@@ -136,6 +136,7 @@ struct priv {
unsigned int reta_idx_n; /* RETA index size. */
struct fdir_filter_list *fdir_filter_list; /* Flow director rules. */
struct fdir_queue *fdir_drop_queue; /* Flow director drop queue. */
+ LIST_HEAD(mlx5_flows, rte_flow) flows; /* RTE Flow rules. */
uint32_t link_speed_capa; /* Link speed capabilities. */
rte_spinlock_t lock; /* Lock for control functions. */
};
@@ -283,5 +284,7 @@ struct rte_flow *mlx5_flow_create(struct rte_eth_dev *,
int mlx5_flow_destroy(struct rte_eth_dev *, struct rte_flow *,
struct rte_flow_error *);
int mlx5_flow_flush(struct rte_eth_dev *, struct rte_flow_error *);
+int priv_flow_start(struct priv *);
+void priv_flow_stop(struct priv *);
#endif /* RTE_PMD_MLX5_H_ */
@@ -31,12 +31,380 @@
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
+#include <sys/queue.h>
+#include <string.h>
+
+/* Verbs header. */
+/* ISO C doesn't support unnamed structs/unions, disabling -pedantic. */
+#ifdef PEDANTIC
+#pragma GCC diagnostic ignored "-Wpedantic"
+#endif
+#include <infiniband/verbs.h>
+#ifdef PEDANTIC
+#pragma GCC diagnostic error "-Wpedantic"
+#endif
+
#include <rte_ethdev.h>
#include <rte_flow.h>
#include <rte_flow_driver.h>
+#include <rte_malloc.h>
#include "mlx5.h"
+static int
+mlx5_flow_create_eth(const struct rte_flow_item *item, void *data);
+
+static int
+mlx5_flow_create_ipv4(const struct rte_flow_item *item, void *data);
+
+static int
+mlx5_flow_create_ipv6(const struct rte_flow_item *item, void *data);
+
+static int
+mlx5_flow_create_udp(const struct rte_flow_item *item, void *data);
+
+static int
+mlx5_flow_create_tcp(const struct rte_flow_item *item, void *data);
+
+struct rte_flow {
+ LIST_ENTRY(rte_flow) next; /**< Pointer to the next flow structure. */
+ struct ibv_exp_flow_attr *ibv_attr; /**< Pointer to Verbs attributes. */
+ struct ibv_exp_rwq_ind_table *ind_table; /**< Indirection table. */
+ struct ibv_qp *qp; /**< Verbs queue pair. */
+ struct ibv_exp_flow *ibv_flow; /**< Verbs flow. */
+ struct ibv_exp_wq *wq; /**< Verbs work queue. */
+ struct ibv_cq *cq; /**< Verbs completion queue. */
+ struct rxq *rxq; /**< Pointer to the queue, NULL if drop 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 following items. */
+ const enum rte_flow_item_type *const 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;
+ /** 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 data
+ * Internal structure to store the conversion.
+ *
+ * @return
+ * 0 on success, negative value otherwise.
