+#include "mlx5_defs.h"
+#include "mlx5_prm.h"
+
+/* Define minimal priority for control plane flows. */
+#define MLX5_CTRL_FLOW_PRIORITY 4
+
+/* Internet Protocol versions. */
+#define MLX5_IPV4 4
+#define MLX5_IPV6 6
+
+#ifndef HAVE_IBV_DEVICE_COUNTERS_SET_SUPPORT
+struct ibv_counter_set_init_attr {
+ int dummy;
+};
+struct ibv_flow_spec_counter_action {
+ int dummy;
+};
+struct ibv_counter_set {
+ int dummy;
+};
+
+static inline int
+ibv_destroy_counter_set(struct ibv_counter_set *cs)
+{
+ (void)cs;
+ return -ENOTSUP;
+}
+#endif
+
+/* Dev ops structure defined in mlx5.c */
+extern const struct eth_dev_ops mlx5_dev_ops;
+extern const struct eth_dev_ops mlx5_dev_ops_isolate;
+
+static int
+mlx5_flow_create_eth(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data);
+
+static int
+mlx5_flow_create_vlan(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data);
+
+static int
+mlx5_flow_create_ipv4(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data);
+
+static int
+mlx5_flow_create_ipv6(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data);
+
+static int
+mlx5_flow_create_udp(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data);
+
+static int
+mlx5_flow_create_tcp(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data);
+
+static int
+mlx5_flow_create_vxlan(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data);
+
+struct mlx5_flow_parse;
+
+static void
+mlx5_flow_create_copy(struct mlx5_flow_parse *parser, void *src,
+ unsigned int size);
+
+static int
+mlx5_flow_create_flag_mark(struct mlx5_flow_parse *parser, uint32_t mark_id);
+
+static int
+mlx5_flow_create_count(struct priv *priv, struct mlx5_flow_parse *parser);
+
+/* Hash RX queue types. */
+enum hash_rxq_type {
+ HASH_RXQ_TCPV4,
+ HASH_RXQ_UDPV4,
+ HASH_RXQ_IPV4,
+ HASH_RXQ_TCPV6,
+ HASH_RXQ_UDPV6,
+ HASH_RXQ_IPV6,
+ HASH_RXQ_ETH,
+};
+
+/* Initialization data for hash RX queue. */
+struct hash_rxq_init {
+ uint64_t hash_fields; /* Fields that participate in the hash. */
+ uint64_t dpdk_rss_hf; /* Matching DPDK RSS hash fields. */
+ unsigned int flow_priority; /* Flow priority to use. */
+ unsigned int ip_version; /* Internet protocol. */
+};
+
+/* Initialization data for hash RX queues. */
+const struct hash_rxq_init hash_rxq_init[] = {
+ [HASH_RXQ_TCPV4] = {
+ .hash_fields = (IBV_RX_HASH_SRC_IPV4 |
+ IBV_RX_HASH_DST_IPV4 |
+ IBV_RX_HASH_SRC_PORT_TCP |
+ IBV_RX_HASH_DST_PORT_TCP),
+ .dpdk_rss_hf = ETH_RSS_NONFRAG_IPV4_TCP,
+ .flow_priority = 0,
+ .ip_version = MLX5_IPV4,
+ },
+ [HASH_RXQ_UDPV4] = {
+ .hash_fields = (IBV_RX_HASH_SRC_IPV4 |
+ IBV_RX_HASH_DST_IPV4 |
+ IBV_RX_HASH_SRC_PORT_UDP |
+ IBV_RX_HASH_DST_PORT_UDP),
+ .dpdk_rss_hf = ETH_RSS_NONFRAG_IPV4_UDP,
+ .flow_priority = 0,
+ .ip_version = MLX5_IPV4,
+ },
+ [HASH_RXQ_IPV4] = {
+ .hash_fields = (IBV_RX_HASH_SRC_IPV4 |
+ IBV_RX_HASH_DST_IPV4),
+ .dpdk_rss_hf = (ETH_RSS_IPV4 |
+ ETH_RSS_FRAG_IPV4),
+ .flow_priority = 1,
+ .ip_version = MLX5_IPV4,
+ },
+ [HASH_RXQ_TCPV6] = {
+ .hash_fields = (IBV_RX_HASH_SRC_IPV6 |
+ IBV_RX_HASH_DST_IPV6 |
+ IBV_RX_HASH_SRC_PORT_TCP |
+ IBV_RX_HASH_DST_PORT_TCP),
+ .dpdk_rss_hf = ETH_RSS_NONFRAG_IPV6_TCP,
+ .flow_priority = 0,
+ .ip_version = MLX5_IPV6,
+ },
+ [HASH_RXQ_UDPV6] = {
+ .hash_fields = (IBV_RX_HASH_SRC_IPV6 |
+ IBV_RX_HASH_DST_IPV6 |
+ IBV_RX_HASH_SRC_PORT_UDP |
+ IBV_RX_HASH_DST_PORT_UDP),
+ .dpdk_rss_hf = ETH_RSS_NONFRAG_IPV6_UDP,
+ .flow_priority = 0,
+ .ip_version = MLX5_IPV6,
+ },
+ [HASH_RXQ_IPV6] = {
+ .hash_fields = (IBV_RX_HASH_SRC_IPV6 |
+ IBV_RX_HASH_DST_IPV6),
+ .dpdk_rss_hf = (ETH_RSS_IPV6 |
+ ETH_RSS_FRAG_IPV6),
+ .flow_priority = 1,
+ .ip_version = MLX5_IPV6,
+ },
+ [HASH_RXQ_ETH] = {
+ .hash_fields = 0,
+ .dpdk_rss_hf = 0,
+ .flow_priority = 2,
+ },
+};
+
+/* Number of entries in hash_rxq_init[]. */
+const unsigned int hash_rxq_init_n = RTE_DIM(hash_rxq_init);
+
+/** Structure for holding counter stats. */
+struct mlx5_flow_counter_stats {
+ uint64_t hits; /**< Number of packets matched by the rule. */
+ uint64_t bytes; /**< Number of bytes matched by the rule. */
+};
+
+/** Structure for Drop queue. */
+struct mlx5_hrxq_drop {
+ struct ibv_rwq_ind_table *ind_table; /**< Indirection table. */
+ struct ibv_qp *qp; /**< Verbs queue pair. */
+ struct ibv_wq *wq; /**< Verbs work queue. */
+ struct ibv_cq *cq; /**< Verbs completion queue. */
+};
+
+/* Flows structures. */
+struct mlx5_flow {
+ uint64_t hash_fields; /**< Fields that participate in the hash. */
+ struct ibv_flow_attr *ibv_attr; /**< Pointer to Verbs attributes. */
+ struct ibv_flow *ibv_flow; /**< Verbs flow. */
+ struct mlx5_hrxq *hrxq; /**< Hash Rx queues. */
+};
+
+/* Drop flows structures. */
+struct mlx5_flow_drop {
+ struct ibv_flow_attr *ibv_attr; /**< Pointer to Verbs attributes. */
+ struct ibv_flow *ibv_flow; /**< Verbs flow. */
+};
+
+struct rte_flow {
+ TAILQ_ENTRY(rte_flow) next; /**< Pointer to the next flow structure. */
+ uint32_t mark:1; /**< Set if the flow is marked. */
+ uint32_t drop:1; /**< Drop queue. */
+ uint16_t queues_n; /**< Number of entries in queue[]. */
+ uint16_t (*queues)[]; /**< Queues indexes to use. */
+ struct rte_eth_rss_conf rss_conf; /**< RSS configuration */
+ uint8_t rss_key[40]; /**< copy of the RSS key. */
+ struct ibv_counter_set *cs; /**< Holds the counters for the rule. */
+ struct mlx5_flow_counter_stats counter_stats;/**<The counter stats. */
+ struct mlx5_flow frxq[RTE_DIM(hash_rxq_init)];
+ /**< Flow with Rx queue. */
+};
+
+/** Static initializer for items. */
+#define ITEMS(...) \
+ (const enum rte_flow_item_type []){ \
+ __VA_ARGS__, RTE_FLOW_ITEM_TYPE_END, \
+ }
+
+/** Structure to generate a simple graph of layers supported by the NIC. */
+struct mlx5_flow_items {
+ /** List of possible actions for these items. */
+ const enum rte_flow_action_type *const actions;
+ /** Bit-masks corresponding to the possibilities for the item. */
+ const void *mask;
+ /**
+ * Default bit-masks to use when item->mask is not provided. When
+ * \default_mask is also NULL, the full supported bit-mask (\mask) is
+ * used instead.
