net/txgbe: parse RSS filter
authorJiawen Wu <jiawenwu@trustnetic.com>
Fri, 18 Dec 2020 09:36:47 +0000 (17:36 +0800)
committerFerruh Yigit <ferruh.yigit@intel.com>
Wed, 13 Jan 2021 17:51:58 +0000 (18:51 +0100)
Check if the rule is a RSS filter rule, and get the RSS info.

Signed-off-by: Jiawen Wu <jiawenwu@trustnetic.com>
drivers/net/txgbe/txgbe_flow.c

index 47bebb9..313f985 100644 (file)
@@ -2370,6 +2370,120 @@ step_next:
        return ret;
 }
 
+static int
+txgbe_parse_rss_filter(struct rte_eth_dev *dev,
+                       const struct rte_flow_attr *attr,
+                       const struct rte_flow_action actions[],
+                       struct txgbe_rte_flow_rss_conf *rss_conf,
+                       struct rte_flow_error *error)
+{
+       const struct rte_flow_action *act;
+       const struct rte_flow_action_rss *rss;
+       uint16_t n;
+
+       /**
+        * rss only supports forwarding,
+        * check if the first not void action is RSS.
+        */
+       act = next_no_void_action(actions, NULL);
+       if (act->type != RTE_FLOW_ACTION_TYPE_RSS) {
+               memset(rss_conf, 0, sizeof(struct txgbe_rte_flow_rss_conf));
+               rte_flow_error_set(error, EINVAL,
+                       RTE_FLOW_ERROR_TYPE_ACTION,
+                       act, "Not supported action.");
+               return -rte_errno;
+       }
+
+       rss = (const struct rte_flow_action_rss *)act->conf;
+
+       if (!rss || !rss->queue_num) {
+               rte_flow_error_set(error, EINVAL,
+                               RTE_FLOW_ERROR_TYPE_ACTION,
+                               act,
+                          "no valid queues");
+               return -rte_errno;
+       }
+
+       for (n = 0; n < rss->queue_num; n++) {
+               if (rss->queue[n] >= dev->data->nb_rx_queues) {
+                       rte_flow_error_set(error, EINVAL,
+                                  RTE_FLOW_ERROR_TYPE_ACTION,
+                                  act,
+                                  "queue id > max number of queues");
+                       return -rte_errno;
+               }
+       }
+
+       if (rss->func != RTE_ETH_HASH_FUNCTION_DEFAULT)
+               return rte_flow_error_set
+                       (error, ENOTSUP, RTE_FLOW_ERROR_TYPE_ACTION, act,
+                        "non-default RSS hash functions are not supported");
+       if (rss->level)
+               return rte_flow_error_set
+                       (error, ENOTSUP, RTE_FLOW_ERROR_TYPE_ACTION, act,
+                        "a nonzero RSS encapsulation level is not supported");
+       if (rss->key_len && rss->key_len != RTE_DIM(rss_conf->key))
+               return rte_flow_error_set
+                       (error, ENOTSUP, RTE_FLOW_ERROR_TYPE_ACTION, act,
+                        "RSS hash key must be exactly 40 bytes");
+       if (rss->queue_num > RTE_DIM(rss_conf->queue))
+               return rte_flow_error_set
+                       (error, ENOTSUP, RTE_FLOW_ERROR_TYPE_ACTION, act,
+                        "too many queues for RSS context");
+       if (txgbe_rss_conf_init(rss_conf, rss))
+               return rte_flow_error_set
+                       (error, EINVAL, RTE_FLOW_ERROR_TYPE_ACTION, act,
+                        "RSS context initialization failure");
+
+       /* check if the next not void item is END */
+       act = next_no_void_action(actions, act);
+       if (act->type != RTE_FLOW_ACTION_TYPE_END) {
+               memset(rss_conf, 0, sizeof(struct rte_eth_rss_conf));
+               rte_flow_error_set(error, EINVAL,
+                       RTE_FLOW_ERROR_TYPE_ACTION,
+                       act, "Not supported action.");
+               return -rte_errno;
+       }
+
+       /* parse attr */
+       /* must be input direction */
+       if (!attr->ingress) {
+               memset(rss_conf, 0, sizeof(struct txgbe_rte_flow_rss_conf));
+               rte_flow_error_set(error, EINVAL,
+                                  RTE_FLOW_ERROR_TYPE_ATTR_INGRESS,
+                                  attr, "Only support ingress.");
+               return -rte_errno;
+       }
+
+       /* not supported */
+       if (attr->egress) {
+               memset(rss_conf, 0, sizeof(struct txgbe_rte_flow_rss_conf));
+               rte_flow_error_set(error, EINVAL,
+                                  RTE_FLOW_ERROR_TYPE_ATTR_EGRESS,
+                                  attr, "Not support egress.");
+               return -rte_errno;
+       }
+
+       /* not supported */
+       if (attr->transfer) {
+               memset(rss_conf, 0, sizeof(struct txgbe_rte_flow_rss_conf));
+               rte_flow_error_set(error, EINVAL,
+                                  RTE_FLOW_ERROR_TYPE_ATTR_TRANSFER,
+                                  attr, "No support for transfer.");
+               return -rte_errno;
+       }
+
+       if (attr->priority > 0xFFFF) {
+               memset(rss_conf, 0, sizeof(struct txgbe_rte_flow_rss_conf));
+               rte_flow_error_set(error, EINVAL,
+                                  RTE_FLOW_ERROR_TYPE_ATTR_PRIORITY,
+                                  attr, "Error priority.");
+               return -rte_errno;
+       }
+
+       return 0;
+}
+
 /**
  * Create or destroy a flow rule.
  * Theorically one rule can match more than one filters.
@@ -2404,6 +2518,7 @@ txgbe_flow_validate(struct rte_eth_dev *dev,
        struct rte_eth_syn_filter syn_filter;
        struct txgbe_l2_tunnel_conf l2_tn_filter;
        struct txgbe_fdir_rule fdir_rule;
+       struct txgbe_rte_flow_rss_conf rss_conf;
        int ret = 0;
 
        memset(&ntuple_filter, 0, sizeof(struct rte_eth_ntuple_filter));
@@ -2436,6 +2551,10 @@ txgbe_flow_validate(struct rte_eth_dev *dev,
        if (!ret)
                return 0;
 
+       memset(&rss_conf, 0, sizeof(struct txgbe_rte_flow_rss_conf));
+       ret = txgbe_parse_rss_filter(dev, attr,
+                                       actions, &rss_conf, error);
+
        return ret;
 }