ethdev: add GTP extension header to flow API
[dpdk.git] / app / test-pmd / cmdline_flow.c
index 659908f..cdaabcc 100644 (file)
@@ -13,8 +13,8 @@
 #include <arpa/inet.h>
 #include <sys/socket.h>
 
+#include <rte_string_fns.h>
 #include <rte_common.h>
-#include <rte_eth_ctrl.h>
 #include <rte_ethdev.h>
 #include <rte_byteorder.h>
 #include <cmdline_parse.h>
@@ -28,6 +28,8 @@ enum index {
        /* Special tokens. */
        ZERO = 0,
        END,
+       START_SET,
+       END_SET,
 
        /* Common tokens. */
        INTEGER,
@@ -35,6 +37,7 @@ enum index {
        PREFIX,
        BOOLEAN,
        STRING,
+       HEX,
        MAC_ADDR,
        IPV4_ADDR,
        IPV6_ADDR,
@@ -44,8 +47,13 @@ enum index {
        PRIORITY_LEVEL,
 
        /* Top-level command. */
-       FLOW,
+       SET,
+       /* Sub-leve commands. */
+       SET_RAW_ENCAP,
+       SET_RAW_DECAP,
 
+       /* Top-level command. */
+       FLOW,
        /* Sub-level commands. */
        VALIDATE,
        CREATE,
@@ -145,8 +153,14 @@ enum index {
        ITEM_NVGRE_TNI,
        ITEM_MPLS,
        ITEM_MPLS_LABEL,
+       ITEM_MPLS_TC,
+       ITEM_MPLS_S,
        ITEM_GRE,
        ITEM_GRE_PROTO,
+       ITEM_GRE_C_RSVD0_VER,
+       ITEM_GRE_C_BIT,
+       ITEM_GRE_K_BIT,
+       ITEM_GRE_S_BIT,
        ITEM_FUZZY,
        ITEM_FUZZY_THRESH,
        ITEM_GTP,
@@ -180,6 +194,11 @@ enum index {
        ITEM_ICMP6_ND_OPT_TLA_ETH_TLA,
        ITEM_META,
        ITEM_META_DATA,
+       ITEM_GRE_KEY,
+       ITEM_GRE_KEY_VALUE,
+       ITEM_GTP_PSC,
+       ITEM_GTP_PSC_QFI,
+       ITEM_GTP_PSC_PDU_T,
 
        /* Validate/create actions. */
        ACTIONS,
@@ -271,6 +290,16 @@ enum index {
        ACTION_SET_MAC_SRC_MAC_SRC,
        ACTION_SET_MAC_DST,
        ACTION_SET_MAC_DST_MAC_DST,
+       ACTION_INC_TCP_SEQ,
+       ACTION_INC_TCP_SEQ_VALUE,
+       ACTION_DEC_TCP_SEQ,
+       ACTION_DEC_TCP_SEQ_VALUE,
+       ACTION_INC_TCP_ACK,
+       ACTION_INC_TCP_ACK_VALUE,
+       ACTION_DEC_TCP_ACK,
+       ACTION_DEC_TCP_ACK_VALUE,
+       ACTION_RAW_ENCAP,
+       ACTION_RAW_DECAP,
 };
 
 /** Maximum size for pattern in struct rte_flow_item_raw. */
@@ -293,6 +322,25 @@ struct action_rss_data {
 /** Maximum number of items in struct rte_flow_action_vxlan_encap. */
 #define ACTION_VXLAN_ENCAP_ITEMS_NUM 6
 
+#define ACTION_RAW_ENCAP_MAX_DATA 128
+
+/** Storage for struct rte_flow_action_raw_encap. */
+struct raw_encap_conf {
+       uint8_t data[ACTION_RAW_ENCAP_MAX_DATA];
+       uint8_t preserve[ACTION_RAW_ENCAP_MAX_DATA];
+       size_t size;
+};
+
+struct raw_encap_conf raw_encap_conf = {.size = 0};
+
+/** Storage for struct rte_flow_action_raw_decap. */
+struct raw_decap_conf {
+       uint8_t data[ACTION_RAW_ENCAP_MAX_DATA];
+       size_t size;
+};
+
+struct raw_decap_conf raw_decap_conf = {.size = 0};
+
 /** Storage for struct rte_flow_action_vxlan_encap including external data. */
 struct action_vxlan_encap_data {
        struct rte_flow_action_vxlan_encap conf;
@@ -484,6 +532,14 @@ struct token {
                .size = sizeof(((s *)0)->f), \
        })
 
+/** Same as ARGS_ENTRY_HTON() for a single argument, without structure. */
+#define ARG_ENTRY_HTON(s) \
+       (&(const struct arg){ \
+               .hton = 1, \
+               .offset = 0, \
+               .size = sizeof(s), \
+       })
+
 /** Parser output buffer layout expected by cmd_flow_parsed(). */
 struct buffer {
        enum index command; /**< Flow command. */
@@ -609,6 +665,9 @@ static const enum index next_item[] = {
        ITEM_ICMP6_ND_OPT_SLA_ETH,
        ITEM_ICMP6_ND_OPT_TLA_ETH,
        ITEM_META,
+       ITEM_GRE_KEY,
+       ITEM_GTP_PSC,
+       END_SET,
        ZERO,
 };
 
@@ -748,12 +807,24 @@ static const enum index item_nvgre[] = {
 
 static const enum index item_mpls[] = {
        ITEM_MPLS_LABEL,
+       ITEM_MPLS_TC,
+       ITEM_MPLS_S,
        ITEM_NEXT,
        ZERO,
 };
 
 static const enum index item_gre[] = {
        ITEM_GRE_PROTO,
+       ITEM_GRE_C_RSVD0_VER,
+       ITEM_GRE_C_BIT,
+       ITEM_GRE_K_BIT,
+       ITEM_GRE_S_BIT,
+       ITEM_NEXT,
+       ZERO,
+};
+
+static const enum index item_gre_key[] = {
+       ITEM_GRE_KEY_VALUE,
        ITEM_NEXT,
        ZERO,
 };
@@ -835,6 +906,13 @@ static const enum index item_meta[] = {
        ZERO,
 };
 
+static const enum index item_gtp_psc[] = {
+       ITEM_GTP_PSC_QFI,
+       ITEM_GTP_PSC_PDU_T,
+       ITEM_NEXT,
+       ZERO,
+};
+
 static const enum index next_action[] = {
        ACTION_END,
        ACTION_VOID,
@@ -884,6 +962,12 @@ static const enum index next_action[] = {
        ACTION_SET_TTL,
        ACTION_SET_MAC_SRC,
        ACTION_SET_MAC_DST,
+       ACTION_INC_TCP_SEQ,
+       ACTION_DEC_TCP_SEQ,
+       ACTION_INC_TCP_ACK,
+       ACTION_DEC_TCP_ACK,
+       ACTION_RAW_ENCAP,
+       ACTION_RAW_DECAP,
        ZERO,
 };
 
@@ -1046,6 +1130,36 @@ static const enum index action_set_mac_dst[] = {
        ZERO,
 };
 
