netdev-dpdk: fix mbuf leaks
[cascardo/ovs.git] / datapath / flow_netlink.c
index a3f34f1..6ffcc53 100644 (file)
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2007-2014 Nicira, Inc.
+ * Copyright (c) 2007-2015 Nicira, Inc.
  *
  * This program is free software; you can redistribute it and/or
  * modify it under the terms of version 2 of the GNU General Public
@@ -18,9 +18,6 @@
 
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
-#include "flow.h"
-#include "datapath.h"
-#include "mpls.h"
 #include <linux/uaccess.h>
 #include <linux/netdevice.h>
 #include <linux/etherdevice.h>
 #include <linux/rculist.h>
 #include <net/geneve.h>
 #include <net/ip.h>
-#include <net/ip_tunnels.h>
 #include <net/ipv6.h>
 #include <net/ndisc.h>
+#include <net/mpls.h>
+#include <net/vxlan.h>
 
+#include "datapath.h"
+#include "conntrack.h"
+#include "flow.h"
 #include "flow_netlink.h"
+#include "gso.h"
+
+struct ovs_len_tbl {
+       int len;
+       const struct ovs_len_tbl *next;
+};
+
+#define OVS_ATTR_NESTED -1
+#define OVS_ATTR_VARIABLE -2
 
 static void update_range(struct sw_flow_match *match,
                         size_t offset, size_t size, bool is_mask)
@@ -112,13 +122,14 @@ static void update_range(struct sw_flow_match *match,
        } while (0)
 
 static bool match_validate(const struct sw_flow_match *match,
-                          u64 key_attrs, u64 mask_attrs)
+                          u64 key_attrs, u64 mask_attrs, bool log)
 {
        u64 key_expected = 1ULL << OVS_KEY_ATTR_ETHERNET;
        u64 mask_allowed = key_attrs;  /* At most allow all key attributes */
 
        /* The following mask attributes allowed only if they
-        * pass the validation tests. */
+        * pass the validation tests.
+        */
        mask_allowed &= ~((1ULL << OVS_KEY_ATTR_IPV4)
                        | (1ULL << OVS_KEY_ATTR_IPV6)
                        | (1ULL << OVS_KEY_ATTR_TCP)
@@ -144,7 +155,6 @@ static bool match_validate(const struct sw_flow_match *match,
                        mask_allowed |= 1ULL << OVS_KEY_ATTR_ARP;
        }
 
-
        if (eth_p_mpls(match->key->eth.type)) {
                key_expected |= 1ULL << OVS_KEY_ATTR_MPLS;
                if (match->mask && (match->mask->key.eth.type == htons(0xffff)))
@@ -231,15 +241,17 @@ static bool match_validate(const struct sw_flow_match *match,
 
        if ((key_attrs & key_expected) != key_expected) {
                /* Key attributes check failed. */
-               OVS_NLERR("Missing expected key attributes (key_attrs=%llx, expected=%llx).\n",
-                               (unsigned long long)key_attrs, (unsigned long long)key_expected);
+               OVS_NLERR(log, "Missing key (keys=%llx, expected=%llx)",
+                         (unsigned long long)key_attrs,
+                         (unsigned long long)key_expected);
                return false;
        }
 
        if ((mask_attrs & mask_allowed) != mask_attrs) {
                /* Mask attributes check failed. */
-               OVS_NLERR("Contain more than allowed mask fields (mask_attrs=%llx, mask_allowed=%llx).\n",
-                               (unsigned long long)mask_attrs, (unsigned long long)mask_allowed);
+               OVS_NLERR(log, "Unexpected mask (mask=%llx, allowed=%llx)",
+                         (unsigned long long)mask_attrs,
+                         (unsigned long long)mask_allowed);
                return false;
        }
 
@@ -249,7 +261,8 @@ static bool match_validate(const struct sw_flow_match *match,
 size_t ovs_tun_key_attr_size(void)
 {
        /* Whenever adding new OVS_TUNNEL_KEY_ FIELDS, we should consider
-        * updating this function.  */
+        * 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 */
@@ -259,6 +272,9 @@ size_t ovs_tun_key_attr_size(void)
                + nla_total_size(0)    /* OVS_TUNNEL_KEY_ATTR_CSUM */
                + nla_total_size(0)    /* OVS_TUNNEL_KEY_ATTR_OAM */
                + nla_total_size(256)  /* OVS_TUNNEL_KEY_ATTR_GENEVE_OPTS */
+               /* OVS_TUNNEL_KEY_ATTR_VXLAN_OPTS is mutually exclusive with
+                * OVS_TUNNEL_KEY_ATTR_GENEVE_OPTS and covered by it.
+                */
                + nla_total_size(2)    /* OVS_TUNNEL_KEY_ATTR_TP_SRC */
                + nla_total_size(2);   /* OVS_TUNNEL_KEY_ATTR_TP_DST */
 }
@@ -266,8 +282,9 @@ size_t ovs_tun_key_attr_size(void)
 size_t ovs_key_attr_size(void)
 {
        /* Whenever adding new OVS_KEY_ FIELDS, we should consider
-        * updating this function.  */
-       BUILD_BUG_ON(OVS_KEY_ATTR_TUNNEL_INFO != 22);
+        * updating this function.
+        */
+       BUILD_BUG_ON(OVS_KEY_ATTR_TUNNEL_INFO != 26);
 
        return    nla_total_size(4)   /* OVS_KEY_ATTR_PRIORITY */
                + nla_total_size(0)   /* OVS_KEY_ATTR_TUNNEL */
@@ -276,6 +293,10 @@ size_t ovs_key_attr_size(void)
                + nla_total_size(4)   /* OVS_KEY_ATTR_SKB_MARK */
                + nla_total_size(4)   /* OVS_KEY_ATTR_DP_HASH */
                + nla_total_size(4)   /* OVS_KEY_ATTR_RECIRC_ID */
+               + nla_total_size(4)   /* OVS_KEY_ATTR_CT_STATE */
+               + nla_total_size(2)   /* OVS_KEY_ATTR_CT_ZONE */
+               + nla_total_size(4)   /* OVS_KEY_ATTR_CT_MARK */
+               + nla_total_size(16)  /* OVS_KEY_ATTR_CT_LABELS */
                + nla_total_size(12)  /* OVS_KEY_ATTR_ETHERNET */
                + nla_total_size(2)   /* OVS_KEY_ATTR_ETHERTYPE */
                + nla_total_size(4)   /* OVS_KEY_ATTR_VLAN */
@@ -286,31 +307,63 @@ size_t ovs_key_attr_size(void)
                + nla_total_size(28); /* OVS_KEY_ATTR_ND */
 }
 
+static const struct ovs_len_tbl ovs_vxlan_ext_key_lens[OVS_VXLAN_EXT_MAX + 1] = {
+       [OVS_VXLAN_EXT_GBP]         = { .len = sizeof(u32) },
+};
+
+static const struct ovs_len_tbl ovs_tunnel_key_lens[OVS_TUNNEL_KEY_ATTR_MAX + 1] = {
+       [OVS_TUNNEL_KEY_ATTR_ID]            = { .len = sizeof(u64) },
+       [OVS_TUNNEL_KEY_ATTR_IPV4_SRC]      = { .len = sizeof(u32) },
+       [OVS_TUNNEL_KEY_ATTR_IPV4_DST]      = { .len = sizeof(u32) },
+       [OVS_TUNNEL_KEY_ATTR_TOS]           = { .len = 1 },
+       [OVS_TUNNEL_KEY_ATTR_TTL]           = { .len = 1 },
+       [OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT] = { .len = 0 },
+       [OVS_TUNNEL_KEY_ATTR_CSUM]          = { .len = 0 },
+       [OVS_TUNNEL_KEY_ATTR_TP_SRC]        = { .len = sizeof(u16) },
+       [OVS_TUNNEL_KEY_ATTR_TP_DST]        = { .len = sizeof(u16) },
+       [OVS_TUNNEL_KEY_ATTR_OAM]           = { .len = 0 },
+       [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 },
+};
+
 /* The size of the argument for each %OVS_KEY_ATTR_* Netlink attribute.  */
-static const int ovs_key_lens[OVS_KEY_ATTR_MAX + 1] = {
-       [OVS_KEY_ATTR_ENCAP] = -1,
-       [OVS_KEY_ATTR_PRIORITY] = sizeof(u32),
-       [OVS_KEY_ATTR_IN_PORT] = sizeof(u32),
-       [OVS_KEY_ATTR_SKB_MARK] = sizeof(u32),
-       [OVS_KEY_ATTR_ETHERNET] = sizeof(struct ovs_key_ethernet),
-       [OVS_KEY_ATTR_VLAN] = sizeof(__be16),
-       [OVS_KEY_ATTR_ETHERTYPE] = sizeof(__be16),
-       [OVS_KEY_ATTR_IPV4] = sizeof(struct ovs_key_ipv4),
-       [OVS_KEY_ATTR_IPV6] = sizeof(struct ovs_key_ipv6),
-       [OVS_KEY_ATTR_TCP] = sizeof(struct ovs_key_tcp),
-       [OVS_KEY_ATTR_TCP_FLAGS] = sizeof(__be16),
-       [OVS_KEY_ATTR_UDP] = sizeof(struct ovs_key_udp),
-       [OVS_KEY_ATTR_SCTP] = sizeof(struct ovs_key_sctp),
-       [OVS_KEY_ATTR_ICMP] = sizeof(struct ovs_key_icmp),
-       [OVS_KEY_ATTR_ICMPV6] = sizeof(struct ovs_key_icmpv6),
-       [OVS_KEY_ATTR_ARP] = sizeof(struct ovs_key_arp),
-       [OVS_KEY_ATTR_ND] = sizeof(struct ovs_key_nd),
-       [OVS_KEY_ATTR_DP_HASH] = sizeof(u32),
-       [OVS_KEY_ATTR_RECIRC_ID] = sizeof(u32),
-       [OVS_KEY_ATTR_TUNNEL] = -1,
-       [OVS_KEY_ATTR_MPLS] = sizeof(struct ovs_key_mpls),
+static const struct ovs_len_tbl ovs_key_lens[OVS_KEY_ATTR_MAX + 1] = {
+       [OVS_KEY_ATTR_ENCAP]     = { .len = OVS_ATTR_NESTED },
+       [OVS_KEY_ATTR_PRIORITY]  = { .len = sizeof(u32) },
+       [OVS_KEY_ATTR_IN_PORT]   = { .len = sizeof(u32) },
+       [OVS_KEY_ATTR_SKB_MARK]  = { .len = sizeof(u32) },
+       [OVS_KEY_ATTR_ETHERNET]  = { .len = sizeof(struct ovs_key_ethernet) },
+       [OVS_KEY_ATTR_VLAN]      = { .len = sizeof(__be16) },
+       [OVS_KEY_ATTR_ETHERTYPE] = { .len = sizeof(__be16) },
+       [OVS_KEY_ATTR_IPV4]      = { .len = sizeof(struct ovs_key_ipv4) },
+       [OVS_KEY_ATTR_IPV6]      = { .len = sizeof(struct ovs_key_ipv6) },
+       [OVS_KEY_ATTR_TCP]       = { .len = sizeof(struct ovs_key_tcp) },
+       [OVS_KEY_ATTR_TCP_FLAGS] = { .len = sizeof(__be16) },
+       [OVS_KEY_ATTR_UDP]       = { .len = sizeof(struct ovs_key_udp) },
+       [OVS_KEY_ATTR_SCTP]      = { .len = sizeof(struct ovs_key_sctp) },
+       [OVS_KEY_ATTR_ICMP]      = { .len = sizeof(struct ovs_key_icmp) },
+       [OVS_KEY_ATTR_ICMPV6]    = { .len = sizeof(struct ovs_key_icmpv6) },
+       [OVS_KEY_ATTR_ARP]       = { .len = sizeof(struct ovs_key_arp) },
+       [OVS_KEY_ATTR_ND]        = { .len = sizeof(struct ovs_key_nd) },
+       [OVS_KEY_ATTR_RECIRC_ID] = { .len = sizeof(u32) },
+       [OVS_KEY_ATTR_DP_HASH]   = { .len = sizeof(u32) },
+       [OVS_KEY_ATTR_TUNNEL]    = { .len = OVS_ATTR_NESTED,
+                                    .next = ovs_tunnel_key_lens, },
+       [OVS_KEY_ATTR_MPLS]      = { .len = sizeof(struct ovs_key_mpls) },
+       [OVS_KEY_ATTR_CT_STATE]  = { .len = sizeof(u32) },
+       [OVS_KEY_ATTR_CT_ZONE]   = { .len = sizeof(u16) },
+       [OVS_KEY_ATTR_CT_MARK]   = { .len = sizeof(u32) },
+       [OVS_KEY_ATTR_CT_LABELS] = { .len = sizeof(struct ovs_key_ct_labels) },
 };
 