+ */
+ const void *default_mask;
+ /** Bit-masks size in bytes. */
+ const unsigned int mask_sz;
+ /**
+ * Conversion function from rte_flow to NIC specific flow.
+ *
+ * @param item
+ * rte_flow item to convert.
+ * @param default_mask
+ * Default bit-masks to use when item->mask is not provided.
+ * @param data
+ * Internal structure to store the conversion.
+ *
+ * @return
+ * 0 on success, negative value otherwise.
+ */
+ int (*convert)(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data);
+ /** Size in bytes of the destination structure. */
+ const unsigned int dst_sz;
+ /** List of possible following items. */
+ const enum rte_flow_item_type *const items;
+};
+
+/** Valid action for this PMD. */
+static const enum rte_flow_action_type valid_actions[] = {
+ RTE_FLOW_ACTION_TYPE_DROP,
+ RTE_FLOW_ACTION_TYPE_QUEUE,
+ RTE_FLOW_ACTION_TYPE_MARK,
+ RTE_FLOW_ACTION_TYPE_FLAG,
+#ifdef HAVE_IBV_DEVICE_COUNTERS_SET_SUPPORT
+ RTE_FLOW_ACTION_TYPE_COUNT,
+#endif
+ RTE_FLOW_ACTION_TYPE_END,
+};
+
+/** Graph of supported items and associated actions. */
+static const struct mlx5_flow_items mlx5_flow_items[] = {
+ [RTE_FLOW_ITEM_TYPE_END] = {
+ .items = ITEMS(RTE_FLOW_ITEM_TYPE_ETH,
+ RTE_FLOW_ITEM_TYPE_VXLAN),
+ },
+ [RTE_FLOW_ITEM_TYPE_ETH] = {
+ .items = ITEMS(RTE_FLOW_ITEM_TYPE_VLAN,
+ RTE_FLOW_ITEM_TYPE_IPV4,
+ RTE_FLOW_ITEM_TYPE_IPV6),
+ .actions = valid_actions,
+ .mask = &(const struct rte_flow_item_eth){
+ .dst.addr_bytes = "\xff\xff\xff\xff\xff\xff",
+ .src.addr_bytes = "\xff\xff\xff\xff\xff\xff",
+ .type = -1,
+ },
+ .default_mask = &rte_flow_item_eth_mask,
+ .mask_sz = sizeof(struct rte_flow_item_eth),
+ .convert = mlx5_flow_create_eth,
+ .dst_sz = sizeof(struct ibv_flow_spec_eth),
+ },
+ [RTE_FLOW_ITEM_TYPE_VLAN] = {
+ .items = ITEMS(RTE_FLOW_ITEM_TYPE_IPV4,
+ RTE_FLOW_ITEM_TYPE_IPV6),
+ .actions = valid_actions,
+ .mask = &(const struct rte_flow_item_vlan){
+ .tci = -1,
+ },
+ .default_mask = &rte_flow_item_vlan_mask,
+ .mask_sz = sizeof(struct rte_flow_item_vlan),
+ .convert = mlx5_flow_create_vlan,
+ .dst_sz = 0,
+ },
+ [RTE_FLOW_ITEM_TYPE_IPV4] = {
+ .items = ITEMS(RTE_FLOW_ITEM_TYPE_UDP,
+ RTE_FLOW_ITEM_TYPE_TCP),
+ .actions = valid_actions,
+ .mask = &(const struct rte_flow_item_ipv4){
+ .hdr = {
+ .src_addr = -1,
+ .dst_addr = -1,
+ .type_of_service = -1,
+ .next_proto_id = -1,
+ .time_to_live = -1,
+ },
+ },
+ .default_mask = &rte_flow_item_ipv4_mask,
+ .mask_sz = sizeof(struct rte_flow_item_ipv4),
+ .convert = mlx5_flow_create_ipv4,
+ .dst_sz = sizeof(struct ibv_flow_spec_ipv4_ext),
+ },
+ [RTE_FLOW_ITEM_TYPE_IPV6] = {
+ .items = ITEMS(RTE_FLOW_ITEM_TYPE_UDP,
+ RTE_FLOW_ITEM_TYPE_TCP),
+ .actions = valid_actions,
+ .mask = &(const struct rte_flow_item_ipv6){
+ .hdr = {
+ .src_addr = {
+ 0xff, 0xff, 0xff, 0xff,
+ 0xff, 0xff, 0xff, 0xff,
+ 0xff, 0xff, 0xff, 0xff,
+ 0xff, 0xff, 0xff, 0xff,
+ },
+ .dst_addr = {
+ 0xff, 0xff, 0xff, 0xff,
+ 0xff, 0xff, 0xff, 0xff,
+ 0xff, 0xff, 0xff, 0xff,
+ 0xff, 0xff, 0xff, 0xff,
+ },
+ .vtc_flow = -1,
+ .proto = -1,
+ .hop_limits = -1,
+ },
+ },
+ .default_mask = &rte_flow_item_ipv6_mask,
+ .mask_sz = sizeof(struct rte_flow_item_ipv6),
+ .convert = mlx5_flow_create_ipv6,
+ .dst_sz = sizeof(struct ibv_flow_spec_ipv6),
+ },
+ [RTE_FLOW_ITEM_TYPE_UDP] = {
+ .items = ITEMS(RTE_FLOW_ITEM_TYPE_VXLAN),
+ .actions = valid_actions,
+ .mask = &(const struct rte_flow_item_udp){
+ .hdr = {
+ .src_port = -1,
+ .dst_port = -1,
+ },
+ },
+ .default_mask = &rte_flow_item_udp_mask,
+ .mask_sz = sizeof(struct rte_flow_item_udp),
+ .convert = mlx5_flow_create_udp,
+ .dst_sz = sizeof(struct ibv_flow_spec_tcp_udp),
+ },
+ [RTE_FLOW_ITEM_TYPE_TCP] = {
+ .actions = valid_actions,
+ .mask = &(const struct rte_flow_item_tcp){
+ .hdr = {
+ .src_port = -1,
+ .dst_port = -1,
+ },
+ },
+ .default_mask = &rte_flow_item_tcp_mask,
+ .mask_sz = sizeof(struct rte_flow_item_tcp),
+ .convert = mlx5_flow_create_tcp,
+ .dst_sz = sizeof(struct ibv_flow_spec_tcp_udp),
+ },
+ [RTE_FLOW_ITEM_TYPE_VXLAN] = {
+ .items = ITEMS(RTE_FLOW_ITEM_TYPE_ETH),
+ .actions = valid_actions,
+ .mask = &(const struct rte_flow_item_vxlan){
+ .vni = "\xff\xff\xff",
+ },
+ .default_mask = &rte_flow_item_vxlan_mask,
+ .mask_sz = sizeof(struct rte_flow_item_vxlan),
+ .convert = mlx5_flow_create_vxlan,
+ .dst_sz = sizeof(struct ibv_flow_spec_tunnel),
+ },
+};
+
+/** Structure to pass to the conversion function. */
+struct mlx5_flow_parse {
+ uint32_t inner; /**< Set once VXLAN is encountered. */
+ uint32_t allmulti:1; /**< Set once allmulti dst MAC is encountered. */
+ uint32_t create:1;
+ /**< Whether resources should remain after a validate. */
+ uint32_t drop:1; /**< Target is a drop queue. */
+ uint32_t mark:1; /**< Mark is present in the flow. */
+ uint32_t count:1; /**< Count is present in the flow. */
+ uint32_t mark_id; /**< Mark identifier. */
+ uint16_t queues[RTE_MAX_QUEUES_PER_PORT]; /**< Queues indexes to use. */
+ uint16_t queues_n; /**< Number of entries in queue[]. */
+ struct rte_eth_rss_conf rss_conf; /**< RSS configuration */
+ uint8_t rss_key[40]; /**< copy of the RSS key. */
+ enum hash_rxq_type layer; /**< Last pattern layer detected. */
+ struct ibv_counter_set *cs; /**< Holds the counter set for the rule */
+ struct {
+ struct ibv_flow_attr *ibv_attr;
+ /**< Pointer to Verbs attributes. */
+ unsigned int offset;
+ /**< Current position or total size of the attribute. */
+ } queue[RTE_DIM(hash_rxq_init)];
+};
+
+static const struct rte_flow_ops mlx5_flow_ops = {
+ .validate = mlx5_flow_validate,
+ .create = mlx5_flow_create,
+ .destroy = mlx5_flow_destroy,
+ .flush = mlx5_flow_flush,
+#ifdef HAVE_IBV_DEVICE_COUNTERS_SET_SUPPORT
+ .query = mlx5_flow_query,
+#else
+ .query = NULL,
+#endif
+ .isolate = mlx5_flow_isolate,
+};
+
+/* Convert FDIR request to Generic flow. */
+struct mlx5_fdir {
+ struct rte_flow_attr attr;
+ struct rte_flow_action actions[2];
+ struct rte_flow_item items[4];
+ struct rte_flow_item_eth l2;
+ struct rte_flow_item_eth l2_mask;
+ union {
+ struct rte_flow_item_ipv4 ipv4;
+ struct rte_flow_item_ipv6 ipv6;
+ } l3;
+ union {
+ struct rte_flow_item_udp udp;
+ struct rte_flow_item_tcp tcp;
+ } l4;
+ struct rte_flow_action_queue queue;
+};
+
+/* Verbs specification header. */
+struct ibv_spec_header {
+ enum ibv_flow_spec_type type;
+ uint16_t size;
+};
+
+/**
+ * Check support for a given item.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param mask[in]
+ * Bit-masks covering supported fields to compare with spec, last and mask in
+ * \item.