+static const enum index action_inc_tcp_seq[] = {
+       ACTION_INC_TCP_SEQ_VALUE,
+       ACTION_NEXT,
+       ZERO,
+};
+
+static const enum index action_dec_tcp_seq[] = {
+       ACTION_DEC_TCP_SEQ_VALUE,
+       ACTION_NEXT,
+       ZERO,
+};
+
+static const enum index action_inc_tcp_ack[] = {
+       ACTION_INC_TCP_ACK_VALUE,
+       ACTION_NEXT,
+       ZERO,
+};
+
+static const enum index action_dec_tcp_ack[] = {
+       ACTION_DEC_TCP_ACK_VALUE,
+       ACTION_NEXT,
+       ZERO,
+};
+
+static int parse_set_raw_encap_decap(struct context *, const struct token *,
+                                    const char *, unsigned int,
+                                    void *, unsigned int);
+static int parse_set_init(struct context *, const struct token *,
+                         const char *, unsigned int,
+                         void *, unsigned int);
 static int parse_init(struct context *, const struct token *,
                      const char *, unsigned int,
                      void *, unsigned int);
@@ -1092,6 +1206,12 @@ static int parse_vc_action_mplsoudp_encap(struct context *,
 static int parse_vc_action_mplsoudp_decap(struct context *,
                                          const struct token *, const char *,
                                          unsigned int, void *, unsigned int);
+static int parse_vc_action_raw_encap(struct context *,
+                                    const struct token *, const char *,
+                                    unsigned int, void *, unsigned int);
+static int parse_vc_action_raw_decap(struct context *,
+                                    const struct token *, const char *,
+                                    unsigned int, void *, unsigned int);
 static int parse_destroy(struct context *, const struct token *,
                         const char *, unsigned int,
                         void *, unsigned int);
@@ -1122,6 +1242,9 @@ static int parse_boolean(struct context *, const struct token *,
 static int parse_string(struct context *, const struct token *,
                        const char *, unsigned int,
                        void *, unsigned int);
+static int parse_hex(struct context *ctx, const struct token *token,
+                       const char *str, unsigned int len,
+                       void *buf, unsigned int size);
 static int parse_mac_addr(struct context *, const struct token *,
                          const char *, unsigned int,
                          void *, unsigned int);
@@ -1162,6 +1285,16 @@ static const struct token token_list[] = {
                .type = "RETURN",
                .help = "command may end here",
        },
+       [START_SET] = {
+               .name = "START_SET",
+               .help = "null entry, abused as the entry point for set",
+               .next = NEXT(NEXT_ENTRY(SET)),
+       },
+       [END_SET] = {
+               .name = "end_set",
+               .type = "RETURN",
+               .help = "set command may end here",
+       },
        /* Common tokens. */
        [INTEGER] = {
                .name = "{int}",
@@ -1198,6 +1331,13 @@ static const struct token token_list[] = {
                .call = parse_string,
                .comp = comp_none,
        },
+       [HEX] = {
+               .name = "{hex}",
+               .type = "HEX",
+               .help = "fixed string",
+               .call = parse_hex,
+               .comp = comp_none,
+       },
        [MAC_ADDR] = {
                .name = "{MAC address}",
                .type = "MAC-48",
@@ -1873,6 +2013,22 @@ static const struct token token_list[] = {
                                                  label_tc_s,
                                                  "\xff\xff\xf0")),
        },
+       [ITEM_MPLS_TC] = {
+               .name = "tc",
+               .help = "MPLS Traffic Class",
+               .next = NEXT(item_mpls, NEXT_ENTRY(UNSIGNED), item_param),
+               .args = ARGS(ARGS_ENTRY_MASK_HTON(struct rte_flow_item_mpls,
+                                                 label_tc_s,
+                                                 "\x00\x00\x0e")),
+       },
+       [ITEM_MPLS_S] = {
+               .name = "s",
+               .help = "MPLS Bottom-of-Stack",
+               .next = NEXT(item_mpls, NEXT_ENTRY(UNSIGNED), item_param),
+               .args = ARGS(ARGS_ENTRY_MASK_HTON(struct rte_flow_item_mpls,
+                                                 label_tc_s,
+                                                 "\x00\x00\x01")),
+       },
        [ITEM_GRE] = {
                .name = "gre",
                .help = "match GRE header",
@@ -1887,6 +2043,40 @@ static const struct token token_list[] = {
                .args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_gre,
                                             protocol)),
        },
+       [ITEM_GRE_C_RSVD0_VER] = {
+               .name = "c_rsvd0_ver",
+               .help =
+                       "checksum (1b), undefined (1b), key bit (1b),"
+                       " sequence number (1b), reserved 0 (9b),"
+                       " version (3b)",
+               .next = NEXT(item_gre, NEXT_ENTRY(UNSIGNED), item_param),
+               .args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_gre,
+                                            c_rsvd0_ver)),
+       },
+       [ITEM_GRE_C_BIT] = {
+               .name = "c_bit",
+               .help = "checksum bit (C)",
+               .next = NEXT(item_gre, NEXT_ENTRY(BOOLEAN), item_param),
+               .args = ARGS(ARGS_ENTRY_MASK_HTON(struct rte_flow_item_gre,
+                                                 c_rsvd0_ver,
+                                                 "\x80\x00\x00\x00")),
+       },
+       [ITEM_GRE_S_BIT] = {
+               .name = "s_bit",
+               .help = "sequence number bit (S)",
+               .next = NEXT(item_gre, NEXT_ENTRY(BOOLEAN), item_param),
+               .args = ARGS(ARGS_ENTRY_MASK_HTON(struct rte_flow_item_gre,
+                                                 c_rsvd0_ver,
+                                                 "\x10\x00\x00\x00")),
+       },
+       [ITEM_GRE_K_BIT] = {
+               .name = "k_bit",
+               .help = "key bit (K)",
+               .next = NEXT(item_gre, NEXT_ENTRY(BOOLEAN), item_param),
+               .args = ARGS(ARGS_ENTRY_MASK_HTON(struct rte_flow_item_gre,
+                                                 c_rsvd0_ver,
+                                                 "\x20\x00\x00\x00")),
+       },
        [ITEM_FUZZY] = {
                .name = "fuzzy",
                .help = "fuzzy pattern match, expect faster than default",
@@ -2139,6 +2329,41 @@ static const struct token token_list[] = {
                .args = ARGS(ARGS_ENTRY_MASK_HTON(struct rte_flow_item_meta,
                                                  data, "\xff\xff\xff\xff")),
        },
+       [ITEM_GRE_KEY] = {
+               .name = "gre_key",
+               .help = "match GRE key",
+               .priv = PRIV_ITEM(GRE_KEY, sizeof(rte_be32_t)),
+               .next = NEXT(item_gre_key),
+               .call = parse_vc,
+       },
+       [ITEM_GRE_KEY_VALUE] = {
+               .name = "value",
+               .help = "key value",
+               .next = NEXT(item_gre_key, NEXT_ENTRY(UNSIGNED), item_param),
+               .args = ARGS(ARG_ENTRY_HTON(rte_be32_t)),
+       },
+       [ITEM_GTP_PSC] = {
+               .name = "gtp_psc",
+               .help = "match GTP extension header with type 0x85",
+               .priv = PRIV_ITEM(GTP_PSC,
+                               sizeof(struct rte_flow_item_gtp_psc)),
+               .next = NEXT(item_gtp_psc),
+               .call = parse_vc,
+       },
+       [ITEM_GTP_PSC_QFI] = {
+               .name = "qfi",
+               .help = "QoS flow identifier",
+               .next = NEXT(item_gtp_psc, NEXT_ENTRY(UNSIGNED), item_param),
+               .args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_gtp_psc,
+                                       qfi)),
+       },
+       [ITEM_GTP_PSC_PDU_T] = {
+               .name = "pdu_t",
+               .help = "PDU type",
+               .next = NEXT(item_gtp_psc, NEXT_ENTRY(UNSIGNED), item_param),
+               .args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_gtp_psc,
+                                       pdu_type)),
+       },
 