+static bool check_attr_len(unsigned int attr_len, unsigned int expected_len)
+{
+       return expected_len == attr_len ||
+              expected_len == OVS_ATTR_NESTED ||
+              expected_len == OVS_ATTR_VARIABLE;
+}
+
 static bool is_all_zero(const u8 *fp, size_t size)
 {
        int i;
@@ -327,7 +380,7 @@ static bool is_all_zero(const u8 *fp, size_t size)
 
 static int __parse_flow_nlattrs(const struct nlattr *attr,
                                const struct nlattr *a[],
-                               u64 *attrsp, bool nz)
+                               u64 *attrsp, bool log, bool nz)
 {
        const struct nlattr *nla;
        u64 attrs;
@@ -339,21 +392,20 @@ static int __parse_flow_nlattrs(const struct nlattr *attr,
                int expected_len;
 
                if (type > OVS_KEY_ATTR_MAX) {
-                       OVS_NLERR("Unknown key attribute (type=%d, max=%d).\n",
+                       OVS_NLERR(log, "Key type %d is out of range max %d",
                                  type, OVS_KEY_ATTR_MAX);
                        return -EINVAL;
                }
 
                if (attrs & (1ULL << type)) {
-                       OVS_NLERR("Duplicate key attribute (type %d).\n", type);
+                       OVS_NLERR(log, "Duplicate key (type %d).", type);
                        return -EINVAL;
                }
 
-               expected_len = ovs_key_lens[type];
-               if (nla_len(nla) != expected_len && expected_len != -1) {
-                       OVS_NLERR("Key attribute has unexpected length (type=%d"
-                                 ", length=%d, expected=%d).\n", type,
-                                 nla_len(nla), expected_len);
+               expected_len = ovs_key_lens[type].len;
+               if (!check_attr_len(nla_len(nla), expected_len)) {
+                       OVS_NLERR(log, "Key %d has unexpected len %d expected %d",
+                                 type, nla_len(nla), expected_len);
                        return -EINVAL;
                }
 
@@ -363,7 +415,7 @@ static int __parse_flow_nlattrs(const struct nlattr *attr,
                }
        }
        if (rem) {
-               OVS_NLERR("Message has %d unknown bytes.\n", rem);
+               OVS_NLERR(log, "Message has %d unknown bytes.", rem);
                return -EINVAL;
        }
 
@@ -372,52 +424,150 @@ static int __parse_flow_nlattrs(const struct nlattr *attr,
 }
 
 static int parse_flow_mask_nlattrs(const struct nlattr *attr,
-                                  const struct nlattr *a[], u64 *attrsp)
+                                  const struct nlattr *a[], u64 *attrsp,
+                                  bool log)
 {
-       return __parse_flow_nlattrs(attr, a, attrsp, true);
+       return __parse_flow_nlattrs(attr, a, attrsp, log, true);
 }
 
 static int parse_flow_nlattrs(const struct nlattr *attr,
-                             const struct nlattr *a[], u64 *attrsp)
+                             const struct nlattr *a[], u64 *attrsp,
+                             bool log)
 {
-       return __parse_flow_nlattrs(attr, a, attrsp, false);
+       return __parse_flow_nlattrs(attr, a, attrsp, log, false);
+}
+
+static int genev_tun_opt_from_nlattr(const struct nlattr *a,
+                                    struct sw_flow_match *match, bool is_mask,
+                                    bool log)
+{
+       unsigned long opt_key_offset;
+
+       if (nla_len(a) > sizeof(match->key->tun_opts)) {
+               OVS_NLERR(log, "Geneve option length err (len %d, max %zu).",
+                         nla_len(a), sizeof(match->key->tun_opts));
+               return -EINVAL;
+       }
+
+       if (nla_len(a) % 4 != 0) {
+               OVS_NLERR(log, "Geneve opt len %d is not a multiple of 4.",
+                         nla_len(a));
+               return -EINVAL;
+       }
+
+       /* We need to record the length of the options passed
+        * down, otherwise packets with the same format but
+        * additional options will be silently matched.
+        */
+       if (!is_mask) {
+               SW_FLOW_KEY_PUT(match, tun_opts_len, nla_len(a),
+                               false);
+       } else {
+               /* This is somewhat unusual because it looks at
+                * both the key and mask while parsing the
+                * attributes (and by extension assumes the key
+                * is parsed first). Normally, we would verify
+                * that each is the correct length and that the
+                * attributes line up in the validate function.
+                * However, that is difficult because this is
+                * variable length and we won't have the
+                * information later.
+                */
+               if (match->key->tun_opts_len != nla_len(a)) {
+                       OVS_NLERR(log, "Geneve option len %d != mask len %d",
+                                 match->key->tun_opts_len, nla_len(a));
+                       return -EINVAL;
+               }
+
+               SW_FLOW_KEY_PUT(match, tun_opts_len, 0xff, true);
+       }
+
+       opt_key_offset = TUN_METADATA_OFFSET(nla_len(a));
+       SW_FLOW_KEY_MEMCPY_OFFSET(match, opt_key_offset, nla_data(a),
+                                 nla_len(a), is_mask);
+       return 0;
+}
+
+static int vxlan_tun_opt_from_nlattr(const struct nlattr *attr,
+                                    struct sw_flow_match *match, bool is_mask,
+                                    bool log)
+{
+       struct nlattr *a;
+       int rem;
+       unsigned long opt_key_offset;
+       struct vxlan_metadata opts;
+
+       BUILD_BUG_ON(sizeof(opts) > sizeof(match->key->tun_opts));
+
+       memset(&opts, 0, sizeof(opts));
+       nla_for_each_nested(a, attr, rem) {
+               int type = nla_type(a);
+
+               if (type > OVS_VXLAN_EXT_MAX) {
+                       OVS_NLERR(log, "VXLAN extension %d out of range max %d",
+                                 type, OVS_VXLAN_EXT_MAX);
+                       return -EINVAL;
+               }
+
+               if (!check_attr_len(nla_len(a),
+                                   ovs_vxlan_ext_key_lens[type].len)) {
+                       OVS_NLERR(log, "VXLAN extension %d has unexpected len %d expected %d",
+                                 type, nla_len(a),
+                                 ovs_vxlan_ext_key_lens[type].len);
+                       return -EINVAL;
+               }
+
+               switch (type) {
+               case OVS_VXLAN_EXT_GBP:
+                       opts.gbp = nla_get_u32(a);
+                       break;
+               default:
+                       OVS_NLERR(log, "Unknown VXLAN extension attribute %d",
+                                 type);
+                       return -EINVAL;
+               }
+       }
+       if (rem) {
+               OVS_NLERR(log, "VXLAN extension message has %d unknown bytes.",
+                         rem);
+               return -EINVAL;
+       }
+
+       if (!is_mask)
+               SW_FLOW_KEY_PUT(match, tun_opts_len, sizeof(opts), false);
+       else
+               SW_FLOW_KEY_PUT(match, tun_opts_len, 0xff, true);
+
+       opt_key_offset = TUN_METADATA_OFFSET(sizeof(opts));
+       SW_FLOW_KEY_MEMCPY_OFFSET(match, opt_key_offset, &opts, sizeof(opts),
+                                 is_mask);
+       return 0;
 }
 
 static int ipv4_tun_from_nlattr(const struct nlattr *attr,
-                               struct sw_flow_match *match, bool is_mask)
+                               struct sw_flow_match *match, bool is_mask,
+                               bool log)
 {
        struct nlattr *a;
        int rem;
        bool ttl = false;
        __be16 tun_flags = 0;
+       int opts_type = 0;
 
        nla_for_each_nested(a, attr, rem) {
                int type = nla_type(a);
-               static const u32 ovs_tunnel_key_lens[OVS_TUNNEL_KEY_ATTR_MAX + 1] = {
-                       [OVS_TUNNEL_KEY_ATTR_ID] = sizeof(u64),
-                       [OVS_TUNNEL_KEY_ATTR_IPV4_SRC] = sizeof(u32),
-                       [OVS_TUNNEL_KEY_ATTR_IPV4_DST] = sizeof(u32),
-                       [OVS_TUNNEL_KEY_ATTR_TOS] = 1,
-                       [OVS_TUNNEL_KEY_ATTR_TTL] = 1,
-                       [OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT] = 0,
-                       [OVS_TUNNEL_KEY_ATTR_CSUM] = 0,
-                       [OVS_TUNNEL_KEY_ATTR_TP_SRC] = sizeof(u16),
-                       [OVS_TUNNEL_KEY_ATTR_TP_DST] = sizeof(u16),
-                       [OVS_TUNNEL_KEY_ATTR_OAM] = 0,
-                       [OVS_TUNNEL_KEY_ATTR_GENEVE_OPTS] = -1,
-               };
+               int err;
 
                if (type > OVS_TUNNEL_KEY_ATTR_MAX) {
-                       OVS_NLERR("Unknown IPv4 tunnel attribute (type=%d, max=%d).\n",
-                       type, OVS_TUNNEL_KEY_ATTR_MAX);
+                       OVS_NLERR(log, "Tunnel attr %d out of range max %d",
+                                 type, OVS_TUNNEL_KEY_ATTR_MAX);
                        return -EINVAL;
                }
 
