vxlan_opt_to_nlattr(struct sk_buff *skb,
 	return 0;
 }
 
-static int __ipv4_tun_to_nlattr(struct sk_buff *skb,
-				const struct ip_tunnel_key *output,
-				const void *tun_opts, int swkey_tun_opts_len)
+static int __ip_tun_to_nlattr(struct sk_buff *skb,
+			      const struct ip_tunnel_key *output,
+			      const void *tun_opts, int swkey_tun_opts_len,
+			      unsigned short tun_proto)
 {
 	if (output->tun_flags & TUNNEL_KEY &&
 	    nla_put_be64(skb, OVS_TUNNEL_KEY_ATTR_ID, output->tun_id))
 		return -EMSGSIZE;
-	if (output->u.ipv4.src &&
-	    nla_put_in_addr(skb, OVS_TUNNEL_KEY_ATTR_IPV4_SRC,
-			    output->u.ipv4.src))
-		return -EMSGSIZE;
-	if (output->u.ipv4.dst &&
-	    nla_put_in_addr(skb, OVS_TUNNEL_KEY_ATTR_IPV4_DST,
-			    output->u.ipv4.dst))
-		return -EMSGSIZE;
+	switch (tun_proto) {
+	case AF_INET:
+		if (output->u.ipv4.src &&
+		    nla_put_in_addr(skb, OVS_TUNNEL_KEY_ATTR_IPV4_SRC,
+				    output->u.ipv4.src))
+			return -EMSGSIZE;
+		if (output->u.ipv4.dst &&
+		    nla_put_in_addr(skb, OVS_TUNNEL_KEY_ATTR_IPV4_DST,
+				    output->u.ipv4.dst))
+			return -EMSGSIZE;
+		break;
+	case AF_INET6:
+		if (!ipv6_addr_any(&output->u.ipv6.src) &&
+		    nla_put_in6_addr(skb, OVS_TUNNEL_KEY_ATTR_IPV6_SRC,
+				     &output->u.ipv6.src))
+			return -EMSGSIZE;
+		if (!ipv6_addr_any(&output->u.ipv6.dst) &&
+		    nla_put_in6_addr(skb, OVS_TUNNEL_KEY_ATTR_IPV6_DST,
+				     &output->u.ipv6.dst))
+			return -EMSGSIZE;
+		break;
+	}
 	if (output->tos &&
 	    nla_put_u8(skb, OVS_TUNNEL_KEY_ATTR_TOS, output->tos))
 		return -EMSGSIZE;
@@ -730,9 +777,10 @@ static int __ipv4_tun_to_nlattr(struct sk_buff *skb,
 	return 0;
 }
 
-static int ipv4_tun_to_nlattr(struct sk_buff *skb,
-			      const struct ip_tunnel_key *output,
-			      const void *tun_opts, int swkey_tun_opts_len)
+static int ip_tun_to_nlattr(struct sk_buff *skb,
+			    const struct ip_tunnel_key *output,
+			    const void *tun_opts, int swkey_tun_opts_len,
+			    unsigned short tun_proto)
 {
 	struct nlattr *nla;
 	int err;
@@ -741,7 +789,8 @@ static int ipv4_tun_to_nlattr(struct sk_buff *skb,
 	if (!nla)
 		return -EMSGSIZE;
 
-	err = __ipv4_tun_to_nlattr(skb, output, tun_opts, swkey_tun_opts_len);
+	err = __ip_tun_to_nlattr(skb, output, tun_opts, swkey_tun_opts_len,
+				 tun_proto);
 	if (err)
 		return err;
 
@@ -753,9 +802,10 @@ int ovs_nla_put_egress_tunnel_key(struct sk_buff *skb,
 				  const struct ip_tunnel_info *egress_tun_info,
 				  const void *egress_tun_opts)
 {
-	return __ipv4_tun_to_nlattr(skb, &egress_tun_info->key,
-				    egress_tun_opts,
-				    egress_tun_info->options_len);
+	return __ip_tun_to_nlattr(skb, &egress_tun_info->key,
+				  egress_tun_opts,
+				  egress_tun_info->options_len,
+				  ip_tunnel_info_af(egress_tun_info));
 }
 