        /* Validate/create actions. */
        [ACTIONS] = {
@@ -2306,7 +2531,7 @@ static const struct token token_list[] = {
        [ACTION_RSS_KEY] = {
                .name = "key",
                .help = "RSS hash key",
-               .next = NEXT(action_rss, NEXT_ENTRY(STRING)),
+               .next = NEXT(action_rss, NEXT_ENTRY(HEX)),
                .args = ARGS(ARGS_ENTRY_ARB(0, 0),
                             ARGS_ENTRY_ARB
                             (offsetof(struct action_rss_data, conf) +
@@ -2843,6 +3068,101 @@ static const struct token token_list[] = {
                             (struct rte_flow_action_set_mac, mac_addr)),
                .call = parse_vc_conf,
        },
+       [ACTION_INC_TCP_SEQ] = {
+               .name = "inc_tcp_seq",
+               .help = "increase TCP sequence number",
+               .priv = PRIV_ACTION(INC_TCP_SEQ, sizeof(rte_be32_t)),
+               .next = NEXT(action_inc_tcp_seq),
+               .call = parse_vc,
+       },
+       [ACTION_INC_TCP_SEQ_VALUE] = {
+               .name = "value",
+               .help = "the value to increase TCP sequence number by",
+               .next = NEXT(action_inc_tcp_seq, NEXT_ENTRY(UNSIGNED)),
+               .args = ARGS(ARG_ENTRY_HTON(rte_be32_t)),
+               .call = parse_vc_conf,
+       },
+       [ACTION_DEC_TCP_SEQ] = {
+               .name = "dec_tcp_seq",
+               .help = "decrease TCP sequence number",
+               .priv = PRIV_ACTION(DEC_TCP_SEQ, sizeof(rte_be32_t)),
+               .next = NEXT(action_dec_tcp_seq),
+               .call = parse_vc,
+       },
+       [ACTION_DEC_TCP_SEQ_VALUE] = {
+               .name = "value",
+               .help = "the value to decrease TCP sequence number by",
+               .next = NEXT(action_dec_tcp_seq, NEXT_ENTRY(UNSIGNED)),
+               .args = ARGS(ARG_ENTRY_HTON(rte_be32_t)),
+               .call = parse_vc_conf,
+       },
+       [ACTION_INC_TCP_ACK] = {
+               .name = "inc_tcp_ack",
+               .help = "increase TCP acknowledgment number",
+               .priv = PRIV_ACTION(INC_TCP_ACK, sizeof(rte_be32_t)),
+               .next = NEXT(action_inc_tcp_ack),
+               .call = parse_vc,
+       },
+       [ACTION_INC_TCP_ACK_VALUE] = {
+               .name = "value",
+               .help = "the value to increase TCP acknowledgment number by",
+               .next = NEXT(action_inc_tcp_ack, NEXT_ENTRY(UNSIGNED)),
+               .args = ARGS(ARG_ENTRY_HTON(rte_be32_t)),
+               .call = parse_vc_conf,
+       },
+       [ACTION_DEC_TCP_ACK] = {
+               .name = "dec_tcp_ack",
+               .help = "decrease TCP acknowledgment number",
+               .priv = PRIV_ACTION(DEC_TCP_ACK, sizeof(rte_be32_t)),
+               .next = NEXT(action_dec_tcp_ack),
+               .call = parse_vc,
+       },
+       [ACTION_DEC_TCP_ACK_VALUE] = {
+               .name = "value",
+               .help = "the value to decrease TCP acknowledgment number by",
+               .next = NEXT(action_dec_tcp_ack, NEXT_ENTRY(UNSIGNED)),
+               .args = ARGS(ARG_ENTRY_HTON(rte_be32_t)),
+               .call = parse_vc_conf,
+       },
+       [ACTION_RAW_ENCAP] = {
+               .name = "raw_encap",
+               .help = "encapsulation data, defined by set raw_encap",
+               .priv = PRIV_ACTION(RAW_ENCAP,
+                       sizeof(struct rte_flow_action_raw_encap)),
+               .next = NEXT(NEXT_ENTRY(ACTION_NEXT)),
+               .call = parse_vc_action_raw_encap,
+       },
+       [ACTION_RAW_DECAP] = {
+               .name = "raw_decap",
+               .help = "decapsulation data, defined by set raw_encap",
+               .priv = PRIV_ACTION(RAW_DECAP,
+                       sizeof(struct rte_flow_action_raw_decap)),
+               .next = NEXT(NEXT_ENTRY(ACTION_NEXT)),
+               .call = parse_vc_action_raw_decap,
+       },
+       /* Top level command. */
+       [SET] = {
+               .name = "set",
+               .help = "set raw encap/decap data",
+               .type = "set raw_encap|raw_decap <pattern>",
+               .next = NEXT(NEXT_ENTRY
+                            (SET_RAW_ENCAP,
+                             SET_RAW_DECAP)),
+               .call = parse_set_init,
+       },
+       /* Sub-level commands. */
+       [SET_RAW_ENCAP] = {
+               .name = "raw_encap",
+               .help = "set raw encap data",
+               .next = NEXT(next_item),
+               .call = parse_set_raw_encap_decap,
+       },
+       [SET_RAW_DECAP] = {
+               .name = "raw_decap",
+               .help = "set raw decap data",
+               .next = NEXT(next_item),
+               .call = parse_set_raw_encap_decap,
+       }
 };
 
 /** Remove and return last entry from argument stack. */
@@ -3262,8 +3582,12 @@ parse_vc_action_rss(struct context *ctx, const struct token *token,
        if (!port_id_is_invalid(ctx->port, DISABLED_WARN) &&
            ctx->port != (portid_t)RTE_PORT_ALL) {
                struct rte_eth_dev_info info;
+               int ret2;
+
+               ret2 = rte_eth_dev_info_get(ctx->port, &info);
+               if (ret2 != 0)
+                       return ret2;
 
-               rte_eth_dev_info_get(ctx->port, &info);
                action_rss_data->conf.key_len =
                        RTE_MIN(sizeof(action_rss_data->key),
                                info.hash_key_size);
@@ -3367,6 +3691,7 @@ parse_vc_action_rss_queue(struct context *ctx, const struct token *token,
 {
        static const enum index next[] = NEXT_ENTRY(ACTION_RSS_QUEUE);
        struct action_rss_data *action_rss_data;
+       const struct arg *arg;
        int ret;
        int i;
 