-               if (ovs_tunnel_key_lens[type] != nla_len(a) &&
-                   ovs_tunnel_key_lens[type] != -1) {
-                       OVS_NLERR("IPv4 tunnel attribute type has unexpected "
-                                 " length (type=%d, length=%d, expected=%d).\n",
-                                 type, nla_len(a), ovs_tunnel_key_lens[type]);
+               if (!check_attr_len(nla_len(a),
+                                   ovs_tunnel_key_lens[type].len)) {
+                       OVS_NLERR(log, "Tunnel attr %d has unexpected len %d expected %d",
+                                 type, nla_len(a), ovs_tunnel_key_lens[type].len);
                        return -EINVAL;
                }
 
@@ -428,19 +578,19 @@ static int ipv4_tun_from_nlattr(const struct nlattr *attr,
                        tun_flags |= TUNNEL_KEY;
                        break;
                case OVS_TUNNEL_KEY_ATTR_IPV4_SRC:
-                       SW_FLOW_KEY_PUT(match, tun_key.ipv4_src,
-                                       nla_get_be32(a), is_mask);
+                       SW_FLOW_KEY_PUT(match, tun_key.u.ipv4.src,
+                                       nla_get_in_addr(a), is_mask);
                        break;
                case OVS_TUNNEL_KEY_ATTR_IPV4_DST:
-                       SW_FLOW_KEY_PUT(match, tun_key.ipv4_dst,
-                                       nla_get_be32(a), is_mask);
+                       SW_FLOW_KEY_PUT(match, tun_key.u.ipv4.dst,
+                                       nla_get_in_addr(a), is_mask);
                        break;
                case OVS_TUNNEL_KEY_ATTR_TOS:
-                       SW_FLOW_KEY_PUT(match, tun_key.ipv4_tos,
+                       SW_FLOW_KEY_PUT(match, tun_key.tos,
                                        nla_get_u8(a), is_mask);
                        break;
                case OVS_TUNNEL_KEY_ATTR_TTL:
-                       SW_FLOW_KEY_PUT(match, tun_key.ipv4_ttl,
+                       SW_FLOW_KEY_PUT(match, tun_key.ttl,
                                        nla_get_u8(a), is_mask);
                        ttl = true;
                        break;
@@ -462,58 +612,34 @@ static int ipv4_tun_from_nlattr(const struct nlattr *attr,
                        tun_flags |= TUNNEL_OAM;
                        break;
                case OVS_TUNNEL_KEY_ATTR_GENEVE_OPTS:
-                       tun_flags |= TUNNEL_OPTIONS_PRESENT;
-                       if (nla_len(a) > sizeof(match->key->tun_opts)) {
-                               OVS_NLERR("Geneve option length exceeds "
-                                         "maximum size (len %d, max %zu).\n",
-                                         nla_len(a),
-                                         sizeof(match->key->tun_opts));
+                       if (opts_type) {
+                               OVS_NLERR(log, "Multiple metadata blocks provided");
                                return -EINVAL;
                        }
 
-                       if (nla_len(a) % 4 != 0) {
-                               OVS_NLERR("Geneve option length is not "
-                                         "a multiple of 4 (len %d).\n",
-                                         nla_len(a));
+                       err = genev_tun_opt_from_nlattr(a, match, is_mask, log);
+                       if (err)
+                               return err;
+
+                       tun_flags |= TUNNEL_GENEVE_OPT;
+                       opts_type = type;
+                       break;
+               case OVS_TUNNEL_KEY_ATTR_VXLAN_OPTS:
+                       if (opts_type) {
+                               OVS_NLERR(log, "Multiple metadata blocks provided");
                                return -EINVAL;
                        }
 
-                       /* We need to record the length of the options passed
-                        * down, otherwise packets with the same format but
-                        * additional options will be silently matched.
-                        */
-                       if (!is_mask) {
-                               SW_FLOW_KEY_PUT(match, tun_opts_len, nla_len(a),
-                                               false);
-                       } else {
-                               /* This is somewhat unusual because it looks at
-                                * both the key and mask while parsing the
-                                * attributes (and by extension assumes the key
-                                * is parsed first). Normally, we would verify
-                                * that each is the correct length and that the
-                                * attributes line up in the validate function.
-                                * However, that is difficult because this is
-                                * variable length and we won't have the
-                                * information later.
-                                */
-                               if (match->key->tun_opts_len != nla_len(a)) {
-                                       OVS_NLERR("Geneve option key length (%d)"
-                                          " is different from mask length (%d).",
-                                          match->key->tun_opts_len, nla_len(a));
-                                       return -EINVAL;
-                               }
-
-                               SW_FLOW_KEY_PUT(match, tun_opts_len, 0xff,
-                                               true);
-                       }
+                       err = vxlan_tun_opt_from_nlattr(a, match, is_mask, log);
+                       if (err)
+                               return err;
 
-                       SW_FLOW_KEY_MEMCPY_OFFSET(match,
-                               (unsigned long)GENEVE_OPTS((struct sw_flow_key *)0,
-                                                          nla_len(a)),
-                               nla_data(a), nla_len(a), is_mask);
+                       tun_flags |= TUNNEL_VXLAN_OPT;
+                       opts_type = type;
                        break;
                default:
-                       OVS_NLERR("Unknown IPv4 tunnel attribute (%d).\n", type);
+                       OVS_NLERR(log, "Unknown IPv4 tunnel attribute %d",
+                                 type);
                        return -EINVAL;
                }
        }
@@ -521,72 +647,94 @@ static int ipv4_tun_from_nlattr(const struct nlattr *attr,
        SW_FLOW_KEY_PUT(match, tun_key.tun_flags, tun_flags, is_mask);
 
        if (rem > 0) {
-               OVS_NLERR("IPv4 tunnel attribute has %d unknown bytes.\n", rem);
+               OVS_NLERR(log, "IPv4 tunnel attribute has %d unknown bytes.",
+                         rem);
                return -EINVAL;
        }
 
        if (!is_mask) {
-               if (!match->key->tun_key.ipv4_dst) {
-                       OVS_NLERR("IPv4 tunnel destination address is zero.\n");
+               if (!match->key->tun_key.u.ipv4.dst) {
+                       OVS_NLERR(log, "IPv4 tunnel dst address is zero");
                        return -EINVAL;
                }
 
                if (!ttl) {
-                       OVS_NLERR("IPv4 tunnel TTL not specified.\n");
+                       OVS_NLERR(log, "IPv4 tunnel TTL not specified.");
                        return -EINVAL;
                }
        }
 
+       return opts_type;
+}
+
+static int vxlan_opt_to_nlattr(struct sk_buff *skb,
+                              const void *tun_opts, int swkey_tun_opts_len)
+{
+       const struct vxlan_metadata *opts = tun_opts;
+       struct nlattr *nla;
+
+       nla = nla_nest_start(skb, OVS_TUNNEL_KEY_ATTR_VXLAN_OPTS);
+       if (!nla)
+               return -EMSGSIZE;
+
+       if (nla_put_u32(skb, OVS_VXLAN_EXT_GBP, opts->gbp) < 0)
+               return -EMSGSIZE;
+
+       nla_nest_end(skb, nla);
        return 0;
 }
 
 static int __ipv4_tun_to_nlattr(struct sk_buff *skb,
-                               const struct ovs_key_ipv4_tunnel *output,
-                               const struct geneve_opt *tun_opts,
-                               int swkey_tun_opts_len)
+                               const struct ip_tunnel_key *output,
+                               const void *tun_opts, int swkey_tun_opts_len)
 {
        if (output->tun_flags & TUNNEL_KEY &&
            nla_put_be64(skb, OVS_TUNNEL_KEY_ATTR_ID, output->tun_id))
                return -EMSGSIZE;
-       if (output->ipv4_src &&
-               nla_put_be32(skb, OVS_TUNNEL_KEY_ATTR_IPV4_SRC, output->ipv4_src))
+       if (output->u.ipv4.src &&
+           nla_put_in_addr(skb, OVS_TUNNEL_KEY_ATTR_IPV4_SRC,
+                           output->u.ipv4.src))
                return -EMSGSIZE;
-       if (output->ipv4_dst &&
-               nla_put_be32(skb, OVS_TUNNEL_KEY_ATTR_IPV4_DST, output->ipv4_dst))
+       if (output->u.ipv4.dst &&
+           nla_put_in_addr(skb, OVS_TUNNEL_KEY_ATTR_IPV4_DST,
+                           output->u.ipv4.dst))
                return -EMSGSIZE;
-       if (output->ipv4_tos &&
-               nla_put_u8(skb, OVS_TUNNEL_KEY_ATTR_TOS, output->ipv4_tos))
+       if (output->tos &&
+           nla_put_u8(skb, OVS_TUNNEL_KEY_ATTR_TOS, output->tos))
                return -EMSGSIZE;
-       if (nla_put_u8(skb, OVS_TUNNEL_KEY_ATTR_TTL, output->ipv4_ttl))
+       if (nla_put_u8(skb, OVS_TUNNEL_KEY_ATTR_TTL, output->ttl))
                return -EMSGSIZE;
        if ((output->tun_flags & TUNNEL_DONT_FRAGMENT) &&
-               nla_put_flag(skb, OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT))
+           nla_put_flag(skb, OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT))
                return -EMSGSIZE;
        if ((output->tun_flags & TUNNEL_CSUM) &&
-               nla_put_flag(skb, OVS_TUNNEL_KEY_ATTR_CSUM))
+           nla_put_flag(skb, OVS_TUNNEL_KEY_ATTR_CSUM))
                return -EMSGSIZE;
        if (output->tp_src &&
-               nla_put_be16(skb, OVS_TUNNEL_KEY_ATTR_TP_SRC, output->tp_src))
+           nla_put_be16(skb, OVS_TUNNEL_KEY_ATTR_TP_SRC, output->tp_src))
                return -EMSGSIZE;
        if (output->tp_dst &&
-               nla_put_be16(skb, OVS_TUNNEL_KEY_ATTR_TP_DST, output->tp_dst))
+           nla_put_be16(skb, OVS_TUNNEL_KEY_ATTR_TP_DST, output->tp_dst))
                return -EMSGSIZE;
        if ((output->tun_flags & TUNNEL_OAM) &&
-               nla_put_flag(skb, OVS_TUNNEL_KEY_ATTR_OAM))
-               return -EMSGSIZE;
-       if (tun_opts &&
-           nla_put(skb, OVS_TUNNEL_KEY_ATTR_GENEVE_OPTS,
-                   swkey_tun_opts_len, tun_opts))
+           nla_put_flag(skb, OVS_TUNNEL_KEY_ATTR_OAM))
                return -EMSGSIZE;
+       if (tun_opts) {
+               if (output->tun_flags & TUNNEL_GENEVE_OPT &&
+                   nla_put(skb, OVS_TUNNEL_KEY_ATTR_GENEVE_OPTS,
+                           swkey_tun_opts_len, tun_opts))
+                       return -EMSGSIZE;
+               else if (output->tun_flags & TUNNEL_VXLAN_OPT &&
+                        vxlan_opt_to_nlattr(skb, tun_opts, swkey_tun_opts_len))
+                       return -EMSGSIZE;
+       }
 
        return 0;
 }
 
-
 static int ipv4_tun_to_nlattr(struct sk_buff *skb,
-                             const struct ovs_key_ipv4_tunnel *output,
-                             const struct geneve_opt *tun_opts,
-                             int swkey_tun_opts_len)
+                             const struct ip_tunnel_key *output,
+                             const void *tun_opts, int swkey_tun_opts_len)
 {
        struct nlattr *nla;
        int err;
@@ -604,15 +752,17 @@ static int ipv4_tun_to_nlattr(struct sk_buff *skb,
 }
 
 int ovs_nla_put_egress_tunnel_key(struct sk_buff *skb,
-                                 const struct ovs_tunnel_info *egress_tun_info)
+                                 const struct ip_tunnel_info *egress_tun_info,
+                                 const void *egress_tun_opts)
 {
-       return __ipv4_tun_to_nlattr(skb, &egress_tun_info->tunnel,
-                                   egress_tun_info->options,
+       return __ipv4_tun_to_nlattr(skb, &egress_tun_info->key,
+                                   egress_tun_opts,
                                    egress_tun_info->options_len);
 }
 