+ * @param size
+ * Bit-Mask size in bytes.
+ *
+ * @return
+ * 0 on success.
+ */
+static int
+mlx5_flow_item_validate(const struct rte_flow_item *item,
+ const uint8_t *mask, unsigned int size)
+{
+ int ret = 0;
+
+ if (!item->spec && (item->mask || item->last))
+ return -1;
+ if (item->spec && !item->mask) {
+ unsigned int i;
+ const uint8_t *spec = item->spec;
+
+ for (i = 0; i < size; ++i)
+ if ((spec[i] | mask[i]) != mask[i])
+ return -1;
+ }
+ if (item->last && !item->mask) {
+ unsigned int i;
+ const uint8_t *spec = item->last;
+
+ for (i = 0; i < size; ++i)
+ if ((spec[i] | mask[i]) != mask[i])
+ return -1;
+ }
+ if (item->mask) {
+ unsigned int i;
+ const uint8_t *spec = item->spec;
+
+ for (i = 0; i < size; ++i)
+ if ((spec[i] | mask[i]) != mask[i])
+ return -1;
+ }
+ if (item->spec && item->last) {
+ uint8_t spec[size];
+ uint8_t last[size];
+ const uint8_t *apply = mask;
+ unsigned int i;
+
+ if (item->mask)
+ apply = item->mask;
+ for (i = 0; i < size; ++i) {
+ spec[i] = ((const uint8_t *)item->spec)[i] & apply[i];
+ last[i] = ((const uint8_t *)item->last)[i] & apply[i];
+ }
+ ret = memcmp(spec, last, size);
+ }
+ return ret;
+}
+
+/**
+ * Copy the RSS configuration from the user ones.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param parser
+ * Internal parser structure.
+ * @param rss_conf
+ * User RSS configuration to save.
+ *
+ * @return
+ * 0 on success, errno value on failure.
+ */
+static int
+priv_flow_convert_rss_conf(struct priv *priv,
+ struct mlx5_flow_parse *parser,
+ const struct rte_eth_rss_conf *rss_conf)
+{
+ const struct rte_eth_rss_conf *rss;
+
+ if (rss_conf) {
+ if (rss_conf->rss_hf & MLX5_RSS_HF_MASK)
+ return EINVAL;
+ rss = rss_conf;
+ } else {
+ rss = &priv->rss_conf;
+ }
+ if (rss->rss_key_len > 40)
+ return EINVAL;
+ parser->rss_conf.rss_key_len = rss->rss_key_len;
+ parser->rss_conf.rss_hf = rss->rss_hf;
+ memcpy(parser->rss_key, rss->rss_key, rss->rss_key_len);
+ parser->rss_conf.rss_key = parser->rss_key;
+ return 0;
+}
+
+/**
+ * Extract attribute to the parser.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param[in] attr
+ * Flow rule attributes.
+ * @param[out] error
+ * Perform verbose error reporting if not NULL.
+ * @param[in, out] parser
+ * Internal parser structure.
+ *
+ * @return
+ * 0 on success, a negative errno value otherwise and rte_errno is set.
+ */
+static int
+priv_flow_convert_attributes(struct priv *priv,
+ const struct rte_flow_attr *attr,
+ struct rte_flow_error *error,
+ struct mlx5_flow_parse *parser)
+{
+ (void)priv;
+ (void)parser;
+ if (attr->group) {
+ rte_flow_error_set(error, ENOTSUP,
+ RTE_FLOW_ERROR_TYPE_ATTR_GROUP,
+ NULL,
+ "groups are not supported");
+ return -rte_errno;
+ }
+ if (attr->priority && attr->priority != MLX5_CTRL_FLOW_PRIORITY) {
+ rte_flow_error_set(error, ENOTSUP,
+ RTE_FLOW_ERROR_TYPE_ATTR_PRIORITY,
+ NULL,
+ "priorities are not supported");
+ return -rte_errno;
+ }
+ if (attr->egress) {
+ rte_flow_error_set(error, ENOTSUP,
+ RTE_FLOW_ERROR_TYPE_ATTR_EGRESS,
+ NULL,
+ "egress is not supported");
+ return -rte_errno;
+ }
+ if (!attr->ingress) {
+ rte_flow_error_set(error, ENOTSUP,
+ RTE_FLOW_ERROR_TYPE_ATTR_INGRESS,
+ NULL,
+ "only ingress is supported");
+ return -rte_errno;
+ }
+ return 0;
+}
+
+/**
+ * Extract actions request to the parser.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param[in] actions
+ * Associated actions (list terminated by the END action).
+ * @param[out] error
+ * Perform verbose error reporting if not NULL.
+ * @param[in, out] parser
+ * Internal parser structure.
+ *
+ * @return
+ * 0 on success, a negative errno value otherwise and rte_errno is set.
+ */
+static int
+priv_flow_convert_actions(struct priv *priv,
+ const struct rte_flow_action actions[],
+ struct rte_flow_error *error,
+ struct mlx5_flow_parse *parser)
+{
+ /*
+ * Add default RSS configuration necessary for Verbs to create QP even
+ * if no RSS is necessary.