@@ -3382,10 +3707,10 @@ parse_vc_action_rss_queue(struct context *ctx, const struct token *token,
        }
        if (i >= ACTION_RSS_QUEUE_NUM)
                return -1;
-       if (push_args(ctx,
-                     ARGS_ENTRY_ARB(offsetof(struct action_rss_data, queue) +
-                                    i * sizeof(action_rss_data->queue[i]),
-                                    sizeof(action_rss_data->queue[i]))))
+       arg = ARGS_ENTRY_ARB(offsetof(struct action_rss_data, queue) +
+                            i * sizeof(action_rss_data->queue[i]),
+                            sizeof(action_rss_data->queue[i]));
+       if (push_args(ctx, arg))
                return -1;
        ret = parse_int(ctx, token, str, len, NULL, 0);
        if (ret < 0) {
@@ -3482,9 +3807,9 @@ parse_vc_action_vxlan_encap(struct context *ctx, const struct token *token,
                .item_vxlan.flags = 0,
        };
        memcpy(action_vxlan_encap_data->item_eth.dst.addr_bytes,
-              vxlan_encap_conf.eth_dst, ETHER_ADDR_LEN);
+              vxlan_encap_conf.eth_dst, RTE_ETHER_ADDR_LEN);
        memcpy(action_vxlan_encap_data->item_eth.src.addr_bytes,
-              vxlan_encap_conf.eth_src, ETHER_ADDR_LEN);
+              vxlan_encap_conf.eth_src, RTE_ETHER_ADDR_LEN);
        if (!vxlan_encap_conf.select_ipv4) {
                memcpy(&action_vxlan_encap_data->item_ipv6.hdr.src_addr,
                       &vxlan_encap_conf.ipv6_src,
@@ -3501,6 +3826,38 @@ parse_vc_action_vxlan_encap(struct context *ctx, const struct token *token,
        if (!vxlan_encap_conf.select_vlan)
                action_vxlan_encap_data->items[1].type =
                        RTE_FLOW_ITEM_TYPE_VOID;
+       if (vxlan_encap_conf.select_tos_ttl) {
+               if (vxlan_encap_conf.select_ipv4) {
+                       static struct rte_flow_item_ipv4 ipv4_mask_tos;
+
+                       memcpy(&ipv4_mask_tos, &rte_flow_item_ipv4_mask,
+                              sizeof(ipv4_mask_tos));
+                       ipv4_mask_tos.hdr.type_of_service = 0xff;
+                       ipv4_mask_tos.hdr.time_to_live = 0xff;
+                       action_vxlan_encap_data->item_ipv4.hdr.type_of_service =
+                                       vxlan_encap_conf.ip_tos;
+                       action_vxlan_encap_data->item_ipv4.hdr.time_to_live =
+                                       vxlan_encap_conf.ip_ttl;
+                       action_vxlan_encap_data->items[2].mask =
+                                                       &ipv4_mask_tos;
+               } else {
+                       static struct rte_flow_item_ipv6 ipv6_mask_tos;
+
+                       memcpy(&ipv6_mask_tos, &rte_flow_item_ipv6_mask,
+                              sizeof(ipv6_mask_tos));
+                       ipv6_mask_tos.hdr.vtc_flow |=
+                               RTE_BE32(0xfful << RTE_IPV6_HDR_TC_SHIFT);
+                       ipv6_mask_tos.hdr.hop_limits = 0xff;
+                       action_vxlan_encap_data->item_ipv6.hdr.vtc_flow |=
+                               rte_cpu_to_be_32
+                                       ((uint32_t)vxlan_encap_conf.ip_tos <<
+                                        RTE_IPV6_HDR_TC_SHIFT);
+                       action_vxlan_encap_data->item_ipv6.hdr.hop_limits =
+                                       vxlan_encap_conf.ip_ttl;
+                       action_vxlan_encap_data->items[2].mask =
+                                                       &ipv6_mask_tos;
+               }
+       }
        memcpy(action_vxlan_encap_data->item_vxlan.vni, vxlan_encap_conf.vni,
               RTE_DIM(vxlan_encap_conf.vni));
        action->conf = &action_vxlan_encap_data->conf;
@@ -3573,9 +3930,9 @@ parse_vc_action_nvgre_encap(struct context *ctx, const struct token *token,
                .item_nvgre.flow_id = 0,
        };
        memcpy(action_nvgre_encap_data->item_eth.dst.addr_bytes,
-              nvgre_encap_conf.eth_dst, ETHER_ADDR_LEN);
+              nvgre_encap_conf.eth_dst, RTE_ETHER_ADDR_LEN);
        memcpy(action_nvgre_encap_data->item_eth.src.addr_bytes,
-              nvgre_encap_conf.eth_src, ETHER_ADDR_LEN);
+              nvgre_encap_conf.eth_src, RTE_ETHER_ADDR_LEN);
        if (!nvgre_encap_conf.select_ipv4) {
                memcpy(&action_nvgre_encap_data->item_ipv6.hdr.src_addr,
                       &nvgre_encap_conf.ipv6_src,
@@ -3637,22 +3994,22 @@ parse_vc_action_l2_encap(struct context *ctx, const struct token *token,
        };
        header = action_encap_data->data;
        if (l2_encap_conf.select_vlan)
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_VLAN);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_VLAN);
        else if (l2_encap_conf.select_ipv4)
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_IPv4);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
        else
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_IPv6);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
        memcpy(eth.dst.addr_bytes,
-              l2_encap_conf.eth_dst, ETHER_ADDR_LEN);
+              l2_encap_conf.eth_dst, RTE_ETHER_ADDR_LEN);
        memcpy(eth.src.addr_bytes,
-              l2_encap_conf.eth_src, ETHER_ADDR_LEN);
+              l2_encap_conf.eth_src, RTE_ETHER_ADDR_LEN);
        memcpy(header, &eth, sizeof(eth));
        header += sizeof(eth);
        if (l2_encap_conf.select_vlan) {
                if (l2_encap_conf.select_ipv4)
-                       vlan.inner_type = rte_cpu_to_be_16(ETHER_TYPE_IPv4);
+                       vlan.inner_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
                else
-                       vlan.inner_type = rte_cpu_to_be_16(ETHER_TYPE_IPv6);
+                       vlan.inner_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
                memcpy(header, &vlan, sizeof(vlan));
                header += sizeof(vlan);
        }
@@ -3701,7 +4058,7 @@ parse_vc_action_l2_decap(struct context *ctx, const struct token *token,
        };
        header = action_decap_data->data;
        if (l2_decap_conf.select_vlan)
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_VLAN);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_VLAN);
        memcpy(header, &eth, sizeof(eth));
        header += sizeof(eth);
        if (l2_decap_conf.select_vlan) {
@@ -3735,11 +4092,14 @@ parse_vc_action_mplsogre_encap(struct context *ctx, const struct token *token,
                        .src_addr = mplsogre_encap_conf.ipv4_src,
                        .dst_addr = mplsogre_encap_conf.ipv4_dst,
                        .next_proto_id = IPPROTO_GRE,
+                       .version_ihl = RTE_IPV4_VHL_DEF,
+                       .time_to_live = IPDEFTTL,
                },
        };
        struct rte_flow_item_ipv6 ipv6 = {
                .hdr =  {
                        .proto = IPPROTO_GRE,
+                       .hop_limits = IPDEFTTL,
                },
        };
        struct rte_flow_item_gre gre = {
@@ -3772,22 +4132,22 @@ parse_vc_action_mplsogre_encap(struct context *ctx, const struct token *token,
        };
        header = action_encap_data->data;
        if (mplsogre_encap_conf.select_vlan)
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_VLAN);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_VLAN);
        else if (mplsogre_encap_conf.select_ipv4)
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_IPv4);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
        else
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_IPv6);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
        memcpy(eth.dst.addr_bytes,
-              mplsogre_encap_conf.eth_dst, ETHER_ADDR_LEN);
+              mplsogre_encap_conf.eth_dst, RTE_ETHER_ADDR_LEN);
        memcpy(eth.src.addr_bytes,
-              mplsogre_encap_conf.eth_src, ETHER_ADDR_LEN);
+              mplsogre_encap_conf.eth_src, RTE_ETHER_ADDR_LEN);
        memcpy(header, &eth, sizeof(eth));
        header += sizeof(eth);
        if (mplsogre_encap_conf.select_vlan) {
                if (mplsogre_encap_conf.select_ipv4)
-                       vlan.inner_type = rte_cpu_to_be_16(ETHER_TYPE_IPv4);
+                       vlan.inner_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
                else
-                       vlan.inner_type = rte_cpu_to_be_16(ETHER_TYPE_IPv6);
+                       vlan.inner_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
                memcpy(header, &vlan, sizeof(vlan));
                header += sizeof(vlan);
        }
@@ -3867,22 +4227,22 @@ parse_vc_action_mplsogre_decap(struct context *ctx, const struct token *token,
        };
        header = action_decap_data->data;
        if (mplsogre_decap_conf.select_vlan)
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_VLAN);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_VLAN);
        else if (mplsogre_encap_conf.select_ipv4)
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_IPv4);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
        else
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_IPv6);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
        memcpy(eth.dst.addr_bytes,
-              mplsogre_encap_conf.eth_dst, ETHER_ADDR_LEN);
+              mplsogre_encap_conf.eth_dst, RTE_ETHER_ADDR_LEN);
        memcpy(eth.src.addr_bytes,
-              mplsogre_encap_conf.eth_src, ETHER_ADDR_LEN);
+              mplsogre_encap_conf.eth_src, RTE_ETHER_ADDR_LEN);
        memcpy(header, &eth, sizeof(eth));
        header += sizeof(eth);
        if (mplsogre_encap_conf.select_vlan) {
                if (mplsogre_encap_conf.select_ipv4)
-                       vlan.inner_type = rte_cpu_to_be_16(ETHER_TYPE_IPv4);
+                       vlan.inner_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
                else
-                       vlan.inner_type = rte_cpu_to_be_16(ETHER_TYPE_IPv6);
+                       vlan.inner_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
                memcpy(header, &vlan, sizeof(vlan));
                header += sizeof(vlan);
        }
@@ -3923,11 +4283,14 @@ parse_vc_action_mplsoudp_encap(struct context *ctx, const struct token *token,
                        .src_addr = mplsoudp_encap_conf.ipv4_src,
                        .dst_addr = mplsoudp_encap_conf.ipv4_dst,
                        .next_proto_id = IPPROTO_UDP,
+                       .version_ihl = RTE_IPV4_VHL_DEF,
+                       .time_to_live = IPDEFTTL,
                },
        };
        struct rte_flow_item_ipv6 ipv6 = {
                .hdr =  {
                        .proto = IPPROTO_UDP,
+                       .hop_limits = IPDEFTTL,
                },
        };
        struct rte_flow_item_udp udp = {
@@ -3963,22 +4326,22 @@ parse_vc_action_mplsoudp_encap(struct context *ctx, const struct token *token,
        };
        header = action_encap_data->data;
        if (mplsoudp_encap_conf.select_vlan)
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_VLAN);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_VLAN);
        else if (mplsoudp_encap_conf.select_ipv4)
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_IPv4);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
        else
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_IPv6);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
        memcpy(eth.dst.addr_bytes,
-              mplsoudp_encap_conf.eth_dst, ETHER_ADDR_LEN);
+              mplsoudp_encap_conf.eth_dst, RTE_ETHER_ADDR_LEN);
        memcpy(eth.src.addr_bytes,
-              mplsoudp_encap_conf.eth_src, ETHER_ADDR_LEN);
+              mplsoudp_encap_conf.eth_src, RTE_ETHER_ADDR_LEN);
        memcpy(header, &eth, sizeof(eth));
        header += sizeof(eth);
        if (mplsoudp_encap_conf.select_vlan) {
                if (mplsoudp_encap_conf.select_ipv4)
-                       vlan.inner_type = rte_cpu_to_be_16(ETHER_TYPE_IPv4);
+                       vlan.inner_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
                else
-                       vlan.inner_type = rte_cpu_to_be_16(ETHER_TYPE_IPv6);
+                       vlan.inner_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
                memcpy(header, &vlan, sizeof(vlan));
                header += sizeof(vlan);
        }
@@ -4060,22 +4423,22 @@ parse_vc_action_mplsoudp_decap(struct context *ctx, const struct token *token,
        };
        header = action_decap_data->data;
        if (mplsoudp_decap_conf.select_vlan)
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_VLAN);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_VLAN);
        else if (mplsoudp_encap_conf.select_ipv4)
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_IPv4);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
        else
-               eth.type = rte_cpu_to_be_16(ETHER_TYPE_IPv6);
+               eth.type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
        memcpy(eth.dst.addr_bytes,
-              mplsoudp_encap_conf.eth_dst, ETHER_ADDR_LEN);
+              mplsoudp_encap_conf.eth_dst, RTE_ETHER_ADDR_LEN);
        memcpy(eth.src.addr_bytes,
-              mplsoudp_encap_conf.eth_src, ETHER_ADDR_LEN);
+              mplsoudp_encap_conf.eth_src, RTE_ETHER_ADDR_LEN);
        memcpy(header, &eth, sizeof(eth));
        header += sizeof(eth);
        if (mplsoudp_encap_conf.select_vlan) {
                if (mplsoudp_encap_conf.select_ipv4)
-                       vlan.inner_type = rte_cpu_to_be_16(ETHER_TYPE_IPv4);
+                       vlan.inner_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
                else
-                       vlan.inner_type = rte_cpu_to_be_16(ETHER_TYPE_IPv6);
+                       vlan.inner_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
                memcpy(header, &vlan, sizeof(vlan));
                header += sizeof(vlan);
        }
@@ -4097,6 +4460,75 @@ parse_vc_action_mplsoudp_decap(struct context *ctx, const struct token *token,
        return ret;
 }
 