-static int metadata_from_nlattrs(struct sw_flow_match *match,  u64 *attrs,
-                                const struct nlattr **a, bool is_mask)
+static int metadata_from_nlattrs(struct net *net, struct sw_flow_match *match,
+                                 u64 *attrs, const struct nlattr **a,
+                                bool is_mask, bool log)
 {
        if (*attrs & (1ULL << OVS_KEY_ATTR_DP_HASH)) {
                u32 hash_val = nla_get_u32(a[OVS_KEY_ATTR_DP_HASH]);
@@ -640,7 +790,7 @@ static int metadata_from_nlattrs(struct sw_flow_match *match,  u64 *attrs,
                if (is_mask) {
                        in_port = 0xffffffff; /* Always exact match in_port. */
                } else if (in_port >= DP_MAX_PORTS) {
-                       OVS_NLERR("Input port (%d) exceeds maximum allowable (%d).\n",
+                       OVS_NLERR(log, "Port %d exceeds max allowable %d",
                                  in_port, DP_MAX_PORTS);
                        return -EINVAL;
                }
@@ -659,19 +809,57 @@ static int metadata_from_nlattrs(struct sw_flow_match *match,  u64 *attrs,
        }
        if (*attrs & (1ULL << OVS_KEY_ATTR_TUNNEL)) {
                if (ipv4_tun_from_nlattr(a[OVS_KEY_ATTR_TUNNEL], match,
-                                        is_mask))
+                                        is_mask, log) < 0)
                        return -EINVAL;
                *attrs &= ~(1ULL << OVS_KEY_ATTR_TUNNEL);
        }
+
+       if (*attrs & (1 << OVS_KEY_ATTR_CT_STATE) &&
+           ovs_ct_verify(net, OVS_KEY_ATTR_CT_STATE)) {
+               u32 ct_state = nla_get_u32(a[OVS_KEY_ATTR_CT_STATE]);
+
+               if (ct_state & ~CT_SUPPORTED_MASK) {
+                       OVS_NLERR(log, "ct_state flags %08x unsupported",
+                                 ct_state);
+                       return -EINVAL;
+               }
+
+               SW_FLOW_KEY_PUT(match, ct.state, ct_state, is_mask);
+               *attrs &= ~(1ULL << OVS_KEY_ATTR_CT_STATE);
+       }
+       if (*attrs & (1 << OVS_KEY_ATTR_CT_ZONE) &&
+           ovs_ct_verify(net, OVS_KEY_ATTR_CT_ZONE)) {
+               u16 ct_zone = nla_get_u16(a[OVS_KEY_ATTR_CT_ZONE]);
+
+               SW_FLOW_KEY_PUT(match, ct.zone, ct_zone, is_mask);
+               *attrs &= ~(1ULL << OVS_KEY_ATTR_CT_ZONE);
+       }
+       if (*attrs & (1 << OVS_KEY_ATTR_CT_MARK) &&
+           ovs_ct_verify(net, OVS_KEY_ATTR_CT_MARK)) {
+               u32 mark = nla_get_u32(a[OVS_KEY_ATTR_CT_MARK]);
+
+               SW_FLOW_KEY_PUT(match, ct.mark, mark, is_mask);
+               *attrs &= ~(1ULL << OVS_KEY_ATTR_CT_MARK);
+       }
+       if (*attrs & (1 << OVS_KEY_ATTR_CT_LABELS) &&
+           ovs_ct_verify(net, OVS_KEY_ATTR_CT_LABELS)) {
+               const struct ovs_key_ct_labels *cl;
+
+               cl = nla_data(a[OVS_KEY_ATTR_CT_LABELS]);
+               SW_FLOW_KEY_MEMCPY(match, ct.labels, cl->ct_labels,
+                                  sizeof(*cl), is_mask);
+               *attrs &= ~(1ULL << OVS_KEY_ATTR_CT_LABELS);
+       }
        return 0;
 }
 
-static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
-                               const struct nlattr **a, bool is_mask)
+static int ovs_key_from_nlattrs(struct net *net, struct sw_flow_match *match,
+                               u64 attrs, const struct nlattr **a,
+                               bool is_mask, bool log)
 {
        int err;
 
-       err = metadata_from_nlattrs(match, &attrs, a, is_mask);
+       err = metadata_from_nlattrs(net, match, &attrs, a, is_mask, log);
        if (err)
                return err;
 
@@ -692,9 +880,9 @@ static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
                tci = nla_get_be16(a[OVS_KEY_ATTR_VLAN]);
                if (!(tci & htons(VLAN_TAG_PRESENT))) {
                        if (is_mask)
-                               OVS_NLERR("VLAN TCI mask does not have exact match for VLAN_TAG_PRESENT bit.\n");
+                               OVS_NLERR(log, "VLAN TCI mask does not have exact match for VLAN_TAG_PRESENT bit.");
                        else
-                               OVS_NLERR("VLAN TCI does not have VLAN_TAG_PRESENT bit set.\n");
+                               OVS_NLERR(log, "VLAN TCI does not have VLAN_TAG_PRESENT bit set.");
 
                        return -EINVAL;
                }
@@ -710,9 +898,9 @@ static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
                if (is_mask) {
                        /* Always exact match EtherType. */
                        eth_type = htons(0xffff);
-               } else if (ntohs(eth_type) < ETH_P_802_3_MIN) {
-                       OVS_NLERR("EtherType is less than minimum (type=%x, min=%x).\n",
-                                       ntohs(eth_type), ETH_P_802_3_MIN);
+               } else if (!eth_proto_is_802_3(eth_type)) {
+                       OVS_NLERR(log, "EtherType %x is less than min %x",
+                                 ntohs(eth_type), ETH_P_802_3_MIN);
                        return -EINVAL;
                }
 
@@ -722,13 +910,13 @@ static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
                SW_FLOW_KEY_PUT(match, eth.type, htons(ETH_P_802_2), is_mask);
        }
 
-       if (attrs & (1ULL << OVS_KEY_ATTR_IPV4)) {
+       if (attrs & (1 << OVS_KEY_ATTR_IPV4)) {
                const struct ovs_key_ipv4 *ipv4_key;
 
                ipv4_key = nla_data(a[OVS_KEY_ATTR_IPV4]);
                if (!is_mask && ipv4_key->ipv4_frag > OVS_FRAG_TYPE_MAX) {
-                       OVS_NLERR("Unknown IPv4 fragment type (value=%d, max=%d).\n",
-                               ipv4_key->ipv4_frag, OVS_FRAG_TYPE_MAX);
+                       OVS_NLERR(log, "IPv4 frag type %d is out of range max %d",
+                                 ipv4_key->ipv4_frag, OVS_FRAG_TYPE_MAX);
                        return -EINVAL;
                }
                SW_FLOW_KEY_PUT(match, ip.proto,
@@ -743,7 +931,7 @@ static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
                                ipv4_key->ipv4_src, is_mask);
                SW_FLOW_KEY_PUT(match, ipv4.addr.dst,
                                ipv4_key->ipv4_dst, is_mask);
-               attrs &= ~(1ULL << OVS_KEY_ATTR_IPV4);
+               attrs &= ~(1 << OVS_KEY_ATTR_IPV4);
        }
 
        if (attrs & (1ULL << OVS_KEY_ATTR_IPV6)) {
@@ -751,15 +939,18 @@ static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
 
                ipv6_key = nla_data(a[OVS_KEY_ATTR_IPV6]);
                if (!is_mask && ipv6_key->ipv6_frag > OVS_FRAG_TYPE_MAX) {
-                       OVS_NLERR("Unknown IPv6 fragment type (value=%d, max=%d).\n",
-                               ipv6_key->ipv6_frag, OVS_FRAG_TYPE_MAX);
+                       OVS_NLERR(log, "IPv6 frag type %d is out of range max %d",
+                                 ipv6_key->ipv6_frag, OVS_FRAG_TYPE_MAX);
                        return -EINVAL;
                }
-               if (ipv6_key->ipv6_label & htonl(0xFFF00000)) {
-                       OVS_NLERR("Invalid IPv6 flow label value (value=%x, max=%x).\n",
+
+               if (!is_mask && ipv6_key->ipv6_label & htonl(0xFFF00000)) {
+                       OVS_NLERR(log,
+                                 "Invalid IPv6 flow label value (value=%x, max=%x).",
                                  ntohl(ipv6_key->ipv6_label), (1 << 20) - 1);
                        return -EINVAL;
                }
+
                SW_FLOW_KEY_PUT(match, ipv6.label,
                                ipv6_key->ipv6_label, is_mask);
                SW_FLOW_KEY_PUT(match, ip.proto,
@@ -787,7 +978,7 @@ static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
 
                arp_key = nla_data(a[OVS_KEY_ATTR_ARP]);
                if (!is_mask && (arp_key->arp_op & htons(0xff00))) {
-                       OVS_NLERR("Unknown ARP opcode (opcode=%d).\n",
+                       OVS_NLERR(log, "Unknown ARP opcode (opcode=%d).",
                                  arp_key->arp_op);
                        return -EINVAL;
                }
@@ -888,7 +1079,7 @@ static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
        }
 
        if (attrs != 0) {
-               OVS_NLERR("Unknown key attributes (%llx).\n",
+               OVS_NLERR(log, "Unknown key attributes %llx",
                          (unsigned long long)attrs);
                return -EINVAL;
        }
@@ -896,26 +1087,30 @@ static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
        return 0;
 }
 
-static void nlattr_set(struct nlattr *attr, u8 val, bool is_attr_mask_key)
+static void nlattr_set(struct nlattr *attr, u8 val,
+                      const struct ovs_len_tbl *tbl)
 {
        struct nlattr *nla;
        int rem;
 