+ */
+ int (*convert)(const struct rte_flow_item *item, void *data);
+ /** Size in bytes of the destination structure. */
+ const unsigned int dst_sz;
+};
+
+/** 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_END,
+};
+
+/** Graph of supported items and associated actions. */
+static const struct mlx5_flow_items mlx5_flow_items[] = {
+ [RTE_FLOW_ITEM_TYPE_VOID] = {
+ .items = ITEMS(RTE_FLOW_ITEM_TYPE_VOID,
+ RTE_FLOW_ITEM_TYPE_ETH),
+ .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",
+ },
+ .mask_sz = sizeof(struct rte_flow_item_eth),
+ .convert = mlx5_flow_create_eth,
+ .dst_sz = sizeof(struct ibv_exp_flow_spec_eth),
+ },
+ [RTE_FLOW_ITEM_TYPE_ETH] = {
+ .items = ITEMS(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",
+ },
+ .mask_sz = sizeof(struct rte_flow_item_eth),
+ .convert = mlx5_flow_create_eth,
+ .dst_sz = sizeof(struct ibv_exp_flow_spec_eth),
+ },
+ [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,
+ },
+ },
+ .mask_sz = sizeof(struct rte_flow_item_ipv4),
+ .convert = mlx5_flow_create_ipv4,
+ .dst_sz = sizeof(struct ibv_exp_flow_spec_ipv4),
+ },
+ [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,
+ },
+ },
+ },
+ .mask_sz = sizeof(struct rte_flow_item_ipv6),
+ .convert = mlx5_flow_create_ipv6,
+ .dst_sz = sizeof(struct ibv_exp_flow_spec_ipv6),
+ },
+ [RTE_FLOW_ITEM_TYPE_UDP] = {
+ .actions = valid_actions,
+ .mask = &(const struct rte_flow_item_udp){
+ .hdr = {
+ .src_port = -1,
+ .dst_port = -1,
+ },
+ },
+ .mask_sz = sizeof(struct rte_flow_item_udp),
+ .convert = mlx5_flow_create_udp,
+ .dst_sz = sizeof(struct ibv_exp_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,
+ },
+ },
+ .mask_sz = sizeof(struct rte_flow_item_tcp),
+ .convert = mlx5_flow_create_tcp,
+ .dst_sz = sizeof(struct ibv_exp_flow_spec_tcp_udp),
+ },
+};
+
+/** Structure to pass to the conversion function. */
+struct mlx5_flow {
+ struct ibv_exp_flow_attr *ibv_attr; /**< Verbs attribute. */
+ unsigned int offset; /**< Offset in bytes in the ibv_attr buffer. */
+};
+
+struct mlx5_flow_action {
+ uint32_t queue:1; /**< Target is a receive queue. */
+ uint32_t drop:1; /**< Target is a drop queue. */
+ uint32_t queue_id; /**< Identifier of the queue. */
+};
+
+/**
+ * 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) {
+ 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->mask;
+
+ 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;
+}
+
+/**
+ * Validate 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] flow
+ * Flow structure to update.
+ *
+ * @return
+ * 0 on success, a negative errno value otherwise and rte_errno is set.
+ */
+static int
+priv_flow_validate(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 *flow)
+{
+ const struct mlx5_flow_items *cur_item = mlx5_flow_items;
+ struct mlx5_flow_action action = {
+ .queue = 0,
+ .drop = 0,
+ };
+
+ (void)priv;
+ 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) {
+ 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;
+ }
+ for (; items->type != RTE_FLOW_ITEM_TYPE_END; ++items) {
+ const struct mlx5_flow_items *token = NULL;
+ unsigned int i;
+ 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,
+ sizeof(cur_item->mask_sz));
+ if (err)
+ goto exit_item_not_supported;
+ if (flow->ibv_attr && cur_item->convert) {
+ err = cur_item->convert(items, flow);
+ if (err)
+ goto exit_item_not_supported;
+ }
+ flow->offset += cur_item->dst_sz;
+ }
+ 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) {
+ action.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;
+
+ if (!queue || (queue->index > (priv->rxqs_n - 1)))
+ goto exit_action_not_supported;
+ action.queue = 1;
+ } else {
+ goto exit_action_not_supported;
+ }
+ }
+ if (!action.queue && !action.drop) {
+ rte_flow_error_set(error, ENOTSUP, RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "no valid action");
+ return -rte_errno;
+ }
+ return 0;
+exit_item_not_supported:
+ rte_flow_error_set(error, ENOTSUP, RTE_FLOW_ERROR_TYPE_ITEM,
+ items, "item not supported");
+ return -rte_errno;
+exit_action_not_supported:
+ rte_flow_error_set(error, ENOTSUP, RTE_FLOW_ERROR_TYPE_ACTION,
+ actions, "action not supported");
+ return -rte_errno;
+}
+
/**
* Validate a flow supported by the NIC.