+ */
+ priv_flow_convert_rss_conf(priv, parser,
+ (const struct rte_eth_rss_conf *)
+ &priv->rss_conf);
+ for (; actions->type != RTE_FLOW_ACTION_TYPE_END; ++actions) {
+ if (actions->type == RTE_FLOW_ACTION_TYPE_VOID) {
+ continue;
+ } else if (actions->type == RTE_FLOW_ACTION_TYPE_DROP) {
+ parser->drop = 1;
+ } else if (actions->type == RTE_FLOW_ACTION_TYPE_QUEUE) {
+ const struct rte_flow_action_queue *queue =
+ (const struct rte_flow_action_queue *)
+ actions->conf;
+ uint16_t n;
+ uint16_t found = 0;
+
+ if (!queue || (queue->index > (priv->rxqs_n - 1)))
+ goto exit_action_not_supported;
+ for (n = 0; n < parser->queues_n; ++n) {
+ if (parser->queues[n] == queue->index) {
+ found = 1;
+ break;
+ }
+ }
+ if (parser->queues_n > 1 && !found) {
+ rte_flow_error_set(error, ENOTSUP,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ actions,
+ "queue action not in RSS queues");
+ return -rte_errno;
+ }
+ if (!found) {
+ parser->queues_n = 1;
+ parser->queues[0] = queue->index;
+ }
+ } else if (actions->type == RTE_FLOW_ACTION_TYPE_RSS) {
+ const struct rte_flow_action_rss *rss =
+ (const struct rte_flow_action_rss *)
+ actions->conf;
+ uint16_t n;
+
+ if (!rss || !rss->num) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ actions,
+ "no valid queues");
+ return -rte_errno;
+ }
+ if (parser->queues_n == 1) {
+ uint16_t found = 0;
+
+ assert(parser->queues_n);
+ for (n = 0; n < rss->num; ++n) {
+ if (parser->queues[0] ==
+ rss->queue[n]) {
+ found = 1;
+ break;
+ }
+ }
+ if (!found) {
+ rte_flow_error_set(error, ENOTSUP,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ actions,
+ "queue action not in RSS"
+ " queues");
+ return -rte_errno;
+ }
+ }
+ for (n = 0; n < rss->num; ++n) {
+ if (rss->queue[n] >= priv->rxqs_n) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ actions,
+ "queue id > number of"
+ " queues");
+ return -rte_errno;
+ }
+ }
+ for (n = 0; n < rss->num; ++n)
+ parser->queues[n] = rss->queue[n];
+ parser->queues_n = rss->num;
+ if (priv_flow_convert_rss_conf(priv, parser,
+ rss->rss_conf)) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ actions,
+ "wrong RSS configuration");
+ return -rte_errno;
+ }
+ } else if (actions->type == RTE_FLOW_ACTION_TYPE_MARK) {
+ const struct rte_flow_action_mark *mark =
+ (const struct rte_flow_action_mark *)
+ actions->conf;
+
+ if (!mark) {
+ rte_flow_error_set(error, EINVAL,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ actions,
+ "mark must be defined");
+ return -rte_errno;
+ } else if (mark->id >= MLX5_FLOW_MARK_MAX) {
+ rte_flow_error_set(error, ENOTSUP,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ actions,
+ "mark must be between 0"
+ " and 16777199");
+ return -rte_errno;
+ }
+ parser->mark = 1;
+ parser->mark_id = mark->id;
+ } else if (actions->type == RTE_FLOW_ACTION_TYPE_FLAG) {
+ parser->mark = 1;
+ } else if (actions->type == RTE_FLOW_ACTION_TYPE_COUNT &&
+ priv->config.flow_counter_en) {
+ parser->count = 1;
+ } else {
+ goto exit_action_not_supported;
+ }
+ }
+ if (parser->drop && parser->mark)
+ parser->mark = 0;
+ if (!parser->queues_n && !parser->drop) {
+ rte_flow_error_set(error, ENOTSUP, RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "no valid action");
+ return -rte_errno;
+ }
+ return 0;
+exit_action_not_supported:
+ rte_flow_error_set(error, ENOTSUP, RTE_FLOW_ERROR_TYPE_ACTION,
+ actions, "action not supported");
+ return -rte_errno;
+}
+
+/**
+ * Validate items.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param[in] items
+ * Pattern specification (list terminated by the END pattern item).
+ * @param[out] error
+ * Perform verbose error reporting if not NULL.
+ * @param[in, out] parser
+ * Internal parser structure.
+ *
+ * @return
+ * 0 on success, a negative errno value otherwise and rte_errno is set.
+ */
+static int
+priv_flow_convert_items_validate(struct priv *priv,
+ const struct rte_flow_item items[],
+ struct rte_flow_error *error,
+ struct mlx5_flow_parse *parser)
+{
+ const struct mlx5_flow_items *cur_item = mlx5_flow_items;
+ unsigned int i;
+
+ (void)priv;
+ /* Initialise the offsets to start after verbs attribute. */
+ for (i = 0; i != hash_rxq_init_n; ++i)
+ parser->queue[i].offset = sizeof(struct ibv_flow_attr);
+ for (; items->type != RTE_FLOW_ITEM_TYPE_END; ++items) {
+ const struct mlx5_flow_items *token = NULL;
+ unsigned int n;
+ int err;
+
+ if (items->type == RTE_FLOW_ITEM_TYPE_VOID)
+ continue;
+ for (i = 0;
+ cur_item->items &&
+ cur_item->items[i] != RTE_FLOW_ITEM_TYPE_END;
+ ++i) {
+ if (cur_item->items[i] == items->type) {
+ token = &mlx5_flow_items[items->type];
+ break;
+ }
+ }
+ if (!token)
+ goto exit_item_not_supported;
+ cur_item = token;
+ err = mlx5_flow_item_validate(items,
+ (const uint8_t *)cur_item->mask,
+ cur_item->mask_sz);
+ if (err)
+ goto exit_item_not_supported;
+ if (items->type == RTE_FLOW_ITEM_TYPE_VXLAN) {
+ if (parser->inner) {
+ rte_flow_error_set(error, ENOTSUP,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ items,
+ "cannot recognize multiple"
+ " VXLAN encapsulations");
+ return -rte_errno;
+ }
+ parser->inner = IBV_FLOW_SPEC_INNER;
+ }
+ if (parser->drop || parser->queues_n == 1) {
+ parser->queue[HASH_RXQ_ETH].offset += cur_item->dst_sz;
+ } else {
+ for (n = 0; n != hash_rxq_init_n; ++n)
+ parser->queue[n].offset += cur_item->dst_sz;
+ }
+ }
+ if (parser->drop) {
+ parser->queue[HASH_RXQ_ETH].offset +=
+ sizeof(struct ibv_flow_spec_action_drop);
+ }
+ if (parser->mark) {
+ for (i = 0; i != hash_rxq_init_n; ++i)
+ parser->queue[i].offset +=
+ sizeof(struct ibv_flow_spec_action_tag);
+ }
+ if (parser->count) {
+ unsigned int size = sizeof(struct ibv_flow_spec_counter_action);
+
+ for (i = 0; i != hash_rxq_init_n; ++i)
+ parser->queue[i].offset += size;
+ }
+ return 0;
+exit_item_not_supported:
+ rte_flow_error_set(error, ENOTSUP, RTE_FLOW_ERROR_TYPE_ITEM,
+ items, "item not supported");
+ return -rte_errno;
+}
+
+/**
+ * Allocate memory space to store verbs flow attributes.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param[in] priority
+ * Flow priority.
+ * @param[in] size
+ * Amount of byte to allocate.
+ * @param[out] error
+ * Perform verbose error reporting if not NULL.
+ *
+ * @return
+ * A verbs flow attribute on success, NULL otherwise.
+ */
+static struct ibv_flow_attr*
+priv_flow_convert_allocate(struct priv *priv,
+ unsigned int priority,
+ unsigned int size,
+ struct rte_flow_error *error)
+{
+ struct ibv_flow_attr *ibv_attr;
+
+ (void)priv;
+ ibv_attr = rte_calloc(__func__, 1, size, 0);
+ if (!ibv_attr) {
+ rte_flow_error_set(error, ENOMEM,
+ RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+ NULL,
+ "cannot allocate verbs spec attributes.");
+ return NULL;
+ }
+ ibv_attr->priority = priority;
+ return ibv_attr;
+}
+
+/**
+ * Finalise verbs flow attributes.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param[in, out] parser
+ * Internal parser structure.
+ */
+static void
+priv_flow_convert_finalise(struct priv *priv, struct mlx5_flow_parse *parser)
+{
+ const unsigned int ipv4 =
+ hash_rxq_init[parser->layer].ip_version == MLX5_IPV4;
+ const enum hash_rxq_type hmin = ipv4 ? HASH_RXQ_TCPV4 : HASH_RXQ_TCPV6;
+ const enum hash_rxq_type hmax = ipv4 ? HASH_RXQ_IPV4 : HASH_RXQ_IPV6;
+ const enum hash_rxq_type ohmin = ipv4 ? HASH_RXQ_TCPV6 : HASH_RXQ_TCPV4;
+ const enum hash_rxq_type ohmax = ipv4 ? HASH_RXQ_IPV6 : HASH_RXQ_IPV4;
+ const enum hash_rxq_type ip = ipv4 ? HASH_RXQ_IPV4 : HASH_RXQ_IPV6;
+ unsigned int i;
+
+ (void)priv;
+ if (parser->layer == HASH_RXQ_ETH) {
+ goto fill;
+ } else {
+ /*
+ * This layer becomes useless as the pattern define under
+ * layers.