        /* The nlattr stream should already have been validated */
        nla_for_each_nested(nla, attr, rem) {
-               /* We assume that ovs_key_lens[type] == -1 means that type is a
-                * nested attribute
-                */
-               if (is_attr_mask_key && ovs_key_lens[nla_type(nla)] == -1)
-                       nlattr_set(nla, val, false);
-               else
+               if (tbl[nla_type(nla)].len == OVS_ATTR_NESTED) {
+                       if (tbl[nla_type(nla)].next)
+                               tbl = tbl[nla_type(nla)].next;
+                       nlattr_set(nla, val, tbl);
+               } else {
                        memset(nla_data(nla), val, nla_len(nla));
+               }
+
+               if (nla_type(nla) == OVS_KEY_ATTR_CT_STATE)
+                       *(u32 *)nla_data(nla) &= CT_SUPPORTED_MASK;
        }
 }
 
 static void mask_set_nlattr(struct nlattr *attr, u8 val)
 {
-       nlattr_set(attr, val, true);
+       nlattr_set(attr, val, ovs_key_lens);
 }
 
 /**
@@ -923,16 +1118,21 @@ static void mask_set_nlattr(struct nlattr *attr, u8 val)
  * mask. In case the 'mask' is NULL, the flow is treated as exact match
  * flow. Otherwise, it is treated as a wildcarded flow, except the mask
  * does not include any don't care bit.
+ * @net: Used to determine per-namespace field support.
  * @match: receives the extracted flow match information.
  * @key: Netlink attribute holding nested %OVS_KEY_ATTR_* Netlink attribute
  * sequence. The fields should of the packet that triggered the creation
  * of this flow.
  * @mask: Optional. Netlink attribute holding nested %OVS_KEY_ATTR_* Netlink
  * attribute specifies the mask field of the wildcarded flow.
+ * @log: Boolean to allow kernel error logging.  Normally true, but when
+ * probing for feature compatibility this should be passed in as false to
+ * suppress unnecessary error logging.
  */
-int ovs_nla_get_match(struct sw_flow_match *match,
+int ovs_nla_get_match(struct net *net, struct sw_flow_match *match,
                      const struct nlattr *nla_key,
-                     const struct nlattr *nla_mask)
+                     const struct nlattr *nla_mask,
+                     bool log)
 {
        const struct nlattr *a[OVS_KEY_ATTR_MAX + 1];
        const struct nlattr *encap;
@@ -942,7 +1142,7 @@ int ovs_nla_get_match(struct sw_flow_match *match,
        bool encap_valid = false;
        int err;
 
-       err = parse_flow_nlattrs(nla_key, a, &key_attrs);
+       err = parse_flow_nlattrs(nla_key, a, &key_attrs, log);
        if (err)
                return err;
 
@@ -953,7 +1153,7 @@ int ovs_nla_get_match(struct sw_flow_match *match,
 
                if (!((key_attrs & (1ULL << OVS_KEY_ATTR_VLAN)) &&
                      (key_attrs & (1ULL << OVS_KEY_ATTR_ENCAP)))) {
-                       OVS_NLERR("Invalid Vlan frame.\n");
+                       OVS_NLERR(log, "Invalid Vlan frame.");
                        return -EINVAL;
                }
 
@@ -964,27 +1164,26 @@ int ovs_nla_get_match(struct sw_flow_match *match,
                encap_valid = true;
 
                if (tci & htons(VLAN_TAG_PRESENT)) {
-                       err = parse_flow_nlattrs(encap, a, &key_attrs);
+                       err = parse_flow_nlattrs(encap, a, &key_attrs, log);
                        if (err)
                                return err;
                } else if (!tci) {
                        /* Corner case for truncated 802.1Q header. */
                        if (nla_len(encap)) {
-                               OVS_NLERR("Truncated 802.1Q header has non-zero encap attribute.\n");
+                               OVS_NLERR(log, "Truncated 802.1Q header has non-zero encap attribute.");
                                return -EINVAL;
                        }
                } else {
-                       OVS_NLERR("Encap attribute is set for a non-VLAN frame.\n");
+                       OVS_NLERR(log, "Encap attr is set for non-VLAN frame");
                        return  -EINVAL;
                }
        }
 
-       err = ovs_key_from_nlattrs(match, key_attrs, a, false);
+       err = ovs_key_from_nlattrs(net, match, key_attrs, a, false, log);
        if (err)
                return err;
 
        if (match->mask) {
-
                if (!nla_mask) {
                        /* Create an exact match mask. We need to set to 0xff
                         * all the 'match->mask' fields that have been touched
@@ -992,9 +1191,9 @@ int ovs_nla_get_match(struct sw_flow_match *match,
                         * 'match->mask', because padding bytes and fields not
                         * specified in 'match->key' should be left to 0.
                         * Instead, we use a stream of netlink attributes,
-                        * copied from 'key' and set to 0xff: ovs_key_from_nlattrs()
-                        * will take care of filling 'match->mask'
-                        * appropriately.
+                        * copied from 'key' and set to 0xff.
+                        * ovs_key_from_nlattrs() will take care of filling
+                        * 'match->mask' appropriately.
                         */
                        newmask = kmemdup(nla_key,
                                          nla_total_size(nla_len(nla_key)),
@@ -1007,14 +1206,14 @@ int ovs_nla_get_match(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.ipv4_dst)
+                       if (match->key->tun_key.u.ipv4.dst)
                                SW_FLOW_KEY_MEMSET_FIELD(match, tun_key,
                                                         0xff, true);
 
                        nla_mask = newmask;
                }
 
-               err = parse_flow_mask_nlattrs(nla_mask, a, &mask_attrs);
+               err = parse_flow_mask_nlattrs(nla_mask, a, &mask_attrs, log);
                if (err)
                        goto free_newmask;
 
@@ -1026,7 +1225,7 @@ int ovs_nla_get_match(struct sw_flow_match *match,
                        __be16 tci = 0;
 
                        if (!encap_valid) {
-                               OVS_NLERR("Encap mask attribute is set for non-VLAN frame.\n");
+                               OVS_NLERR(log, "Encap mask attribute is set for non-VLAN frame.");
                                err = -EINVAL;
                                goto free_newmask;
                        }
@@ -1038,12 +1237,13 @@ int ovs_nla_get_match(struct sw_flow_match *match,
                        if (eth_type == htons(0xffff)) {
                                mask_attrs &= ~(1ULL << OVS_KEY_ATTR_ETHERTYPE);
                                encap = a[OVS_KEY_ATTR_ENCAP];
-                               err = parse_flow_mask_nlattrs(encap, a, &mask_attrs);
+                               err = parse_flow_mask_nlattrs(encap, a,
+                                                             &mask_attrs, log);
                                if (err)
                                        goto free_newmask;
                        } else {
-                               OVS_NLERR("VLAN frames must have an exact match on the TPID (mask=%x).\n",
-                                               ntohs(eth_type));
+                               OVS_NLERR(log, "VLAN frames must have an exact match on the TPID (mask=%x).",
+                                         ntohs(eth_type));
                                err = -EINVAL;
                                goto free_newmask;
                        }
@@ -1052,18 +1252,20 @@ int ovs_nla_get_match(struct sw_flow_match *match,
                                tci = nla_get_be16(a[OVS_KEY_ATTR_VLAN]);
 
                        if (!(tci & htons(VLAN_TAG_PRESENT))) {
-                               OVS_NLERR("VLAN tag present bit must have an exact match (tci_mask=%x).\n", ntohs(tci));
+                               OVS_NLERR(log, "VLAN tag present bit must have an exact match (tci_mask=%x).",
+                                         ntohs(tci));
                                err = -EINVAL;
                                goto free_newmask;
                        }
                }
 
-               err = ovs_key_from_nlattrs(match, mask_attrs, a, true);
+               err = ovs_key_from_nlattrs(net, match, mask_attrs, a, true,
+                                          log);
                if (err)
                        goto free_newmask;
        }
 
-       if (!match_validate(match, key_attrs, mask_attrs))
+       if (!match_validate(match, key_attrs, mask_attrs, log))
                err = -EINVAL;
 
 free_newmask:
@@ -1071,26 +1273,84 @@ free_newmask:
        return err;
 }
 
+static size_t get_ufid_len(const struct nlattr *attr, bool log)
+{
+       size_t len;
+
+       if (!attr)
+               return 0;
+
+       len = nla_len(attr);
+       if (len < 1 || len > MAX_UFID_LENGTH) {
+               OVS_NLERR(log, "ufid size %u bytes exceeds the range (1, %d)",
+                         nla_len(attr), MAX_UFID_LENGTH);
+               return 0;
+       }
+
+       return len;
+}
+
+/* Initializes 'flow->ufid', returning true if 'attr' contains a valid UFID,
+ * or false otherwise.
+ */
+bool ovs_nla_get_ufid(struct sw_flow_id *sfid, const struct nlattr *attr,
+                     bool log)
+{
+       sfid->ufid_len = get_ufid_len(attr, log);
+       if (sfid->ufid_len)
+               memcpy(sfid->ufid, nla_data(attr), sfid->ufid_len);
+
+       return sfid->ufid_len;
+}
+
+int ovs_nla_get_identifier(struct sw_flow_id *sfid, const struct nlattr *ufid,
+                          const struct sw_flow_key *key, bool log)
+{
+       struct sw_flow_key *new_key;
+
+       if (ovs_nla_get_ufid(sfid, ufid, log))
+               return 0;
+
+       /* If UFID was not provided, use unmasked key. */
+       new_key = kmalloc(sizeof(*new_key), GFP_KERNEL);
+       if (!new_key)
+               return -ENOMEM;
+       memcpy(new_key, key, sizeof(*key));
+       sfid->unmasked_key = new_key;
+
+       return 0;
+}
+
+u32 ovs_nla_get_ufid_flags(const struct nlattr *attr)
+{
+       return attr ? nla_get_u32(attr) : 0;
+}
+
 /**
  * ovs_nla_get_flow_metadata - parses Netlink attributes into a flow key.
  * @key: Receives extracted in_port, priority, tun_key and skb_mark.
  * @attr: Netlink attribute holding nested %OVS_KEY_ATTR_* Netlink attribute
  * sequence.
+ * @log: Boolean to allow kernel error logging.  Normally true, but when
+ * probing for feature compatibility this should be passed in as false to
+ * suppress unnecessary error logging.
  *
  * This parses a series of Netlink attributes that form a flow key, which must
  * take the same form accepted by flow_from_nlattrs(), but only enough of it to
  * get the metadata, that is, the parts of the flow key that cannot be
  * extracted from the packet itself.
  */
-int ovs_nla_get_flow_metadata(const struct nlattr *attr,
-                             struct sw_flow_key *key)
+
+int ovs_nla_get_flow_metadata(struct net *net, const struct nlattr *attr,
+                             struct sw_flow_key *key,
+                             bool log)
 {
        const struct nlattr *a[OVS_KEY_ATTR_MAX + 1];
        struct sw_flow_match match;
        u64 attrs = 0;
        int err;
 
-       err = parse_flow_nlattrs(attr, a, &attrs);
+       err = parse_flow_nlattrs(attr, a, &attrs, log);
        if (err)
                return -EINVAL;
 