+static int
+parse_vc_action_raw_encap(struct context *ctx, const struct token *token,
+                         const char *str, unsigned int len, void *buf,
+                         unsigned int size)
+{
+       struct buffer *out = buf;
+       struct rte_flow_action *action;
+       struct rte_flow_action_raw_encap *action_raw_encap_conf = NULL;
+       uint8_t *data = NULL;
+       int ret;
+
+       ret = parse_vc(ctx, token, str, len, buf, size);
+       if (ret < 0)
+               return ret;
+       /* Nothing else to do if there is no buffer. */
+       if (!out)
+               return ret;
+       if (!out->args.vc.actions_n)
+               return -1;
+       action = &out->args.vc.actions[out->args.vc.actions_n - 1];
+       /* Point to selected object. */
+       ctx->object = out->args.vc.data;
+       ctx->objmask = NULL;
+       /* Copy the headers to the buffer. */
+       action_raw_encap_conf = ctx->object;
+       /* data stored from tail of data buffer */
+       data = (uint8_t *)&(raw_encap_conf.data) +
+               ACTION_RAW_ENCAP_MAX_DATA - raw_encap_conf.size;
+       action_raw_encap_conf->data = data;
+       action_raw_encap_conf->preserve = NULL;
+       action_raw_encap_conf->size = raw_encap_conf.size;
+       action->conf = action_raw_encap_conf;
+       return ret;
+}
+
+static int
+parse_vc_action_raw_decap(struct context *ctx, const struct token *token,
+                         const char *str, unsigned int len, void *buf,
+                         unsigned int size)
+{
+       struct buffer *out = buf;
+       struct rte_flow_action *action;
+       struct rte_flow_action_raw_decap *action_raw_decap_conf = NULL;
+       uint8_t *data = NULL;
+       int ret;
+
+       ret = parse_vc(ctx, token, str, len, buf, size);
+       if (ret < 0)
+               return ret;
+       /* Nothing else to do if there is no buffer. */
+       if (!out)
+               return ret;
+       if (!out->args.vc.actions_n)
+               return -1;
+       action = &out->args.vc.actions[out->args.vc.actions_n - 1];
+       /* Point to selected object. */
+       ctx->object = out->args.vc.data;
+       ctx->objmask = NULL;
+       /* Copy the headers to the buffer. */
+       action_raw_decap_conf = ctx->object;
+       /* data stored from tail of data buffer */
+       data = (uint8_t *)&(raw_decap_conf.data) +
+               ACTION_RAW_ENCAP_MAX_DATA - raw_decap_conf.size;
+       action_raw_decap_conf->data = data;
+       action_raw_decap_conf->size = raw_decap_conf.size;
+       action->conf = action_raw_decap_conf;
+       return ret;
+}
+
 /** Parse tokens for destroy command. */
 static int
 parse_destroy(struct context *ctx, const struct token *token,
@@ -4443,6 +4875,121 @@ error:
        return -1;
 }
 