+ */
+ rte_free(parser->queue[HASH_RXQ_ETH].ibv_attr);
+ parser->queue[HASH_RXQ_ETH].ibv_attr = NULL;
+ }
+ /* Remove opposite kind of layer e.g. IPv6 if the pattern is IPv4. */
+ for (i = ohmin; i != (ohmax + 1); ++i) {
+ if (!parser->queue[i].ibv_attr)
+ continue;
+ rte_free(parser->queue[i].ibv_attr);
+ parser->queue[i].ibv_attr = NULL;
+ }
+ /* Remove impossible flow according to the RSS configuration. */
+ if (hash_rxq_init[parser->layer].dpdk_rss_hf &
+ parser->rss_conf.rss_hf) {
+ /* Remove any other flow. */
+ for (i = hmin; i != (hmax + 1); ++i) {
+ if ((i == parser->layer) ||
+ (!parser->queue[i].ibv_attr))
+ continue;
+ rte_free(parser->queue[i].ibv_attr);
+ parser->queue[i].ibv_attr = NULL;
+ }
+ } else if (!parser->queue[ip].ibv_attr) {
+ /* no RSS possible with the current configuration. */
+ parser->queues_n = 1;
+ return;
+ }
+fill:
+ /*
+ * Fill missing layers in verbs specifications, or compute the correct
+ * offset to allocate the memory space for the attributes and
+ * specifications.
+ */
+ for (i = 0; i != hash_rxq_init_n - 1; ++i) {
+ union {
+ struct ibv_flow_spec_ipv4_ext ipv4;
+ struct ibv_flow_spec_ipv6 ipv6;
+ struct ibv_flow_spec_tcp_udp udp_tcp;
+ } specs;
+ void *dst;
+ uint16_t size;
+
+ if (i == parser->layer)
+ continue;
+ if (parser->layer == HASH_RXQ_ETH) {
+ if (hash_rxq_init[i].ip_version == MLX5_IPV4) {
+ size = sizeof(struct ibv_flow_spec_ipv4_ext);
+ specs.ipv4 = (struct ibv_flow_spec_ipv4_ext){
+ .type = IBV_FLOW_SPEC_IPV4_EXT,
+ .size = size,
+ };
+ } else {
+ size = sizeof(struct ibv_flow_spec_ipv6);
+ specs.ipv6 = (struct ibv_flow_spec_ipv6){
+ .type = IBV_FLOW_SPEC_IPV6,
+ .size = size,
+ };
+ }
+ if (parser->queue[i].ibv_attr) {
+ dst = (void *)((uintptr_t)
+ parser->queue[i].ibv_attr +
+ parser->queue[i].offset);
+ memcpy(dst, &specs, size);
+ ++parser->queue[i].ibv_attr->num_of_specs;
+ }
+ parser->queue[i].offset += size;
+ }
+ if ((i == HASH_RXQ_UDPV4) || (i == HASH_RXQ_TCPV4) ||
+ (i == HASH_RXQ_UDPV6) || (i == HASH_RXQ_TCPV6)) {
+ size = sizeof(struct ibv_flow_spec_tcp_udp);
+ specs.udp_tcp = (struct ibv_flow_spec_tcp_udp) {
+ .type = ((i == HASH_RXQ_UDPV4 ||
+ i == HASH_RXQ_UDPV6) ?
+ IBV_FLOW_SPEC_UDP :
+ IBV_FLOW_SPEC_TCP),
+ .size = size,
+ };
+ if (parser->queue[i].ibv_attr) {
+ dst = (void *)((uintptr_t)
+ parser->queue[i].ibv_attr +
+ parser->queue[i].offset);
+ memcpy(dst, &specs, size);
+ ++parser->queue[i].ibv_attr->num_of_specs;
+ }
+ parser->queue[i].offset += size;
+ }
+ }
+}
+
+/**
+ * Validate and convert a flow supported by the NIC.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param[in] attr
+ * Flow rule attributes.
+ * @param[in] pattern
+ * Pattern specification (list terminated by the END pattern item).
+ * @param[in] actions
+ * Associated actions (list terminated by the END action).
+ * @param[out] error
+ * Perform verbose error reporting if not NULL.
+ * @param[in, out] parser
+ * Internal parser structure.
+ *
+ * @return
+ * 0 on success, a negative errno value otherwise and rte_errno is set.
+ */
+static int
+priv_flow_convert(struct priv *priv,
+ const struct rte_flow_attr *attr,
+ const struct rte_flow_item items[],
+ const struct rte_flow_action actions[],
+ struct rte_flow_error *error,
+ struct mlx5_flow_parse *parser)
+{
+ const struct mlx5_flow_items *cur_item = mlx5_flow_items;
+ unsigned int i;
+ int ret;
+
+ /* First step. Validate the attributes, items and actions. */
+ *parser = (struct mlx5_flow_parse){
+ .create = parser->create,
+ .layer = HASH_RXQ_ETH,
+ .mark_id = MLX5_FLOW_MARK_DEFAULT,
+ };
+ ret = priv_flow_convert_attributes(priv, attr, error, parser);
+ if (ret)
+ return ret;
+ ret = priv_flow_convert_actions(priv, actions, error, parser);
+ if (ret)
+ return ret;
+ ret = priv_flow_convert_items_validate(priv, items, error, parser);
+ if (ret)
+ return ret;
+ priv_flow_convert_finalise(priv, parser);
+ /*
+ * Second step.
+ * Allocate the memory space to store verbs specifications.
+ */
+ if (parser->drop || parser->queues_n == 1) {
+ unsigned int priority =
+ attr->priority +
+ hash_rxq_init[HASH_RXQ_ETH].flow_priority;
+ unsigned int offset = parser->queue[HASH_RXQ_ETH].offset;
+
+ parser->queue[HASH_RXQ_ETH].ibv_attr =
+ priv_flow_convert_allocate(priv, priority,
+ offset, error);
+ if (!parser->queue[HASH_RXQ_ETH].ibv_attr)
+ return ENOMEM;
+ parser->queue[HASH_RXQ_ETH].offset =
+ sizeof(struct ibv_flow_attr);
+ } else {
+ for (i = 0; i != hash_rxq_init_n; ++i) {
+ unsigned int priority =
+ attr->priority +
+ hash_rxq_init[i].flow_priority;
+ unsigned int offset;
+
+ if (!(parser->rss_conf.rss_hf &
+ hash_rxq_init[i].dpdk_rss_hf) &&
+ (i != HASH_RXQ_ETH))
+ continue;
+ offset = parser->queue[i].offset;
+ parser->queue[i].ibv_attr =
+ priv_flow_convert_allocate(priv, priority,
+ offset, error);
+ if (!parser->queue[i].ibv_attr)
+ goto exit_enomem;
+ parser->queue[i].offset = sizeof(struct ibv_flow_attr);
+ }
+ }
+ /* Third step. Conversion parse, fill the specifications. */
+ parser->inner = 0;
+ for (; items->type != RTE_FLOW_ITEM_TYPE_END; ++items) {
+ if (items->type == RTE_FLOW_ITEM_TYPE_VOID)
+ continue;
+ cur_item = &mlx5_flow_items[items->type];
+ ret = cur_item->convert(items,
+ (cur_item->default_mask ?
+ cur_item->default_mask :
+ cur_item->mask),
+ parser);
+ if (ret) {
+ rte_flow_error_set(error, ret,
+ RTE_FLOW_ERROR_TYPE_ITEM,
+ items, "item not supported");
+ goto exit_free;
+ }
+ }
+ if (parser->mark)
+ mlx5_flow_create_flag_mark(parser, parser->mark_id);
+ if (parser->count && parser->create) {
+ mlx5_flow_create_count(priv, parser);
+ if (!parser->cs)
+ goto exit_count_error;
+ }
+ /*
+ * Last step. Complete missing specification to reach the RSS
+ * configuration.
+ */
+ if (parser->queues_n > 1) {
+ priv_flow_convert_finalise(priv, parser);
+ } else {
+ /*
+ * Action queue have their priority overridden with
+ * Ethernet priority, this priority needs to be adjusted to
+ * their most specific layer priority.