@@ -1098,35 +1358,36 @@ int ovs_nla_get_flow_metadata(const struct nlattr *attr,
        match.key = key;
 
        memset(key, 0, OVS_SW_FLOW_KEY_METADATA_SIZE);
+       memset(&key->ct, 0, sizeof(key->ct));
        key->phy.in_port = DP_MAX_PORTS;
 
-       return metadata_from_nlattrs(&match, &attrs, a, false);
+       return metadata_from_nlattrs(net, &match, &attrs, a, false, log);
 }
 
-int ovs_nla_put_flow(const struct sw_flow_key *swkey,
-                    const struct sw_flow_key *output, struct sk_buff *skb)
+static int __ovs_nla_put_key(const struct sw_flow_key *swkey,
+                            const struct sw_flow_key *output, bool is_mask,
+                            struct sk_buff *skb)
 {
        struct ovs_key_ethernet *eth_key;
        struct nlattr *nla, *encap;
-       bool is_mask = (swkey != output);
 
-       if (nla_put_u32(skb, OVS_KEY_ATTR_DP_HASH, output->ovs_flow_hash))
+       if (nla_put_u32(skb, OVS_KEY_ATTR_RECIRC_ID, output->recirc_id))
                goto nla_put_failure;
 
-       if (nla_put_u32(skb, OVS_KEY_ATTR_RECIRC_ID, output->recirc_id))
+       if (nla_put_u32(skb, OVS_KEY_ATTR_DP_HASH, output->ovs_flow_hash))
                goto nla_put_failure;
 
        if (nla_put_u32(skb, OVS_KEY_ATTR_PRIORITY, output->phy.priority))
                goto nla_put_failure;
 
-       if ((swkey->tun_key.ipv4_dst || is_mask)) {
-               const struct geneve_opt *opts = NULL;
+       if ((swkey->tun_key.u.ipv4.dst || is_mask)) {
+               const void *opts = NULL;
 
                if (output->tun_key.tun_flags & TUNNEL_OPTIONS_PRESENT)
-                       opts = GENEVE_OPTS(output, swkey->tun_opts_len);
+                       opts = TUN_METADATA_OPTS(output, swkey->tun_opts_len);
 
                if (ipv4_tun_to_nlattr(skb, &output->tun_key, opts,
-                                       swkey->tun_opts_len))
+                                      swkey->tun_opts_len))
                        goto nla_put_failure;
        }
 
@@ -1146,6 +1407,9 @@ int ovs_nla_put_flow(const struct sw_flow_key *swkey,
        if (nla_put_u32(skb, OVS_KEY_ATTR_SKB_MARK, output->phy.skb_mark))
                goto nla_put_failure;
 
+       if (ovs_ct_put_key(output, skb))
+               goto nla_put_failure;
+
        nla = nla_reserve(skb, OVS_KEY_ATTR_ETHERNET, sizeof(*eth_key));
        if (!nla)
                goto nla_put_failure;
@@ -1318,15 +1582,57 @@ nla_put_failure:
        return -EMSGSIZE;
 }
 
+int ovs_nla_put_key(const struct sw_flow_key *swkey,
+                   const struct sw_flow_key *output, int attr, bool is_mask,
+                   struct sk_buff *skb)
+{
+       int err;
+       struct nlattr *nla;
+
+       nla = nla_nest_start(skb, attr);
+       if (!nla)
+               return -EMSGSIZE;
+       err = __ovs_nla_put_key(swkey, output, is_mask, skb);
+       if (err)
+               return err;
+       nla_nest_end(skb, nla);
+
+       return 0;
+}
+
+/* Called with ovs_mutex or RCU read lock. */
+int ovs_nla_put_identifier(const struct sw_flow *flow, struct sk_buff *skb)
+{
+       if (ovs_identifier_is_ufid(&flow->id))
+               return nla_put(skb, OVS_FLOW_ATTR_UFID, flow->id.ufid_len,
+                              flow->id.ufid);
+
+       return ovs_nla_put_key(flow->id.unmasked_key, flow->id.unmasked_key,
+                              OVS_FLOW_ATTR_KEY, false, skb);
+}
+
+/* Called with ovs_mutex or RCU read lock. */
+int ovs_nla_put_masked_key(const struct sw_flow *flow, struct sk_buff *skb)
+{
+       return ovs_nla_put_key(&flow->key, &flow->key,
+                               OVS_FLOW_ATTR_KEY, false, skb);
+}
+
+/* Called with ovs_mutex or RCU read lock. */
+int ovs_nla_put_mask(const struct sw_flow *flow, struct sk_buff *skb)
+{
+       return ovs_nla_put_key(&flow->key, &flow->mask->key,
+                               OVS_FLOW_ATTR_MASK, true, skb);
+}
+
 #define MAX_ACTIONS_BUFSIZE    (32 * 1024)
 
-static struct sw_flow_actions *nla_alloc_flow_actions(int size)
+static struct sw_flow_actions *nla_alloc_flow_actions(int size, bool log)
 {
        struct sw_flow_actions *sfa;
 
        if (size > MAX_ACTIONS_BUFSIZE) {
-               OVS_NLERR("Flow action size (%u bytes) exceeds maximum "
-                         "(%u bytes)\n", size, MAX_ACTIONS_BUFSIZE);
+               OVS_NLERR(log, "Flow action size %u bytes exceeds max", size);
                return ERR_PTR(-EINVAL);
        }
 
@@ -1338,23 +1644,56 @@ static struct sw_flow_actions *nla_alloc_flow_actions(int size)
        return sfa;
 }
 
-/* RCU callback used by ovs_nla_free_flow_actions. */
-static void rcu_free_acts_callback(struct rcu_head *rcu)
+static void ovs_nla_free_set_action(const struct nlattr *a)
+{
+       const struct nlattr *ovs_key = nla_data(a);
+       struct ovs_tunnel_info *ovs_tun;
+
+       switch (nla_type(ovs_key)) {
+       case OVS_KEY_ATTR_TUNNEL_INFO:
+               ovs_tun = nla_data(ovs_key);
+               ovs_dst_release((struct dst_entry *)ovs_tun->tun_dst);
+               break;
+       }
+}
+
+void ovs_nla_free_flow_actions(struct sw_flow_actions *sf_acts)
 {
-       struct sw_flow_actions *sf_acts = container_of(rcu,
-                       struct sw_flow_actions, rcu);
+       const struct nlattr *a;
+       int rem;
+
+       if (!sf_acts)
+               return;
+
+       nla_for_each_attr(a, sf_acts->actions, sf_acts->actions_len, rem) {
+               switch (nla_type(a)) {
+               case OVS_ACTION_ATTR_SET:
+                       ovs_nla_free_set_action(a);
+                       break;
+               case OVS_ACTION_ATTR_CT:
+                       ovs_ct_free_action(a);
+                       break;
+               }
+       }
+
        kfree(sf_acts);
 }
 
+static void __ovs_nla_free_flow_actions(struct rcu_head *head)
+{
+       ovs_nla_free_flow_actions(container_of(head, struct sw_flow_actions, rcu));
+}
+
 /* Schedules 'sf_acts' to be freed after the next RCU grace period.
- * The caller must hold rcu_read_lock for this to be sensible. */
-void ovs_nla_free_flow_actions(struct sw_flow_actions *sf_acts)
+ * The caller must hold rcu_read_lock for this to be sensible.
+ */
+void ovs_nla_free_flow_actions_rcu(struct sw_flow_actions *sf_acts)
 {
-       call_rcu(&sf_acts->rcu, rcu_free_acts_callback);
+       call_rcu(&sf_acts->rcu, __ovs_nla_free_flow_actions);
 }
 
 static struct nlattr *reserve_sfa_size(struct sw_flow_actions **sfa,
-                                      int attr_len)
+                                      int attr_len, bool log)
 {
 
        struct sw_flow_actions *acts;
@@ -1374,12 +1713,13 @@ static struct nlattr *reserve_sfa_size(struct sw_flow_actions **sfa,
                new_acts_size = MAX_ACTIONS_BUFSIZE;
        }
 
-       acts = nla_alloc_flow_actions(new_acts_size);
+       acts = nla_alloc_flow_actions(new_acts_size, log);
        if (IS_ERR(acts))
                return (void *)acts;
 
        memcpy(acts->actions, (*sfa)->actions, (*sfa)->actions_len);
        acts->actions_len = (*sfa)->actions_len;
+       acts->orig_len = (*sfa)->orig_len;
        kfree(*sfa);
        *sfa = acts;
 
@@ -1388,12 +1728,12 @@ out:
        return  (struct nlattr *) ((unsigned char *)(*sfa) + next_offset);
 }
 
-static struct nlattr *__add_action(struct sw_flow_actions **sfa, int attrtype,
-                                  void *data, int len)
+static struct nlattr *__add_action(struct sw_flow_actions **sfa,
+                                  int attrtype, void *data, int len, bool log)
 {
        struct nlattr *a;
 
-       a = reserve_sfa_size(sfa, nla_attr_size(len));
+       a = reserve_sfa_size(sfa, nla_attr_size(len), log);
        if (IS_ERR(a))
                return a;
 
@@ -1407,12 +1747,12 @@ static struct nlattr *__add_action(struct sw_flow_actions **sfa, int attrtype,
        return a;
 }
 
-static int add_action(struct sw_flow_actions **sfa, int attrtype,
-                     void *data, int len)
+int ovs_nla_add_action(struct sw_flow_actions **sfa, int attrtype, void *data,
+                      int len, bool log)
 {
        struct nlattr *a;
 
-       a = __add_action(sfa, attrtype, data, len);
+       a = __add_action(sfa, attrtype, data, len, log);
        if (IS_ERR(a))
                return PTR_ERR(a);
 
@@ -1420,12 +1760,12 @@ static int add_action(struct sw_flow_actions **sfa, int attrtype,
 }
 
 static inline int add_nested_action_start(struct sw_flow_actions **sfa,
-                                         int attrtype)
+                                         int attrtype, bool log)
 {
        int used = (*sfa)->actions_len;
        int err;
 
-       err = add_action(sfa, attrtype, NULL, 0);
+       err = ovs_nla_add_action(sfa, attrtype, NULL, 0, log);
        if (err)
                return err;
 
@@ -1441,15 +1781,15 @@ static inline void add_nested_action_end(struct sw_flow_actions *sfa,
        a->nla_len = sfa->actions_len - st_offset;
 }
 
-static int __ovs_nla_copy_actions(const struct nlattr *attr,
+static int __ovs_nla_copy_actions(struct net *net, const struct nlattr *attr,
                                  const struct sw_flow_key *key,
                                  int depth, struct sw_flow_actions **sfa,
-                                 __be16 eth_type, __be16 vlan_tci);
+                                 __be16 eth_type, __be16 vlan_tci, bool log);
 