*
@@ -50,15 +418,417 @@ mlx5_flow_validate(struct rte_eth_dev *dev,
const struct rte_flow_action actions[],
struct rte_flow_error *error)
{
- (void)dev;
- (void)attr;
- (void)items;
- (void)actions;
- (void)error;
- rte_flow_error_set(error, ENOTSUP,
- RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
- NULL, "not implemented yet");
- return -rte_errno;
+ struct priv *priv = dev->data->dev_private;
+ int ret;
+ struct mlx5_flow flow = { .offset = sizeof(struct ibv_exp_flow_attr) };
+
+ priv_lock(priv);
+ ret = priv_flow_validate(priv, attr, items, actions, error, &flow);
+ priv_unlock(priv);
+ return ret;
+}
+
+/**
+ * Convert Ethernet item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_eth(const struct rte_flow_item *item, void *data)
+{
+ const struct rte_flow_item_eth *spec = item->spec;
+ const struct rte_flow_item_eth *mask = item->mask;
+ struct mlx5_flow *flow = (struct mlx5_flow *)data;
+ struct ibv_exp_flow_spec_eth *eth;
+ const unsigned int eth_size = sizeof(struct ibv_exp_flow_spec_eth);
+ unsigned int i;
+
+ eth = (void *)((uintptr_t)flow->ibv_attr + flow->offset);
+ *eth = (struct ibv_exp_flow_spec_eth) {
+ .type = IBV_EXP_FLOW_SPEC_ETH,
+ .size = eth_size,
+ };
+ if (spec) {
+ memcpy(eth->val.dst_mac, spec->dst.addr_bytes, ETHER_ADDR_LEN);
+ memcpy(eth->val.src_mac, spec->src.addr_bytes, ETHER_ADDR_LEN);
+ }
+ if (mask) {
+ memcpy(eth->mask.dst_mac, mask->dst.addr_bytes, ETHER_ADDR_LEN);
+ memcpy(eth->mask.src_mac, mask->src.addr_bytes, ETHER_ADDR_LEN);
+ }
+ /* 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];
+ }
+ /* Finalise the flow. */
+ ++flow->ibv_attr->num_of_specs;
+ flow->ibv_attr->priority = 2;
+ return 0;
+}
+
+/**
+ * Convert IPv4 item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_ipv4(const struct rte_flow_item *item, void *data)
+{
+ const struct rte_flow_item_ipv4 *spec = item->spec;
+ const struct rte_flow_item_ipv4 *mask = item->mask;
+ struct mlx5_flow *flow = (struct mlx5_flow *)data;
+ struct ibv_exp_flow_spec_ipv4 *ipv4;
+ unsigned int ipv4_size = sizeof(struct ibv_exp_flow_spec_ipv4);
+
+ ipv4 = (void *)((uintptr_t)flow->ibv_attr + flow->offset);
+ *ipv4 = (struct ibv_exp_flow_spec_ipv4) {
+ .type = IBV_EXP_FLOW_SPEC_IPV4,
+ .size = ipv4_size,
+ };
+ if (spec) {
+ ipv4->val = (struct ibv_exp_flow_ipv4_filter){
+ .src_ip = spec->hdr.src_addr,
+ .dst_ip = spec->hdr.dst_addr,
+ };
+ }
+ if (mask) {
+ ipv4->mask = (struct ibv_exp_flow_ipv4_filter){
+ .src_ip = mask->hdr.src_addr,
+ .dst_ip = mask->hdr.dst_addr,
+ };
+ }
+ /* Remove unwanted bits from values. */
+ ipv4->val.src_ip &= ipv4->mask.src_ip;
+ ipv4->val.dst_ip &= ipv4->mask.dst_ip;
+ ++flow->ibv_attr->num_of_specs;
+ flow->ibv_attr->priority = 1;
+ return 0;
+}
+
+/**
+ * Convert IPv6 item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_ipv6(const struct rte_flow_item *item, void *data)
+{