+ */
+ parser->queue[HASH_RXQ_ETH].ibv_attr->priority =
+ attr->priority +
+ hash_rxq_init[parser->layer].flow_priority;
+ }
+ if (parser->allmulti &&
+ parser->layer == HASH_RXQ_ETH) {
+ for (i = 0; i != hash_rxq_init_n; ++i) {
+ if (!parser->queue[i].ibv_attr)
+ continue;
+ if (parser->queue[i].ibv_attr->num_of_specs != 1)
+ break;
+ parser->queue[i].ibv_attr->type =
+ IBV_FLOW_ATTR_MC_DEFAULT;
+ }
+ }
+exit_free:
+ /* Only verification is expected, all resources should be released. */
+ if (!parser->create) {
+ for (i = 0; i != hash_rxq_init_n; ++i) {
+ if (parser->queue[i].ibv_attr) {
+ rte_free(parser->queue[i].ibv_attr);
+ parser->queue[i].ibv_attr = NULL;
+ }
+ }
+ }
+ return ret;
+exit_enomem:
+ for (i = 0; i != hash_rxq_init_n; ++i) {
+ if (parser->queue[i].ibv_attr) {
+ rte_free(parser->queue[i].ibv_attr);
+ parser->queue[i].ibv_attr = NULL;
+ }
+ }
+ rte_flow_error_set(error, ENOTSUP, RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+ NULL, "cannot allocate verbs spec attributes.");
+ return ret;
+exit_count_error:
+ rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+ NULL, "cannot create counter.");
+ return rte_errno;
+}
+
+/**
+ * Copy the specification created into the flow.
+ *
+ * @param parser
+ * Internal parser structure.
+ * @param src
+ * Create specification.
+ * @param size
+ * Size in bytes of the specification to copy.
+ */
+static void
+mlx5_flow_create_copy(struct mlx5_flow_parse *parser, void *src,
+ unsigned int size)
+{
+ unsigned int i;
+ void *dst;
+
+ for (i = 0; i != hash_rxq_init_n; ++i) {
+ if (!parser->queue[i].ibv_attr)
+ continue;
+ /* Specification must be the same l3 type or none. */
+ if (parser->layer == HASH_RXQ_ETH ||
+ (hash_rxq_init[parser->layer].ip_version ==
+ hash_rxq_init[i].ip_version) ||
+ (hash_rxq_init[i].ip_version == 0)) {
+ dst = (void *)((uintptr_t)parser->queue[i].ibv_attr +
+ parser->queue[i].offset);
+ memcpy(dst, src, size);
+ ++parser->queue[i].ibv_attr->num_of_specs;
+ parser->queue[i].offset += size;
+ }
+ }
+}
+
+/**
+ * Convert Ethernet item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param default_mask[in]
+ * Default bit-masks to use when item->mask is not provided.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_eth(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data)
+{
+ const struct rte_flow_item_eth *spec = item->spec;
+ const struct rte_flow_item_eth *mask = item->mask;
+ struct mlx5_flow_parse *parser = (struct mlx5_flow_parse *)data;
+ const unsigned int eth_size = sizeof(struct ibv_flow_spec_eth);
+ struct ibv_flow_spec_eth eth = {
+ .type = parser->inner | IBV_FLOW_SPEC_ETH,
+ .size = eth_size,
+ };
+
+ /* Don't update layer for the inner pattern. */
+ if (!parser->inner)
+ parser->layer = HASH_RXQ_ETH;
+ if (spec) {
+ unsigned int i;
+
+ if (!mask)
+ mask = default_mask;
+ memcpy(ð.val.dst_mac, spec->dst.addr_bytes, ETHER_ADDR_LEN);
+ memcpy(ð.val.src_mac, spec->src.addr_bytes, ETHER_ADDR_LEN);
+ eth.val.ether_type = spec->type;
+ memcpy(ð.mask.dst_mac, mask->dst.addr_bytes, ETHER_ADDR_LEN);
+ memcpy(ð.mask.src_mac, mask->src.addr_bytes, ETHER_ADDR_LEN);
+ eth.mask.ether_type = mask->type;
+ /* Remove unwanted bits from values. */
+ for (i = 0; i < ETHER_ADDR_LEN; ++i) {
+ eth.val.dst_mac[i] &= eth.mask.dst_mac[i];
+ eth.val.src_mac[i] &= eth.mask.src_mac[i];
+ }
+ eth.val.ether_type &= eth.mask.ether_type;
+ }
+ mlx5_flow_create_copy(parser, ð, eth_size);
+ parser->allmulti = eth.val.dst_mac[0] & 1;
+ return 0;
+}
+
+/**
+ * Convert VLAN item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param default_mask[in]
+ * Default bit-masks to use when item->mask is not provided.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_vlan(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data)
+{
+ const struct rte_flow_item_vlan *spec = item->spec;
+ const struct rte_flow_item_vlan *mask = item->mask;
+ struct mlx5_flow_parse *parser = (struct mlx5_flow_parse *)data;
+ struct ibv_flow_spec_eth *eth;
+ const unsigned int eth_size = sizeof(struct ibv_flow_spec_eth);
+
+ if (spec) {
+ unsigned int i;
+ if (!mask)
+ mask = default_mask;
+
+ for (i = 0; i != hash_rxq_init_n; ++i) {
+ if (!parser->queue[i].ibv_attr)
+ continue;
+
+ eth = (void *)((uintptr_t)parser->queue[i].ibv_attr +
+ parser->queue[i].offset - eth_size);
+ eth->val.vlan_tag = spec->tci;
+ eth->mask.vlan_tag = mask->tci;
+ eth->val.vlan_tag &= eth->mask.vlan_tag;
+ }
+ }
+ return 0;
+}
+
+/**
+ * Convert IPv4 item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param default_mask[in]
+ * Default bit-masks to use when item->mask is not provided.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_ipv4(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data)
+{
+ const struct rte_flow_item_ipv4 *spec = item->spec;
+ const struct rte_flow_item_ipv4 *mask = item->mask;
+ struct mlx5_flow_parse *parser = (struct mlx5_flow_parse *)data;
+ unsigned int ipv4_size = sizeof(struct ibv_flow_spec_ipv4_ext);
+ struct ibv_flow_spec_ipv4_ext ipv4 = {
+ .type = parser->inner | IBV_FLOW_SPEC_IPV4_EXT,
+ .size = ipv4_size,
+ };
+
+ /* Don't update layer for the inner pattern. */
+ if (!parser->inner)
+ parser->layer = HASH_RXQ_IPV4;
+ if (spec) {
+ if (!mask)
+ mask = default_mask;
+ ipv4.val = (struct ibv_flow_ipv4_ext_filter){
+ .src_ip = spec->hdr.src_addr,
+ .dst_ip = spec->hdr.dst_addr,
+ .proto = spec->hdr.next_proto_id,
+ .tos = spec->hdr.type_of_service,
+ };
+ ipv4.mask = (struct ibv_flow_ipv4_ext_filter){
+ .src_ip = mask->hdr.src_addr,
+ .dst_ip = mask->hdr.dst_addr,
+ .proto = mask->hdr.next_proto_id,
+ .tos = mask->hdr.type_of_service,
+ };
+ /* Remove unwanted bits from values. */
+ ipv4.val.src_ip &= ipv4.mask.src_ip;
+ ipv4.val.dst_ip &= ipv4.mask.dst_ip;
+ ipv4.val.proto &= ipv4.mask.proto;
+ ipv4.val.tos &= ipv4.mask.tos;
+ }
+ mlx5_flow_create_copy(parser, &ipv4, ipv4_size);
+ return 0;
+}
+
+/**
+ * Convert IPv6 item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param default_mask[in]
+ * Default bit-masks to use when item->mask is not provided.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_ipv6(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data)
+{
+ const struct rte_flow_item_ipv6 *spec = item->spec;
+ const struct rte_flow_item_ipv6 *mask = item->mask;
+ struct mlx5_flow_parse *parser = (struct mlx5_flow_parse *)data;