-static int validate_and_copy_sample(const struct nlattr *attr,
+static int validate_and_copy_sample(struct net *net, const struct nlattr *attr,
                                    const struct sw_flow_key *key, int depth,
                                    struct sw_flow_actions **sfa,
-                                   __be16 eth_type, __be16 vlan_tci)
+                                   __be16 eth_type, __be16 vlan_tci, bool log)
 {
        const struct nlattr *attrs[OVS_SAMPLE_ATTR_MAX + 1];
        const struct nlattr *probability, *actions;
@@ -1475,19 +1815,19 @@ static int validate_and_copy_sample(const struct nlattr *attr,
                return -EINVAL;
 
        /* validation done, copy sample action. */
-       start = add_nested_action_start(sfa, OVS_ACTION_ATTR_SAMPLE);
+       start = add_nested_action_start(sfa, OVS_ACTION_ATTR_SAMPLE, log);
        if (start < 0)
                return start;
-       err = add_action(sfa, OVS_SAMPLE_ATTR_PROBABILITY,
-                        nla_data(probability), sizeof(u32));
+       err = ovs_nla_add_action(sfa, OVS_SAMPLE_ATTR_PROBABILITY,
+                                nla_data(probability), sizeof(u32), log);
        if (err)
                return err;
-       st_acts = add_nested_action_start(sfa, OVS_SAMPLE_ATTR_ACTIONS);
+       st_acts = add_nested_action_start(sfa, OVS_SAMPLE_ATTR_ACTIONS, log);
        if (st_acts < 0)
                return st_acts;
 
-       err = __ovs_nla_copy_actions(actions, key, depth + 1, sfa,
-                                    eth_type, vlan_tci);
+       err = __ovs_nla_copy_actions(net, actions, key, depth + 1, sfa,
+                                    eth_type, vlan_tci, log);
        if (err)
                return err;
 
@@ -1497,16 +1837,6 @@ static int validate_and_copy_sample(const struct nlattr *attr,
        return 0;
 }
 
-static int validate_tp_port(const struct sw_flow_key *flow_key,
-                           __be16 eth_type)
-{
-       if ((eth_type == htons(ETH_P_IP) || eth_type == htons(ETH_P_IPV6)) &&
-           (flow_key->tp.src || flow_key->tp.dst))
-               return 0;
-
-       return -EINVAL;
-}
-
 void ovs_match_init(struct sw_flow_match *match,
                    struct sw_flow_key *key,
                    struct sw_flow_mask *mask)
@@ -1523,90 +1853,132 @@ void ovs_match_init(struct sw_flow_match *match,
        }
 }
 
-static int validate_and_copy_set_tun(const struct nlattr *attr,
-                                    struct sw_flow_actions **sfa)
+static int validate_geneve_opts(struct sw_flow_key *key)
 {
-       struct sw_flow_match match;
-       struct sw_flow_key key;
-       struct ovs_tunnel_info *tun_info;
-       struct nlattr *a;
-       int err, start;
+       struct geneve_opt *option;
+       int opts_len = key->tun_opts_len;
+       bool crit_opt = false;
 
-       ovs_match_init(&match, &key, NULL);
-       err = ipv4_tun_from_nlattr(nla_data(attr), &match, false);
-       if (err)
-               return err;
+       option = (struct geneve_opt *)TUN_METADATA_OPTS(key, key->tun_opts_len);
+       while (opts_len > 0) {
+               int len;
 
-       if (key.tun_opts_len) {
-               struct geneve_opt *option = GENEVE_OPTS(&key,
-                                                       key.tun_opts_len);
-               int opts_len = key.tun_opts_len;
-               bool crit_opt = false;
+               if (opts_len < sizeof(*option))
+                       return -EINVAL;
 
-               while (opts_len > 0) {
-                       int len;
+               len = sizeof(*option) + option->length * 4;
+               if (len > opts_len)
+                       return -EINVAL;
 
-                       if (opts_len < sizeof(*option))
-                               return -EINVAL;
+               crit_opt |= !!(option->type & GENEVE_CRIT_OPT_TYPE);
 
-                       len = sizeof(*option) + option->length * 4;
-                       if (len > opts_len)
-                               return -EINVAL;
+               option = (struct geneve_opt *)((u8 *)option + len);
+               opts_len -= len;
+       };
 
-                       crit_opt |= !!(option->type & GENEVE_CRIT_OPT_TYPE);
+       key->tun_key.tun_flags |= crit_opt ? TUNNEL_CRIT_OPT : 0;
 
-                       option = (struct geneve_opt *)((u8 *)option + len);
-                       opts_len -= len;
-               };
+       return 0;
+}
+
+static int validate_and_copy_set_tun(const struct nlattr *attr,
+                                    struct sw_flow_actions **sfa, bool log)
+{
+       struct sw_flow_match match;
+       struct sw_flow_key key;
+       struct metadata_dst *tun_dst;
+       struct ip_tunnel_info *tun_info;
+       struct ovs_tunnel_info *ovs_tun;
+       struct nlattr *a;
+       int err = 0, start, opts_type;
+
+       ovs_match_init(&match, &key, NULL);
+       opts_type = ipv4_tun_from_nlattr(nla_data(attr), &match, false, log);
+       if (opts_type < 0)
+               return opts_type;
 
-               key.tun_key.tun_flags |= crit_opt ? TUNNEL_CRIT_OPT : 0;
+       if (key.tun_opts_len) {
+               switch (opts_type) {
+               case OVS_TUNNEL_KEY_ATTR_GENEVE_OPTS:
+                       err = validate_geneve_opts(&key);
+                       if (err < 0)
+                               return err;
+                       break;
+               case OVS_TUNNEL_KEY_ATTR_VXLAN_OPTS:
+                       break;
+               }
        };
 
-       start = add_nested_action_start(sfa, OVS_ACTION_ATTR_SET);
+       start = add_nested_action_start(sfa, OVS_ACTION_ATTR_SET, log);
        if (start < 0)
                return start;
 
+       tun_dst = metadata_dst_alloc(key.tun_opts_len, GFP_KERNEL);
+       if (!tun_dst)
+               return -ENOMEM;
+
        a = __add_action(sfa, OVS_KEY_ATTR_TUNNEL_INFO, NULL,
-                       sizeof(*tun_info) + key.tun_opts_len);
-       if (IS_ERR(a))
+                        sizeof(*ovs_tun), log);
+       if (IS_ERR(a)) {
+               ovs_dst_release((struct dst_entry *)tun_dst);
                return PTR_ERR(a);
+       }
 
-       tun_info = nla_data(a);
-       tun_info->tunnel = key.tun_key;
-       tun_info->options_len = key.tun_opts_len;
+       ovs_tun = nla_data(a);
+       ovs_tun->tun_dst = tun_dst;
 
-       if (tun_info->options_len) {
-               /* We need to store the options in the action itself since
-                * everything else will go away after flow setup. We can append
-                * it to tun_info and then point there.
-                */
-               memcpy((tun_info + 1), GENEVE_OPTS(&key, key.tun_opts_len),
-                       key.tun_opts_len);
-               tun_info->options = (struct geneve_opt *)(tun_info + 1);
-       } else {
-               tun_info->options = NULL;
-       }
+       tun_info = &tun_dst->u.tun_info;
+       tun_info->mode = IP_TUNNEL_INFO_TX;
+       tun_info->key = key.tun_key;
 
+       /* We need to store the options in the action itself since
+        * everything else will go away after flow setup. We can append
+        * it to tun_info and then point there.
+        */
+       ip_tunnel_info_opts_set(tun_info,
+                               TUN_METADATA_OPTS(&key, key.tun_opts_len),
+                               key.tun_opts_len);
        add_nested_action_end(*sfa, start);
 
        return err;
 }
 
+/* Return false if there are any non-masked bits set.
+ * Mask follows data immediately, before any netlink padding.
+ */
+static bool validate_masked(u8 *data, int len)
+{
+       u8 *mask = data + len;
+
+       while (len--)
+               if (*data++ & ~*mask++)
+                       return false;
+
+       return true;
+}
+
 static int validate_set(const struct nlattr *a,
                        const struct sw_flow_key *flow_key,
                        struct sw_flow_actions **sfa,
-                       bool *set_tun, __be16 eth_type)
+                       bool *skip_copy, __be16 eth_type, bool masked, bool log)
 {
        const struct nlattr *ovs_key = nla_data(a);
        int key_type = nla_type(ovs_key);
+       size_t key_len;
 
        /* There can be only one key in a action */
        if (nla_total_size(nla_len(ovs_key)) != nla_len(a))
                return -EINVAL;
 
+       key_len = nla_len(ovs_key);
+       if (masked)
+               key_len /= 2;
+
        if (key_type > OVS_KEY_ATTR_MAX ||
-           (ovs_key_lens[key_type] != nla_len(ovs_key) &&
-            ovs_key_lens[key_type] != -1))
+           !check_attr_len(key_len, ovs_key_lens[key_type].len))
+               return -EINVAL;
+
+       if (masked && !validate_masked(nla_data(ovs_key), key_len))
                return -EINVAL;
 
        switch (key_type) {
@@ -1616,12 +1988,20 @@ static int validate_set(const struct nlattr *a,
 
        case OVS_KEY_ATTR_PRIORITY:
        case OVS_KEY_ATTR_SKB_MARK:
+       case OVS_KEY_ATTR_CT_MARK:
+       case OVS_KEY_ATTR_CT_LABELS:
        case OVS_KEY_ATTR_ETHERNET:
                break;
 
        case OVS_KEY_ATTR_TUNNEL:
-               *set_tun = true;
-               err = validate_and_copy_set_tun(a, sfa);
+               if (eth_p_mpls(eth_type))
+                       return -EINVAL;
+
+               if (masked)
+                       return -EINVAL; /* Masked tunnel set not supported. */
+
+               *skip_copy = true;
+               err = validate_and_copy_set_tun(a, sfa, log);
                if (err)
                        return err;
                break;
@@ -1630,48 +2010,66 @@ static int validate_set(const struct nlattr *a,
                if (eth_type != htons(ETH_P_IP))
                        return -EINVAL;
 
-               if (!flow_key->ip.proto)
-                       return -EINVAL;
-
                ipv4_key = nla_data(ovs_key);
-               if (ipv4_key->ipv4_proto != flow_key->ip.proto)
-                       return -EINVAL;
 
-               if (ipv4_key->ipv4_frag != flow_key->ip.frag)
-                       return -EINVAL;
+               if (masked) {
+                       const struct ovs_key_ipv4 *mask = ipv4_key + 1;
 
+                       /* Non-writeable fields. */
+                       if (mask->ipv4_proto || mask->ipv4_frag)
+                               return -EINVAL;
+               } else {
+                       if (ipv4_key->ipv4_proto != flow_key->ip.proto)
+                               return -EINVAL;
+
+                       if (ipv4_key->ipv4_frag != flow_key->ip.frag)
+                               return -EINVAL;
+               }
                break;
 
        case OVS_KEY_ATTR_IPV6:
                if (eth_type != htons(ETH_P_IPV6))
                        return -EINVAL;
 
-               if (!flow_key->ip.proto)
-                       return -EINVAL;
-
                ipv6_key = nla_data(ovs_key);
-               if (ipv6_key->ipv6_proto != flow_key->ip.proto)
-                       return -EINVAL;
 