+static int
+parse_hex_string(const char *src, uint8_t *dst, uint32_t *size)
+{
+       char *c = NULL;
+       uint32_t i, len;
+       char tmp[3];
+
+       /* Check input parameters */
+       if ((src == NULL) ||
+               (dst == NULL) ||
+               (size == NULL) ||
+               (*size == 0))
+               return -1;
+
+       /* Convert chars to bytes */
+       for (i = 0, len = 0; i < *size; i += 2) {
+               snprintf(tmp, 3, "%s", src + i);
+               dst[len++] = strtoul(tmp, &c, 16);
+               if (*c != 0) {
+                       len--;
+                       dst[len] = 0;
+                       *size = len;
+                       return -1;
+               }
+       }
+       dst[len] = 0;
+       *size = len;
+
+       return 0;
+}
+
+static int
+parse_hex(struct context *ctx, const struct token *token,
+               const char *str, unsigned int len,
+               void *buf, unsigned int size)
+{
+       const struct arg *arg_data = pop_args(ctx);
+       const struct arg *arg_len = pop_args(ctx);
+       const struct arg *arg_addr = pop_args(ctx);
+       char tmp[16]; /* Ought to be enough. */
+       int ret;
+       unsigned int hexlen = len;
+       unsigned int length = 256;
+       uint8_t hex_tmp[length];
+
+       /* Arguments are expected. */
+       if (!arg_data)
+               return -1;
+       if (!arg_len) {
+               push_args(ctx, arg_data);
+               return -1;
+       }
+       if (!arg_addr) {
+               push_args(ctx, arg_len);
+               push_args(ctx, arg_data);
+               return -1;
+       }
+       size = arg_data->size;
+       /* Bit-mask fill is not supported. */
+       if (arg_data->mask)
+               goto error;
+       if (!ctx->object)
+               return len;
+
+       /* translate bytes string to array. */
+       if (str[0] == '0' && ((str[1] == 'x') ||
+                       (str[1] == 'X'))) {
+               str += 2;
+               hexlen -= 2;
+       }
+       if (hexlen > length)
+               return -1;
+       ret = parse_hex_string(str, hex_tmp, &hexlen);
+       if (ret < 0)
+               goto error;
+       /* Let parse_int() fill length information first. */
+       ret = snprintf(tmp, sizeof(tmp), "%u", hexlen);
+       if (ret < 0)
+               goto error;
+       push_args(ctx, arg_len);
+       ret = parse_int(ctx, token, tmp, ret, NULL, 0);
+       if (ret < 0) {
+               pop_args(ctx);
+               goto error;
+       }
+       buf = (uint8_t *)ctx->object + arg_data->offset;
+       /* Output buffer is not necessarily NUL-terminated. */
+       memcpy(buf, hex_tmp, hexlen);
+       memset((uint8_t *)buf + hexlen, 0x00, size - hexlen);
+       if (ctx->objmask)
+               memset((uint8_t *)ctx->objmask + arg_data->offset,
+                                       0xff, hexlen);
+       /* Save address if requested. */
+       if (arg_addr->size) {
+               memcpy((uint8_t *)ctx->object + arg_addr->offset,
+                      (void *[]){
+                       (uint8_t *)ctx->object + arg_data->offset
+                      },
+                      arg_addr->size);
+               if (ctx->objmask)
+                       memcpy((uint8_t *)ctx->objmask + arg_addr->offset,
+                              (void *[]){
+                               (uint8_t *)ctx->objmask + arg_data->offset
+                              },
+                              arg_addr->size);
+       }
+       return len;
+error:
+       push_args(ctx, arg_addr);
+       push_args(ctx, arg_len);
+       push_args(ctx, arg_data);
+       return -1;
+
+}
+
 /**
  * Parse a MAC address.
  *
@@ -4455,7 +5002,7 @@ parse_mac_addr(struct context *ctx, const struct token *token,
               void *buf, unsigned int size)
 {
        const struct arg *arg = pop_args(ctx);
-       struct ether_addr tmp;
+       struct rte_ether_addr tmp;
        int ret;
 
        (void)token;
@@ -4638,6 +5185,73 @@ parse_port(struct context *ctx, const struct token *token,
        return ret;
 }
 