+ unsigned int ipv6_size = sizeof(struct ibv_flow_spec_ipv6);
+ struct ibv_flow_spec_ipv6 ipv6 = {
+ .type = parser->inner | IBV_FLOW_SPEC_IPV6,
+ .size = ipv6_size,
+ };
+
+ /* Don't update layer for the inner pattern. */
+ if (!parser->inner)
+ parser->layer = HASH_RXQ_IPV6;
+ if (spec) {
+ unsigned int i;
+ uint32_t vtc_flow_val;
+ uint32_t vtc_flow_mask;
+
+ if (!mask)
+ mask = default_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_create_copy(parser, &ipv6, ipv6_size);
+ return 0;
+}
+
+/**
+ * Convert UDP item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param default_mask[in]
+ * Default bit-masks to use when item->mask is not provided.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_udp(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data)
+{
+ const struct rte_flow_item_udp *spec = item->spec;
+ const struct rte_flow_item_udp *mask = item->mask;
+ struct mlx5_flow_parse *parser = (struct mlx5_flow_parse *)data;
+ unsigned int udp_size = sizeof(struct ibv_flow_spec_tcp_udp);
+ struct ibv_flow_spec_tcp_udp udp = {
+ .type = parser->inner | IBV_FLOW_SPEC_UDP,
+ .size = udp_size,
+ };
+
+ /* Don't update layer for the inner pattern. */
+ if (!parser->inner) {
+ if (parser->layer == HASH_RXQ_IPV4)
+ parser->layer = HASH_RXQ_UDPV4;
+ else
+ parser->layer = HASH_RXQ_UDPV6;
+ }
+ if (spec) {
+ if (!mask)
+ mask = default_mask;
+ udp.val.dst_port = spec->hdr.dst_port;
+ udp.val.src_port = spec->hdr.src_port;
+ udp.mask.dst_port = mask->hdr.dst_port;
+ udp.mask.src_port = mask->hdr.src_port;
+ /* Remove unwanted bits from values. */
+ udp.val.src_port &= udp.mask.src_port;
+ udp.val.dst_port &= udp.mask.dst_port;
+ }
+ mlx5_flow_create_copy(parser, &udp, udp_size);
+ return 0;
+}
+
+/**
+ * Convert TCP item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param default_mask[in]
+ * Default bit-masks to use when item->mask is not provided.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_tcp(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data)
+{
+ const struct rte_flow_item_tcp *spec = item->spec;
+ const struct rte_flow_item_tcp *mask = item->mask;
+ struct mlx5_flow_parse *parser = (struct mlx5_flow_parse *)data;
+ unsigned int tcp_size = sizeof(struct ibv_flow_spec_tcp_udp);
+ struct ibv_flow_spec_tcp_udp tcp = {
+ .type = parser->inner | IBV_FLOW_SPEC_TCP,
+ .size = tcp_size,
+ };
+
+ /* Don't update layer for the inner pattern. */
+ if (!parser->inner) {
+ if (parser->layer == HASH_RXQ_IPV4)
+ parser->layer = HASH_RXQ_TCPV4;
+ else
+ parser->layer = HASH_RXQ_TCPV6;
+ }
+ if (spec) {
+ if (!mask)
+ mask = default_mask;
+ tcp.val.dst_port = spec->hdr.dst_port;
+ tcp.val.src_port = spec->hdr.src_port;
+ tcp.mask.dst_port = mask->hdr.dst_port;
+ tcp.mask.src_port = mask->hdr.src_port;
+ /* Remove unwanted bits from values. */
+ tcp.val.src_port &= tcp.mask.src_port;
+ tcp.val.dst_port &= tcp.mask.dst_port;
+ }
+ mlx5_flow_create_copy(parser, &tcp, tcp_size);
+ return 0;
+}
+
+/**
+ * Convert VXLAN item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param default_mask[in]
+ * Default bit-masks to use when item->mask is not provided.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_vxlan(const struct rte_flow_item *item,
+ const void *default_mask,
+ void *data)
+{
+ const struct rte_flow_item_vxlan *spec = item->spec;
+ const struct rte_flow_item_vxlan *mask = item->mask;
+ struct mlx5_flow_parse *parser = (struct mlx5_flow_parse *)data;
+ unsigned int size = sizeof(struct ibv_flow_spec_tunnel);
+ struct ibv_flow_spec_tunnel vxlan = {
+ .type = parser->inner | IBV_FLOW_SPEC_VXLAN_TUNNEL,
+ .size = size,
+ };
+ union vni {
+ uint32_t vlan_id;
+ uint8_t vni[4];
+ } id;
+
+ id.vni[0] = 0;
+ parser->inner = IBV_FLOW_SPEC_INNER;
+ if (spec) {
+ if (!mask)
+ mask = default_mask;
+ memcpy(&id.vni[1], spec->vni, 3);
+ vxlan.val.tunnel_id = id.vlan_id;
+ memcpy(&id.vni[1], mask->vni, 3);
+ vxlan.mask.tunnel_id = id.vlan_id;
+ /* Remove unwanted bits from values. */
+ vxlan.val.tunnel_id &= vxlan.mask.tunnel_id;
+ }
+ /*
+ * Tunnel id 0 is equivalent as not adding a VXLAN layer, if only this
+ * layer is defined in the Verbs specification it is interpreted as
+ * wildcard and all packets will match this rule, if it follows a full
+ * stack layer (ex: eth / ipv4 / udp), all packets matching the layers
+ * before will also match this rule.
+ * To avoid such situation, VNI 0 is currently refused.
+ */
+ if (!vxlan.val.tunnel_id)
+ return EINVAL;
+ mlx5_flow_create_copy(parser, &vxlan, size);
+ return 0;
+}
+
+/**
+ * Convert mark/flag action to Verbs specification.
+ *
+ * @param parser
+ * Internal parser structure.
+ * @param mark_id
+ * Mark identifier.
+ */
+static int
+mlx5_flow_create_flag_mark(struct mlx5_flow_parse *parser, uint32_t mark_id)
+{
+ unsigned int size = sizeof(struct ibv_flow_spec_action_tag);
+ struct ibv_flow_spec_action_tag tag = {
+ .type = IBV_FLOW_SPEC_ACTION_TAG,
+ .size = size,
+ .tag_id = mlx5_flow_mark_set(mark_id),
+ };
+
+ assert(parser->mark);
+ mlx5_flow_create_copy(parser, &tag, size);
+ return 0;
+}
+
+/**
+ * Convert count action to Verbs specification.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param parser
+ * Pointer to MLX5 flow parser structure.
+ *
+ * @return
+ * 0 on success, errno value on failure.
+ */
+static int
+mlx5_flow_create_count(struct priv *priv __rte_unused,
+ struct mlx5_flow_parse *parser __rte_unused)
+{
+#ifdef HAVE_IBV_DEVICE_COUNTERS_SET_SUPPORT
+ unsigned int size = sizeof(struct ibv_flow_spec_counter_action);
+ struct ibv_counter_set_init_attr init_attr = {0};
+ struct ibv_flow_spec_counter_action counter = {
+ .type = IBV_FLOW_SPEC_ACTION_COUNT,
+ .size = size,
+ .counter_set_handle = 0,
+ };
+
+ init_attr.counter_set_id = 0;
+ parser->cs = ibv_create_counter_set(priv->ctx, &init_attr);
+ if (!parser->cs)
+ return EINVAL;
+ counter.counter_set_handle = parser->cs->handle;
+ mlx5_flow_create_copy(parser, &counter, size);
+#endif
+ return 0;
+}
+
+/**
+ * Complete flow rule creation with a drop queue.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param parser
+ * Internal parser structure.
+ * @param flow
+ * Pointer to the rte_flow.
+ * @param[out] error
+ * Perform verbose error reporting if not NULL.
+ *
+ * @return
+ * 0 on success, errno value on failure.
