net/mlx5: add flow IPv6 item
authorNelio Laranjeiro <nelio.laranjeiro@6wind.com>
Thu, 12 Jul 2018 09:30:55 +0000 (11:30 +0200)
committerShahaf Shuler <shahafs@mellanox.com>
Thu, 12 Jul 2018 10:10:02 +0000 (12:10 +0200)
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>
drivers/net/mlx5/mlx5_flow.c

index c05b849..513f70d 100644 (file)
@@ -593,6 +593,118 @@ mlx5_flow_item_ipv4(const struct rte_flow_item *item, struct rte_flow *flow,
        return size;
 }
 
+/**
+ * Convert the @p item into a Verbs specification after ensuring the NIC
+ * will understand and process it correctly.
+ * If the necessary size for the conversion is greater than the @p flow_size,
+ * nothing is written in @p flow, the validation is still performed.
+ *
+ * @param[in] item
+ *   Item specification.
+ * @param[in, out] flow
+ *   Pointer to flow structure.
+ * @param[in] flow_size
+ *   Size in bytes of the available space in @p flow, if too small, nothing is
+ *   written.
+ * @param[out] error
+ *   Pointer to error structure.
+ *
+ * @return
+ *   On success the number of bytes consumed/necessary, if the returned value
+ *   is lesser or equal to @p flow_size, the @p item has fully been converted,
+ *   otherwise another call with this returned memory size should be done.
+ *   On error, a negative errno value is returned and rte_errno is set.
+ */
+static int
+mlx5_flow_item_ipv6(const struct rte_flow_item *item, struct rte_flow *flow,
+                   const size_t flow_size, struct rte_flow_error *error)
+{
+       const struct rte_flow_item_ipv6 *spec = item->spec;
+       const struct rte_flow_item_ipv6 *mask = item->mask;
+       const struct rte_flow_item_ipv6 nic_mask = {
+               .hdr = {
+                       .src_addr =
+                               "\xff\xff\xff\xff\xff\xff\xff\xff"
+                               "\xff\xff\xff\xff\xff\xff\xff\xff",
+                       .dst_addr =
+                               "\xff\xff\xff\xff\xff\xff\xff\xff"
+                               "\xff\xff\xff\xff\xff\xff\xff\xff",
+                       .vtc_flow = RTE_BE32(0xffffffff),
+                       .proto = 0xff,
+                       .hop_limits = 0xff,
+               },
+       };
+       unsigned int size = sizeof(struct ibv_flow_spec_ipv6);
+       struct ibv_flow_spec_ipv6 ipv6 = {
+               .type = IBV_FLOW_SPEC_IPV6,
+               .size = size,
+       };
+       int ret;
+
+       if (flow->layers & MLX5_FLOW_LAYER_OUTER_L3)
+               return rte_flow_error_set(error, ENOTSUP,
+                                         RTE_FLOW_ERROR_TYPE_ITEM,
+                                         item,
+                                         "multiple L3 layers not supported");
+       else if (flow->layers & MLX5_FLOW_LAYER_OUTER_L4)
+               return rte_flow_error_set(error, ENOTSUP,
+                                         RTE_FLOW_ERROR_TYPE_ITEM,
+                                         item,
+                                         "L3 cannot follow an L4 layer.");
+       if (!mask)
+               mask = &rte_flow_item_ipv6_mask;
+       ret = mlx5_flow_item_acceptable
+               (item, (const uint8_t *)mask,
+                (const uint8_t *)&nic_mask,
+                sizeof(struct rte_flow_item_ipv6), error);
+       if (ret < 0)
+               return ret;
+       flow->layers |= MLX5_FLOW_LAYER_OUTER_L3_IPV6;
+       if (size > flow_size)
+               return size;
+       if (spec) {
+               unsigned int i;
+               uint32_t vtc_flow_val;
+               uint32_t vtc_flow_mask;
+
+               memcpy(&ipv6.val.src_ip, spec->hdr.src_addr,
+                      RTE_DIM(ipv6.val.src_ip));
+               memcpy(&ipv6.val.dst_ip, spec->hdr.dst_addr,
+                      RTE_DIM(ipv6.val.dst_ip));
+               memcpy(&ipv6.mask.src_ip, mask->hdr.src_addr,
+                      RTE_DIM(ipv6.mask.src_ip));
+               memcpy(&ipv6.mask.dst_ip, mask->hdr.dst_addr,
+                      RTE_DIM(ipv6.mask.dst_ip));
+               vtc_flow_val = rte_be_to_cpu_32(spec->hdr.vtc_flow);
+               vtc_flow_mask = rte_be_to_cpu_32(mask->hdr.vtc_flow);
+               ipv6.val.flow_label =
+                       rte_cpu_to_be_32((vtc_flow_val & IPV6_HDR_FL_MASK) >>
+                                        IPV6_HDR_FL_SHIFT);
+               ipv6.val.traffic_class = (vtc_flow_val & IPV6_HDR_TC_MASK) >>
+                                        IPV6_HDR_TC_SHIFT;
+               ipv6.val.next_hdr = spec->hdr.proto;
+               ipv6.val.hop_limit = spec->hdr.hop_limits;
+               ipv6.mask.flow_label =
+                       rte_cpu_to_be_32((vtc_flow_mask & IPV6_HDR_FL_MASK) >>
+                                        IPV6_HDR_FL_SHIFT);
+               ipv6.mask.traffic_class = (vtc_flow_mask & IPV6_HDR_TC_MASK) >>
+                                         IPV6_HDR_TC_SHIFT;
+               ipv6.mask.next_hdr = mask->hdr.proto;
+               ipv6.mask.hop_limit = mask->hdr.hop_limits;
+               /* Remove unwanted bits from values. */
+               for (i = 0; i < RTE_DIM(ipv6.val.src_ip); ++i) {
+                       ipv6.val.src_ip[i] &= ipv6.mask.src_ip[i];
+                       ipv6.val.dst_ip[i] &= ipv6.mask.dst_ip[i];
+               }
+               ipv6.val.flow_label &= ipv6.mask.flow_label;
+               ipv6.val.traffic_class &= ipv6.mask.traffic_class;
+               ipv6.val.next_hdr &= ipv6.mask.next_hdr;
+               ipv6.val.hop_limit &= ipv6.mask.hop_limit;
+       }
+       mlx5_flow_spec_verbs_add(flow, &ipv6, size);
+       return size;
+}
+
 /**
  * Convert the @p pattern into a Verbs specifications after ensuring the NIC
  * will understand and process it correctly.
@@ -641,6 +753,9 @@ mlx5_flow_items(const struct rte_flow_item pattern[],
                case RTE_FLOW_ITEM_TYPE_IPV4:
                        ret = mlx5_flow_item_ipv4(pattern, flow, remain, error);
                        break;
+               case RTE_FLOW_ITEM_TYPE_IPV6:
+                       ret = mlx5_flow_item_ipv6(pattern, flow, remain, error);
+                       break;
                default:
                        return rte_flow_error_set(error, ENOTSUP,
                                                  RTE_FLOW_ERROR_TYPE_ITEM,