+/** Parse set command, initialize output buffer for subsequent tokens. */
+static int
+parse_set_raw_encap_decap(struct context *ctx, const struct token *token,
+                         const char *str, unsigned int len,
+                         void *buf, unsigned int size)
+{
+       struct buffer *out = buf;
+
+       /* Token name must match. */
+       if (parse_default(ctx, token, str, len, NULL, 0) < 0)
+               return -1;
+       /* Nothing else to do if there is no buffer. */
+       if (!out)
+               return len;
+       /* Make sure buffer is large enough. */
+       if (size < sizeof(*out))
+               return -1;
+       ctx->objdata = 0;
+       ctx->objmask = NULL;
+       if (!out->command)
+               return -1;
+       out->command = ctx->curr;
+       return len;
+}
+
+/**
+ * Parse set raw_encap/raw_decap command,
+ * initialize output buffer for subsequent tokens.
+ */
+static int
+parse_set_init(struct context *ctx, const struct token *token,
+              const char *str, unsigned int len,
+              void *buf, unsigned int size)
+{
+       struct buffer *out = buf;
+
+       /* Token name must match. */
+       if (parse_default(ctx, token, str, len, NULL, 0) < 0)
+               return -1;
+       /* Nothing else to do if there is no buffer. */
+       if (!out)
+               return len;
+       /* Make sure buffer is large enough. */
+       if (size < sizeof(*out))
+               return -1;
+       /* Initialize buffer. */
+       memset(out, 0x00, sizeof(*out));
+       memset((uint8_t *)out + sizeof(*out), 0x22, size - sizeof(*out));
+       ctx->objdata = 0;
+       ctx->object = out;
+       ctx->objmask = NULL;
+       if (!out->command) {
+               if (ctx->curr != SET)
+                       return -1;
+               if (sizeof(*out) > size)
+                       return -1;
+               out->command = ctx->curr;
+               out->args.vc.data = (uint8_t *)out + size;
+               /* All we need is pattern */
+               out->args.vc.pattern =
+                       (void *)RTE_ALIGN_CEIL((uintptr_t)(out + 1),
+                                              sizeof(double));
+               ctx->object = out->args.vc.pattern;
+       }
+       return len;
+}
+
 /** No completion. */
 static int
 comp_none(struct context *ctx, const struct token *token,
@@ -4662,7 +5276,7 @@ comp_boolean(struct context *ctx, const struct token *token,
        (void)token;
        for (i = 0; boolean_name[i]; ++i)
                if (buf && i == ent)
-                       return snprintf(buf, size, "%s", boolean_name[i]);
+                       return strlcpy(buf, boolean_name[i], size);
        if (buf)
                return -1;
        return i;
@@ -4679,8 +5293,8 @@ comp_action(struct context *ctx, const struct token *token,
        (void)token;
        for (i = 0; next_action[i]; ++i)
                if (buf && i == ent)
-                       return snprintf(buf, size, "%s",
-                                       token_list[next_action[i]].name);
+                       return strlcpy(buf, token_list[next_action[i]].name,
+                                      size);
        if (buf)
                return -1;
        return i;
@@ -4744,7 +5358,7 @@ comp_vc_action_rss_type(struct context *ctx, const struct token *token,
        if (!buf)
                return i + 1;
        if (ent < i)
-               return snprintf(buf, size, "%s", rss_type_table[ent].str);
+               return strlcpy(buf, rss_type_table[ent].str, size);
        if (ent == i)
                return snprintf(buf, size, "end");
        return -1;
@@ -4771,6 +5385,7 @@ static struct context cmd_flow_context;
 
 /** Global parser instance (cmdline API). */
 cmdline_parse_inst_t cmd_flow;
+cmdline_parse_inst_t cmd_set_raw;
 