+ */
+static int
+priv_flow_create_action_queue_drop(struct priv *priv,
+ struct mlx5_flow_parse *parser,
+ struct rte_flow *flow,
+ struct rte_flow_error *error)
+{
+ struct ibv_flow_spec_action_drop *drop;
+ unsigned int size = sizeof(struct ibv_flow_spec_action_drop);
+ int err = 0;
+
+ assert(priv->pd);
+ assert(priv->ctx);
+ flow->drop = 1;
+ drop = (void *)((uintptr_t)parser->queue[HASH_RXQ_ETH].ibv_attr +
+ parser->queue[HASH_RXQ_ETH].offset);
+ *drop = (struct ibv_flow_spec_action_drop){
+ .type = IBV_FLOW_SPEC_ACTION_DROP,
+ .size = size,
+ };
+ ++parser->queue[HASH_RXQ_ETH].ibv_attr->num_of_specs;
+ parser->queue[HASH_RXQ_ETH].offset += size;
+ flow->frxq[HASH_RXQ_ETH].ibv_attr =
+ parser->queue[HASH_RXQ_ETH].ibv_attr;
+ if (parser->count)
+ flow->cs = parser->cs;
+ if (!priv->dev->data->dev_started)
+ return 0;
+ parser->queue[HASH_RXQ_ETH].ibv_attr = NULL;
+ flow->frxq[HASH_RXQ_ETH].ibv_flow =
+ ibv_create_flow(priv->flow_drop_queue->qp,
+ flow->frxq[HASH_RXQ_ETH].ibv_attr);
+ if (!flow->frxq[HASH_RXQ_ETH].ibv_flow) {
+ rte_flow_error_set(error, ENOMEM, RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "flow rule creation failure");
+ err = ENOMEM;
+ goto error;
+ }
+ return 0;
+error:
+ assert(flow);
+ if (flow->frxq[HASH_RXQ_ETH].ibv_flow) {
+ claim_zero(ibv_destroy_flow(flow->frxq[HASH_RXQ_ETH].ibv_flow));
+ flow->frxq[HASH_RXQ_ETH].ibv_flow = NULL;
+ }
+ if (flow->frxq[HASH_RXQ_ETH].ibv_attr) {
+ rte_free(flow->frxq[HASH_RXQ_ETH].ibv_attr);
+ flow->frxq[HASH_RXQ_ETH].ibv_attr = NULL;
+ }
+ if (flow->cs) {
+ claim_zero(ibv_destroy_counter_set(flow->cs));
+ flow->cs = NULL;
+ parser->cs = NULL;
+ }
+ return err;
+}
+
+/**
+ * Create hash Rx queues when RSS is enabled.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param parser
+ * Internal parser structure.
+ * @param flow
+ * Pointer to the rte_flow.
+ * @param[out] error
+ * Perform verbose error reporting if not NULL.
+ *
+ * @return
+ * 0 on success, a errno value otherwise and rte_errno is set.
+ */
+static int
+priv_flow_create_action_queue_rss(struct priv *priv,
+ struct mlx5_flow_parse *parser,
+ struct rte_flow *flow,
+ struct rte_flow_error *error)
+{
+ unsigned int i;
+
+ for (i = 0; i != hash_rxq_init_n; ++i) {
+ uint64_t hash_fields;
+
+ if (!parser->queue[i].ibv_attr)
+ continue;
+ flow->frxq[i].ibv_attr = parser->queue[i].ibv_attr;
+ parser->queue[i].ibv_attr = NULL;
+ hash_fields = hash_rxq_init[i].hash_fields;
+ if (!priv->dev->data->dev_started)
+ continue;
+ flow->frxq[i].hrxq =
+ mlx5_priv_hrxq_get(priv,
+ parser->rss_conf.rss_key,
+ parser->rss_conf.rss_key_len,
+ hash_fields,
+ parser->queues,
+ parser->queues_n);
+ if (flow->frxq[i].hrxq)
+ continue;
+ flow->frxq[i].hrxq =
+ mlx5_priv_hrxq_new(priv,
+ parser->rss_conf.rss_key,
+ parser->rss_conf.rss_key_len,
+ hash_fields,
+ parser->queues,
+ parser->queues_n);
+ if (!flow->frxq[i].hrxq) {
+ rte_flow_error_set(error, ENOMEM,
+ RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "cannot create hash rxq");
+ return ENOMEM;
+ }
+ }
+ return 0;
+}
+
+/**
+ * Complete flow rule creation.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param parser
+ * Internal parser structure.
+ * @param flow
+ * Pointer to the rte_flow.
+ * @param[out] error
+ * Perform verbose error reporting if not NULL.
+ *
+ * @return
+ * 0 on success, a errno value otherwise and rte_errno is set.
+ */
+static int
+priv_flow_create_action_queue(struct priv *priv,
+ struct mlx5_flow_parse *parser,
+ struct rte_flow *flow,
+ struct rte_flow_error *error)
+{
+ int err = 0;
+ unsigned int i;
+
+ assert(priv->pd);
+ assert(priv->ctx);
+ assert(!parser->drop);
+ err = priv_flow_create_action_queue_rss(priv, parser, flow, error);
+ if (err)
+ goto error;
+ if (parser->count)
+ flow->cs = parser->cs;
+ if (!priv->dev->data->dev_started)
+ return 0;
+ for (i = 0; i != hash_rxq_init_n; ++i) {
+ if (!flow->frxq[i].hrxq)
+ continue;
+ flow->frxq[i].ibv_flow =
+ ibv_create_flow(flow->frxq[i].hrxq->qp,
+ flow->frxq[i].ibv_attr);
+ if (!flow->frxq[i].ibv_flow) {
+ rte_flow_error_set(error, ENOMEM,
+ RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "flow rule creation failure");
+ err = ENOMEM;
+ goto error;
+ }
+ DEBUG("%p type %d QP %p ibv_flow %p",
+ (void *)flow, i,
+ (void *)flow->frxq[i].hrxq,
+ (void *)flow->frxq[i].ibv_flow);
+ }
+ for (i = 0; i != parser->queues_n; ++i) {
+ struct mlx5_rxq_data *q =
+ (*priv->rxqs)[parser->queues[i]];
+
+ q->mark |= parser->mark;
+ }
+ return 0;
+error:
+ assert(flow);
+ for (i = 0; i != hash_rxq_init_n; ++i) {
+ if (flow->frxq[i].ibv_flow) {
+ struct ibv_flow *ibv_flow = flow->frxq[i].ibv_flow;
+
+ claim_zero(ibv_destroy_flow(ibv_flow));
+ }
+ if (flow->frxq[i].hrxq)
+ mlx5_priv_hrxq_release(priv, flow->frxq[i].hrxq);
+ if (flow->frxq[i].ibv_attr)
+ rte_free(flow->frxq[i].ibv_attr);
+ }
+ if (flow->cs) {
+ claim_zero(ibv_destroy_counter_set(flow->cs));
+ flow->cs = NULL;
+ parser->cs = NULL;
+ }
+ return err;
+}
+
+/**
+ * Convert a flow.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param list
+ * Pointer to a TAILQ flow list.
+ * @param[in] attr
+ * Flow rule attributes.
+ * @param[in] pattern
+ * Pattern specification (list terminated by the END pattern item).
+ * @param[in] actions
+ * Associated actions (list terminated by the END action).
+ * @param[out] error
+ * Perform verbose error reporting if not NULL.
+ *
+ * @return
+ * A flow on success, NULL otherwise.
+ */
+static struct rte_flow *
+priv_flow_create(struct priv *priv,
+ struct mlx5_flows *list,
+ const struct rte_flow_attr *attr,
+ const struct rte_flow_item items[],
+ const struct rte_flow_action actions[],
+ struct rte_flow_error *error)
+{
+ struct mlx5_flow_parse parser = { .create = 1, };
+ struct rte_flow *flow = NULL;
+ unsigned int i;
+ int err;
+
+ err = priv_flow_convert(priv, attr, items, actions, error, &parser);
+ if (err)
+ goto exit;
+ flow = rte_calloc(__func__, 1,
+ sizeof(*flow) + parser.queues_n * sizeof(uint16_t),
+ 0);
+ if (!flow) {
+ rte_flow_error_set(error, ENOMEM,
+ RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
+ NULL,
+ "cannot allocate flow memory");
+ return NULL;
+ }
+ /* Copy queues configuration. */
+ flow->queues = (uint16_t (*)[])(flow + 1);
+ memcpy(flow->queues, parser.queues, parser.queues_n * sizeof(uint16_t));
+ flow->queues_n = parser.queues_n;
+ flow->mark = parser.mark;
+ /* Copy RSS configuration. */
+ flow->rss_conf = parser.rss_conf;
+ flow->rss_conf.rss_key = flow->rss_key;
+ memcpy(flow->rss_key, parser.rss_key, parser.rss_conf.rss_key_len);
+ /* finalise the flow. */
+ if (parser.drop)
+ err = priv_flow_create_action_queue_drop(priv, &parser, flow,
+ error);
+ else
+ err = priv_flow_create_action_queue(priv, &parser, flow, error);
+ if (err)
+ goto exit;
+ TAILQ_INSERT_TAIL(list, flow, next);
+ DEBUG("Flow created %p", (void *)flow);
+ return flow;
+exit:
+ for (i = 0; i != hash_rxq_init_n; ++i) {
+ if (parser.queue[i].ibv_attr)
+ rte_free(parser.queue[i].ibv_attr);
+ }
+ rte_free(flow);
+ return NULL;
+}