 static int metadata_from_nlattrs(struct net *net, struct sw_flow_match *match,
@@ -806,8 +856,8 @@ static int metadata_from_nlattrs(struct net *net, struct sw_flow_match *match,
 		*attrs &= ~(1 << OVS_KEY_ATTR_SKB_MARK);
 	}
 	if (*attrs & (1 << OVS_KEY_ATTR_TUNNEL)) {
-		if (ipv4_tun_from_nlattr(a[OVS_KEY_ATTR_TUNNEL], match,
-					 is_mask, log) < 0)
+		if (ip_tun_from_nlattr(a[OVS_KEY_ATTR_TUNNEL], match,
+				       is_mask, log) < 0)
 			return -EINVAL;
 		*attrs &= ~(1 << OVS_KEY_ATTR_TUNNEL);
 	}
@@ -1194,7 +1244,7 @@ int ovs_nla_get_match(struct net *net, struct sw_flow_match *match,
 			/* The userspace does not send tunnel attributes that
 			 * are 0, but we should not wildcard them nonetheless.
 			 */
-			if (match->key->tun_key.u.ipv4.dst)
+			if (match->key->tun_proto)
 				SW_FLOW_KEY_MEMSET_FIELD(match, tun_key,
 							 0xff, true);
 
@@ -1367,14 +1417,14 @@ static int __ovs_nla_put_key(const struct sw_flow_key *swkey,
 	if (nla_put_u32(skb, OVS_KEY_ATTR_PRIORITY, output->phy.priority))
 		goto nla_put_failure;
 
-	if ((swkey->tun_key.u.ipv4.dst || is_mask)) {
+	if ((swkey->tun_proto || is_mask)) {
 		const void *opts = NULL;
 
 		if (output->tun_key.tun_flags & TUNNEL_OPTIONS_PRESENT)
 			opts = TUN_METADATA_OPTS(output, swkey->tun_opts_len);
 
-		if (ipv4_tun_to_nlattr(skb, &output->tun_key, opts,
-				       swkey->tun_opts_len))
+		if (ip_tun_to_nlattr(skb, &output->tun_key, opts,
+				     swkey->tun_opts_len, swkey->tun_proto))
 			goto nla_put_failure;
 	}
 
@@ -1877,7 +1927,7 @@ static int validate_and_copy_set_tun(const struct nlattr *attr,
 	int err = 0, start, opts_type;
 
 	ovs_match_init(&match, &key, NULL);
-	opts_type = ipv4_tun_from_nlattr(nla_data(attr), &match, false, log);
+	opts_type = ip_tun_from_nlattr(nla_data(attr), &match, false, log);
 	if (opts_type < 0)
 		return opts_type;
 
@@ -1913,6 +1963,8 @@ static int validate_and_copy_set_tun(const struct nlattr *attr,
 
 	tun_info = &tun_dst->u.tun_info;
 	tun_info->mode = IP_TUNNEL_INFO_TX;
+	if (key.tun_proto == AF_INET6)
+		tun_info->mode |= IP_TUNNEL_INFO_IPV6;
 	tun_info->key = key.tun_key;
 
 	/* We need to store the options in the action itself since
@@ -2374,10 +2426,11 @@ static int set_action_to_attr(const struct nlattr *a, struct sk_buff *skb)
 		if (!start)
 			return -EMSGSIZE;
 
-		err = ipv4_tun_to_nlattr(skb, &tun_info->key,
-					 tun_info->options_len ?
+		err = ip_tun_to_nlattr(skb, &tun_info->key,
+				       tun_info->options_len ?
 					     ip_tunnel_info_opts(tun_info) : NULL,
-					 tun_info->options_len);
+				       tun_info->options_len,
+				       ip_tunnel_info_af(tun_info));
 		if (err)
 			return err;
 		nla_nest_end(skb, start);
diff --git a/net/openvswitch/flow_table.c b/net/openvswitch/flow_table.c
index f2ea83ba4763..95dbcedf0bd4 100644
--- a/net/openvswitch/flow_table.c
+++ b/net/openvswitch/flow_table.c
@@ -427,7 +427,7 @@ static u32 flow_hash(const struct sw_flow_key *key,
 
 static int flow_key_start(const struct sw_flow_key *key)
 {
-	if (key->tun_key.u.ipv4.dst)
+	if (key->tun_proto)
 		return 0;
 	else
 		return rounddown(offsetof(struct sw_flow_key, phy),