 /** Initialize context. */
 static void
@@ -4929,7 +5544,7 @@ cmd_flow_complete_get_elt(cmdline_parse_token_hdr_t *hdr, int index,
        if (index >= i)
                return -1;
        token = &token_list[list[index]];
-       snprintf(dst, size, "%s", token->name);
+       strlcpy(dst, token->name, size);
        /* Save index for cmd_flow_get_help(). */
        ctx->prev = list[index];
        return 0;
@@ -4946,7 +5561,7 @@ cmd_flow_get_help(cmdline_parse_token_hdr_t *hdr, char *dst, unsigned int size)
        if (!size)
                return -1;
        /* Set token type and update global help with details. */
-       snprintf(dst, size, "%s", (token->type ? token->type : "TOKEN"));
+       strlcpy(dst, (token->type ? token->type : "TOKEN"), size);
        if (token->help)
                cmd_flow.help_str = token->help;
        else
@@ -5050,3 +5665,317 @@ cmdline_parse_inst_t cmd_flow = {
                NULL,
        }, /**< Tokens are returned by cmd_flow_tok(). */
 };
+
+/** set cmd facility. Reuse cmd flow's infrastructure as much as possible. */
+
+static void
+update_fields(uint8_t *buf, struct rte_flow_item *item, uint16_t next_proto)
+{
+       struct rte_flow_item_ipv4 *ipv4;
+       struct rte_flow_item_eth *eth;
+       struct rte_flow_item_ipv6 *ipv6;
+       struct rte_flow_item_vxlan *vxlan;
+       struct rte_flow_item_vxlan_gpe *gpe;
+       struct rte_flow_item_nvgre *nvgre;
+       uint32_t ipv6_vtc_flow;
+
+       switch (item->type) {
+       case RTE_FLOW_ITEM_TYPE_ETH:
+               eth = (struct rte_flow_item_eth *)buf;
+               if (next_proto)
+                       eth->type = rte_cpu_to_be_16(next_proto);
+               break;
+       case RTE_FLOW_ITEM_TYPE_IPV4:
+               ipv4 = (struct rte_flow_item_ipv4 *)buf;
+               ipv4->hdr.version_ihl = 0x45;
+               ipv4->hdr.next_proto_id = (uint8_t)next_proto;
+               break;
+       case RTE_FLOW_ITEM_TYPE_IPV6:
+               ipv6 = (struct rte_flow_item_ipv6 *)buf;
+               ipv6->hdr.proto = (uint8_t)next_proto;
+               ipv6_vtc_flow = rte_be_to_cpu_32(ipv6->hdr.vtc_flow);
+               ipv6_vtc_flow &= 0x0FFFFFFF; /*< reset version bits. */
+               ipv6_vtc_flow |= 0x60000000; /*< set ipv6 version. */
+               ipv6->hdr.vtc_flow = rte_cpu_to_be_32(ipv6_vtc_flow);
+               break;
+       case RTE_FLOW_ITEM_TYPE_VXLAN:
+               vxlan = (struct rte_flow_item_vxlan *)buf;
+               vxlan->flags = 0x08;
+               break;
+       case RTE_FLOW_ITEM_TYPE_VXLAN_GPE:
+               gpe = (struct rte_flow_item_vxlan_gpe *)buf;
+               gpe->flags = 0x0C;
+               break;
+       case RTE_FLOW_ITEM_TYPE_NVGRE:
+               nvgre = (struct rte_flow_item_nvgre *)buf;
+               nvgre->protocol = rte_cpu_to_be_16(0x6558);
+               nvgre->c_k_s_rsvd0_ver = rte_cpu_to_be_16(0x2000);
+               break;
+       default:
+               break;
+       }
+}
+
+/** Helper of get item's default mask. */
+static const void *
+flow_item_default_mask(const struct rte_flow_item *item)
+{
+       const void *mask = NULL;
+       static rte_be32_t gre_key_default_mask = RTE_BE32(UINT32_MAX);
+
+       switch (item->type) {
+       case RTE_FLOW_ITEM_TYPE_ANY:
+               mask = &rte_flow_item_any_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_VF:
+               mask = &rte_flow_item_vf_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_PORT_ID:
+               mask = &rte_flow_item_port_id_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_RAW:
+               mask = &rte_flow_item_raw_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_ETH:
+               mask = &rte_flow_item_eth_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_VLAN:
+               mask = &rte_flow_item_vlan_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_IPV4:
+               mask = &rte_flow_item_ipv4_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_IPV6:
+               mask = &rte_flow_item_ipv6_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_ICMP:
+               mask = &rte_flow_item_icmp_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_UDP:
+               mask = &rte_flow_item_udp_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_TCP:
+               mask = &rte_flow_item_tcp_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_SCTP:
+               mask = &rte_flow_item_sctp_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_VXLAN:
+               mask = &rte_flow_item_vxlan_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_VXLAN_GPE:
+               mask = &rte_flow_item_vxlan_gpe_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_E_TAG:
+               mask = &rte_flow_item_e_tag_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_NVGRE:
+               mask = &rte_flow_item_nvgre_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_MPLS:
+               mask = &rte_flow_item_mpls_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_GRE:
+               mask = &rte_flow_item_gre_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_GRE_KEY:
+               mask = &gre_key_default_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_META:
+               mask = &rte_flow_item_meta_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_FUZZY:
+               mask = &rte_flow_item_fuzzy_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_GTP:
+               mask = &rte_flow_item_gtp_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_ESP:
+               mask = &rte_flow_item_esp_mask;
+               break;
+       case RTE_FLOW_ITEM_TYPE_GTP_PSC:
+               mask = &rte_flow_item_gtp_psc_mask;
+               break;
+       default:
+               break;
+       }
+       return mask;
+}
+
+
+
+/** Dispatch parsed buffer to function calls. */
+static void
+cmd_set_raw_parsed(const struct buffer *in)
+{
+       uint32_t n = in->args.vc.pattern_n;
+       int i = 0;
+       struct rte_flow_item *item = NULL;
+       size_t size = 0;
+       uint8_t *data = NULL;
+       uint8_t *data_tail = NULL;
+       size_t *total_size = NULL;
+       uint16_t upper_layer = 0;
+       uint16_t proto = 0;
+
+       RTE_ASSERT(in->command == SET_RAW_ENCAP ||
+                  in->command == SET_RAW_DECAP);
+       if (in->command == SET_RAW_ENCAP) {
+               total_size = &raw_encap_conf.size;
+               data = (uint8_t *)&raw_encap_conf.data;
+       } else {
+               total_size = &raw_decap_conf.size;
+               data = (uint8_t *)&raw_decap_conf.data;
+       }
+       *total_size = 0;
+       memset(data, 0x00, ACTION_RAW_ENCAP_MAX_DATA);
+       /* process hdr from upper layer to low layer (L3/L4 -> L2). */
+       data_tail = data + ACTION_RAW_ENCAP_MAX_DATA;
+       for (i = n - 1 ; i >= 0; --i) {
+               item = in->args.vc.pattern + i;
+               if (item->spec == NULL)
+                       item->spec = flow_item_default_mask(item);
+               switch (item->type) {
+               case RTE_FLOW_ITEM_TYPE_ETH:
+                       size = sizeof(struct rte_flow_item_eth);
+                       break;
+               case RTE_FLOW_ITEM_TYPE_VLAN:
+                       size = sizeof(struct rte_flow_item_vlan);
+                       proto = RTE_ETHER_TYPE_VLAN;
+                       break;
+               case RTE_FLOW_ITEM_TYPE_IPV4:
+                       size = sizeof(struct rte_flow_item_ipv4);
+                       proto = RTE_ETHER_TYPE_IPV4;
+                       break;
+               case RTE_FLOW_ITEM_TYPE_IPV6:
+                       size = sizeof(struct rte_flow_item_ipv6);
+                       proto = RTE_ETHER_TYPE_IPV6;
+                       break;
+               case RTE_FLOW_ITEM_TYPE_UDP:
+                       size = sizeof(struct rte_flow_item_udp);
+                       proto = 0x11;
+                       break;
+               case RTE_FLOW_ITEM_TYPE_TCP:
+                       size = sizeof(struct rte_flow_item_tcp);
+                       proto = 0x06;
+                       break;
+               case RTE_FLOW_ITEM_TYPE_VXLAN:
+                       size = sizeof(struct rte_flow_item_vxlan);
+                       break;
+               case RTE_FLOW_ITEM_TYPE_VXLAN_GPE:
+                       size = sizeof(struct rte_flow_item_vxlan_gpe);
+                       break;
+               case RTE_FLOW_ITEM_TYPE_GRE:
+                       size = sizeof(struct rte_flow_item_gre);
+                       proto = 0x2F;
+                       break;
+               case RTE_FLOW_ITEM_TYPE_GRE_KEY:
+                       size = sizeof(rte_be32_t);
+                       break;
+               case RTE_FLOW_ITEM_TYPE_MPLS:
+                       size = sizeof(struct rte_flow_item_mpls);
+                       break;
+               case RTE_FLOW_ITEM_TYPE_NVGRE:
+                       size = sizeof(struct rte_flow_item_nvgre);
+                       proto = 0x2F;
+                       break;
+               default:
+                       printf("Error - Not supported item\n");
+                       *total_size = 0;
+                       memset(data, 0x00, ACTION_RAW_ENCAP_MAX_DATA);
+                       return;
+               }
+               *total_size += size;
+               rte_memcpy(data_tail - (*total_size), item->spec, size);
+               /* update some fields which cannot be set by cmdline */
+               update_fields((data_tail - (*total_size)), item,
+                             upper_layer);
+               upper_layer = proto;
+       }
+       if (verbose_level & 0x1)
+               printf("total data size is %zu\n", (*total_size));
+       RTE_ASSERT((*total_size) <= ACTION_RAW_ENCAP_MAX_DATA);
+}
+
+/** Populate help strings for current token (cmdline API). */
+static int
+cmd_set_raw_get_help(cmdline_parse_token_hdr_t *hdr, char *dst,
+                    unsigned int size)
+{
+       struct context *ctx = &cmd_flow_context;
+       const struct token *token = &token_list[ctx->prev];
+
+       (void)hdr;
+       if (!size)
+               return -1;
+       /* Set token type and update global help with details. */
+       snprintf(dst, size, "%s", (token->type ? token->type : "TOKEN"));
+       if (token->help)
+               cmd_set_raw.help_str = token->help;
+       else
+               cmd_set_raw.help_str = token->name;
+       return 0;
+}
+
+/** Token definition template (cmdline API). */
+static struct cmdline_token_hdr cmd_set_raw_token_hdr = {
+       .ops = &(struct cmdline_token_ops){
+               .parse = cmd_flow_parse,
+               .complete_get_nb = cmd_flow_complete_get_nb,
+               .complete_get_elt = cmd_flow_complete_get_elt,
+               .get_help = cmd_set_raw_get_help,
+       },
+       .offset = 0,
+};
+
+/** Populate the next dynamic token. */
+static void
+cmd_set_raw_tok(cmdline_parse_token_hdr_t **hdr,
+            cmdline_parse_token_hdr_t **hdr_inst)
+{
+       struct context *ctx = &cmd_flow_context;
+
+       /* Always reinitialize context before requesting the first token. */
+       if (!(hdr_inst - cmd_set_raw.tokens)) {
+               cmd_flow_context_init(ctx);
+               ctx->curr = START_SET;
+       }
+       /* Return NULL when no more tokens are expected. */
+       if (!ctx->next_num && (ctx->curr != START_SET)) {
+               *hdr = NULL;
+               return;
+       }
+       /* Determine if command should end here. */
+       if (ctx->eol && ctx->last && ctx->next_num) {
+               const enum index *list = ctx->next[ctx->next_num - 1];
+               int i;
+
+               for (i = 0; list[i]; ++i) {
+                       if (list[i] != END)
+                               continue;
+                       *hdr = NULL;
+                       return;
+               }
+       }
+       *hdr = &cmd_set_raw_token_hdr;
+}
+
+/** Token generator and output processing callback (cmdline API). */
+static void
+cmd_set_raw_cb(void *arg0, struct cmdline *cl, void *arg2)
+{
+       if (cl == NULL)
+               cmd_set_raw_tok(arg0, arg2);
+       else
+               cmd_set_raw_parsed(arg0);
+}
+
+/** Global parser instance (cmdline API). */
+cmdline_parse_inst_t cmd_set_raw = {
+       .f = cmd_set_raw_cb,
+       .data = NULL, /**< Unused. */
+       .help_str = NULL, /**< Updated by cmd_flow_get_help(). */
+       .tokens = {
+               NULL,
+       }, /**< Tokens are returned by cmd_flow_tok(). */
+};