-               if (ipv6_key->ipv6_frag != flow_key->ip.frag)
-                       return -EINVAL;
+               if (masked) {
+                       const struct ovs_key_ipv6 *mask = ipv6_key + 1;
 
+                       /* Non-writeable fields. */
+                       if (mask->ipv6_proto || mask->ipv6_frag)
+                               return -EINVAL;
+
+                       /* Invalid bits in the flow label mask? */
+                       if (ntohl(mask->ipv6_label) & 0xFFF00000)
+                               return -EINVAL;
+               } else {
+                       if (ipv6_key->ipv6_proto != flow_key->ip.proto)
+                               return -EINVAL;
+
+                       if (ipv6_key->ipv6_frag != flow_key->ip.frag)
+                               return -EINVAL;
+               }
                if (ntohl(ipv6_key->ipv6_label) & 0xFFF00000)
                        return -EINVAL;
 
                break;
 
        case OVS_KEY_ATTR_TCP:
-               if (flow_key->ip.proto != IPPROTO_TCP)
+               if ((eth_type != htons(ETH_P_IP) &&
+                    eth_type != htons(ETH_P_IPV6)) ||
+                   flow_key->ip.proto != IPPROTO_TCP)
                        return -EINVAL;
 
-               return validate_tp_port(flow_key, eth_type);
+               break;
 
        case OVS_KEY_ATTR_UDP:
-               if (flow_key->ip.proto != IPPROTO_UDP)
+               if ((eth_type != htons(ETH_P_IP) &&
+                    eth_type != htons(ETH_P_IPV6)) ||
+                   flow_key->ip.proto != IPPROTO_UDP)
                        return -EINVAL;
 
-               return validate_tp_port(flow_key, eth_type);
+               break;
 
        case OVS_KEY_ATTR_MPLS:
                if (!eth_p_mpls(eth_type))
@@ -1679,15 +2077,45 @@ static int validate_set(const struct nlattr *a,
                break;
 
        case OVS_KEY_ATTR_SCTP:
-               if (flow_key->ip.proto != IPPROTO_SCTP)
+               if ((eth_type != htons(ETH_P_IP) &&
+                    eth_type != htons(ETH_P_IPV6)) ||
+                   flow_key->ip.proto != IPPROTO_SCTP)
                        return -EINVAL;
 
-               return validate_tp_port(flow_key, eth_type);
+               break;
 
        default:
                return -EINVAL;
        }
 
+       /* Convert non-masked non-tunnel set actions to masked set actions. */
+       if (!masked && key_type != OVS_KEY_ATTR_TUNNEL) {
+               int start, len = key_len * 2;
+               struct nlattr *at;
+
+               *skip_copy = true;
+
+               start = add_nested_action_start(sfa,
+                                               OVS_ACTION_ATTR_SET_TO_MASKED,
+                                               log);
+               if (start < 0)
+                       return start;
+
+               at = __add_action(sfa, key_type, NULL, len, log);
+               if (IS_ERR(at))
+                       return PTR_ERR(at);
+
+               memcpy(nla_data(at), nla_data(ovs_key), key_len); /* Key. */
+               memset(nla_data(at) + key_len, 0xff, key_len);    /* Mask. */
+               /* Clear non-writeable bits from otherwise writeable fields. */
+               if (key_type == OVS_KEY_ATTR_IPV6) {
+                       struct ovs_key_ipv6 *mask = nla_data(at) + key_len;
+
+                       mask->ipv6_label &= htonl(0x000FFFFF);
+               }
+               add_nested_action_end(*sfa, start);
+       }
+
        return 0;
 }
 
@@ -1714,12 +2142,12 @@ static int validate_userspace(const struct nlattr *attr)
 }
 
 static int copy_action(const struct nlattr *from,
-                      struct sw_flow_actions **sfa)
+                      struct sw_flow_actions **sfa, bool log)
 {
        int totlen = NLA_ALIGN(from->nla_len);
        struct nlattr *to;
 
-       to = reserve_sfa_size(sfa, from->nla_len);
+       to = reserve_sfa_size(sfa, from->nla_len, log);
        if (IS_ERR(to))
                return PTR_ERR(to);
 
@@ -1727,10 +2155,10 @@ static int copy_action(const struct nlattr *from,
        return 0;
 }
 
-static int __ovs_nla_copy_actions(const struct nlattr *attr,
+static int __ovs_nla_copy_actions(struct net *net, const struct nlattr *attr,
                                  const struct sw_flow_key *key,
                                  int depth, struct sw_flow_actions **sfa,
-                                 __be16 eth_type, __be16 vlan_tci)
+                                 __be16 eth_type, __be16 vlan_tci, bool log)
 {
        const struct nlattr *a;
        int rem, err;
@@ -1749,8 +2177,10 @@ static int __ovs_nla_copy_actions(const struct nlattr *attr,
                        [OVS_ACTION_ATTR_PUSH_VLAN] = sizeof(struct ovs_action_push_vlan),
                        [OVS_ACTION_ATTR_POP_VLAN] = 0,
                        [OVS_ACTION_ATTR_SET] = (u32)-1,
+                       [OVS_ACTION_ATTR_SET_MASKED] = (u32)-1,
                        [OVS_ACTION_ATTR_SAMPLE] = (u32)-1,
-                       [OVS_ACTION_ATTR_HASH] = sizeof(struct ovs_action_hash)
+                       [OVS_ACTION_ATTR_HASH] = sizeof(struct ovs_action_hash),
+                       [OVS_ACTION_ATTR_CT] = (u32)-1,
                };
                const struct ovs_action_push_vlan *vlan;
                int type = nla_type(a);
@@ -1843,24 +2273,40 @@ static int __ovs_nla_copy_actions(const struct nlattr *attr,
                        break;
 
                case OVS_ACTION_ATTR_SET:
-                       err = validate_set(a, key, sfa, &skip_copy, eth_type);
+                       err = validate_set(a, key, sfa,
+                                          &skip_copy, eth_type, false, log);
+                       if (err)
+                               return err;
+                       break;
+
+               case OVS_ACTION_ATTR_SET_MASKED:
+                       err = validate_set(a, key, sfa,
+                                          &skip_copy, eth_type, true, log);
                        if (err)
                                return err;
                        break;
 
                case OVS_ACTION_ATTR_SAMPLE:
-                       err = validate_and_copy_sample(a, key, depth, sfa,
-                                                      eth_type, vlan_tci);
+                       err = validate_and_copy_sample(net, a, key, depth, sfa,
+                                                      eth_type, vlan_tci, log);
+                       if (err)
+                               return err;
+                       skip_copy = true;
+                       break;
+
+               case OVS_ACTION_ATTR_CT:
+                       err = ovs_ct_copy_action(net, a, key, sfa, log);
                        if (err)
                                return err;
                        skip_copy = true;
                        break;
 
                default:
+                       OVS_NLERR(log, "Unknown Action type %d", type);
                        return -EINVAL;
                }
                if (!skip_copy) {
-                       err = copy_action(a, sfa);
+                       err = copy_action(a, sfa, log);
                        if (err)
                                return err;
                }
@@ -1872,20 +2318,22 @@ static int __ovs_nla_copy_actions(const struct nlattr *attr,
        return 0;
 }
 
-int ovs_nla_copy_actions(const struct nlattr *attr,
+/* 'key' must be the masked key. */
+int ovs_nla_copy_actions(struct net *net, const struct nlattr *attr,
                         const struct sw_flow_key *key,
-                        struct sw_flow_actions **sfa)
+                        struct sw_flow_actions **sfa, bool log)
 {
        int err;
 
-       *sfa = nla_alloc_flow_actions(nla_len(attr));
+       *sfa = nla_alloc_flow_actions(nla_len(attr), log);
        if (IS_ERR(*sfa))
                return PTR_ERR(*sfa);
 
-       err = __ovs_nla_copy_actions(attr, key, 0, sfa, key->eth.type,
-                                    key->eth.tci);
+       (*sfa)->orig_len = nla_len(attr);
+       err = __ovs_nla_copy_actions(net, attr, key, 0, sfa, key->eth.type,
+                                    key->eth.tci, log);
        if (err)
-               kfree(*sfa);
+               ovs_nla_free_flow_actions(*sfa);
 
        return err;
 }
@@ -1935,15 +2383,16 @@ static int set_action_to_attr(const struct nlattr *a, struct sk_buff *skb)
 
        switch (key_type) {
        case OVS_KEY_ATTR_TUNNEL_INFO: {
-               struct ovs_tunnel_info *tun_info = nla_data(ovs_key);
+               struct ovs_tunnel_info *ovs_tun = nla_data(ovs_key);
+               struct ip_tunnel_info *tun_info = &ovs_tun->tun_dst->u.tun_info;
 
                start = nla_nest_start(skb, OVS_ACTION_ATTR_SET);
                if (!start)
                        return -EMSGSIZE;
 
-               err = ipv4_tun_to_nlattr(skb, &tun_info->tunnel,
+               err = ipv4_tun_to_nlattr(skb, &tun_info->key,
                                         tun_info->options_len ?
-                                               tun_info->options : NULL,
+                                            ip_tunnel_info_opts(tun_info) : NULL,
                                         tun_info->options_len);
                if (err)
                        return err;
@@ -1959,6 +2408,27 @@ static int set_action_to_attr(const struct nlattr *a, struct sk_buff *skb)
        return 0;
 }
 
+static int masked_set_action_to_set_action_attr(const struct nlattr *a,
+                                               struct sk_buff *skb)
+{
+       const struct nlattr *ovs_key = nla_data(a);
+       struct nlattr *nla;
+       size_t key_len = nla_len(ovs_key) / 2;
+
+       /* Revert the conversion we did from a non-masked set action to
+        * masked set action.
+        */
+       nla = nla_nest_start(skb, OVS_ACTION_ATTR_SET);
+       if (!nla)
+               return -EMSGSIZE;
+
+       if (nla_put(skb, nla_type(ovs_key), key_len, nla_data(ovs_key)))
+               return -EMSGSIZE;
+
+       nla_nest_end(skb, nla);
+       return 0;
+}
+
 int ovs_nla_put_actions(const struct nlattr *attr, int len, struct sk_buff *skb)
 {
        const struct nlattr *a;
@@ -1974,11 +2444,24 @@ int ovs_nla_put_actions(const struct nlattr *attr, int len, struct sk_buff *skb)
                                return err;
                        break;
 
+               case OVS_ACTION_ATTR_SET_TO_MASKED:
+                       err = masked_set_action_to_set_action_attr(a, skb);
+                       if (err)
+                               return err;
+                       break;
+
                case OVS_ACTION_ATTR_SAMPLE:
                        err = sample_action_to_attr(a, skb);
                        if (err)
                                return err;
                        break;
+
+               case OVS_ACTION_ATTR_CT:
+                       err = ovs_ct_action_to_attr(nla_data(a), skb);
+                       if (err)
+                               return err;
+                       break;
+
                default:
                        if (nla_put(skb, type, nla_len(a), nla_data(a)))
                                return -EMSGSIZE;