app/testpmd: support GENEVE header option length
authorViacheslav Ovsiienko <viacheslavo@nvidia.com>
Sun, 17 Jan 2021 10:21:17 +0000 (12:21 +0200)
committerFerruh Yigit <ferruh.yigit@intel.com>
Tue, 19 Jan 2021 02:30:15 +0000 (03:30 +0100)
The patch adds the GENEVE rte flow option length support to
command line interpreter. The flow command with GENEVE
option items looks like:

    flow create 0 ingress pattern eth / ipv4 / udp / geneve vni
    is 100 optlen is 2 / end actions drop / end

The option length should be specified in 32-bit words, this
value specifies the all options length in the GENEVE header.

Signed-off-by: Viacheslav Ovsiienko <viacheslavo@nvidia.com>
Acked-by: Ori Kam <orika@nvidia.com>
app/test-pmd/cmdline_flow.c

index 4e9b766..9940d4f 100644 (file)
@@ -223,6 +223,7 @@ enum index {
        ITEM_GENEVE,
        ITEM_GENEVE_VNI,
        ITEM_GENEVE_PROTO,
+       ITEM_GENEVE_OPTLEN,
        ITEM_VXLAN_GPE,
        ITEM_VXLAN_GPE_VNI,
        ITEM_ARP_ETH_IPV4,
@@ -1101,6 +1102,7 @@ static const enum index item_gtp[] = {
 static const enum index item_geneve[] = {
        ITEM_GENEVE_VNI,
        ITEM_GENEVE_PROTO,
+       ITEM_GENEVE_OPTLEN,
        ITEM_NEXT,
        ZERO,
 };
@@ -2807,6 +2809,14 @@ static const struct token token_list[] = {
                .args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_geneve,
                                             protocol)),
        },
+       [ITEM_GENEVE_OPTLEN] = {
+               .name = "optlen",
+               .help = "GENEVE options length in dwords",
+               .next = NEXT(item_geneve, NEXT_ENTRY(UNSIGNED), item_param),
+               .args = ARGS(ARGS_ENTRY_MASK_HTON(struct rte_flow_item_geneve,
+                                                 ver_opt_len_o_c_rsvd0,
+                                                 "\x3f\x00")),
+       },
        [ITEM_VXLAN_GPE] = {
                .name = "vxlan-gpe",
                .help = "match VXLAN-GPE header",