+ const struct rte_flow_item_ipv6 *spec = item->spec;
+ const struct rte_flow_item_ipv6 *mask = item->mask;
+ struct mlx5_flow *flow = (struct mlx5_flow *)data;
+ struct ibv_exp_flow_spec_ipv6 *ipv6;
+ unsigned int ipv6_size = sizeof(struct ibv_exp_flow_spec_ipv6);
+ unsigned int i;
+
+ ipv6 = (void *)((uintptr_t)flow->ibv_attr + flow->offset);
+ *ipv6 = (struct ibv_exp_flow_spec_ipv6) {
+ .type = IBV_EXP_FLOW_SPEC_IPV6,
+ .size = ipv6_size,
+ };
+ if (spec) {
+ 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));
+ }
+ if (mask) {
+ 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));
+ }
+ /* 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];
+ }
+ ++flow->ibv_attr->num_of_specs;
+ flow->ibv_attr->priority = 1;
+ return 0;
+}
+
+/**
+ * Convert UDP item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_udp(const struct rte_flow_item *item, void *data)
+{
+ const struct rte_flow_item_udp *spec = item->spec;
+ const struct rte_flow_item_udp *mask = item->mask;
+ struct mlx5_flow *flow = (struct mlx5_flow *)data;
+ struct ibv_exp_flow_spec_tcp_udp *udp;
+ unsigned int udp_size = sizeof(struct ibv_exp_flow_spec_tcp_udp);
+
+ udp = (void *)((uintptr_t)flow->ibv_attr + flow->offset);
+ *udp = (struct ibv_exp_flow_spec_tcp_udp) {
+ .type = IBV_EXP_FLOW_SPEC_UDP,
+ .size = udp_size,
+ };
+ if (spec) {
+ udp->val.dst_port = spec->hdr.dst_port;
+ udp->val.src_port = spec->hdr.src_port;
+ }
+ if (mask) {
+ 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;
+ ++flow->ibv_attr->num_of_specs;
+ flow->ibv_attr->priority = 0;
+ return 0;
+}
+
+/**
+ * Convert TCP item to Verbs specification.
+ *
+ * @param item[in]
+ * Item specification.
+ * @param data[in, out]
+ * User structure.
+ */
+static int
+mlx5_flow_create_tcp(const struct rte_flow_item *item, void *data)
+{
+ const struct rte_flow_item_tcp *spec = item->spec;
+ const struct rte_flow_item_tcp *mask = item->mask;
+ struct mlx5_flow *flow = (struct mlx5_flow *)data;
+ struct ibv_exp_flow_spec_tcp_udp *tcp;
+ unsigned int tcp_size = sizeof(struct ibv_exp_flow_spec_tcp_udp);
+
+ tcp = (void *)((uintptr_t)flow->ibv_attr + flow->offset);
+ *tcp = (struct ibv_exp_flow_spec_tcp_udp) {
+ .type = IBV_EXP_FLOW_SPEC_TCP,
+ .size = tcp_size,
+ };
+ if (spec) {
+ tcp->val.dst_port = spec->hdr.dst_port;
+ tcp->val.src_port = spec->hdr.src_port;
+ }
+ if (mask) {
+ 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;
+ ++flow->ibv_attr->num_of_specs;
+ flow->ibv_attr->priority = 0;
+ return 0;
+}
+
+/**
+ * Complete flow rule creation.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param ibv_attr
+ * Verbs flow attributes.
+ * @param action
+ * Target action structure.
+ * @param[out] error
+ * Perform verbose error reporting if not NULL.
+ *
+ * @return
+ * A flow if the rule could be created.
+ */
+static struct rte_flow *
+priv_flow_create_action_queue(struct priv *priv,
+ struct ibv_exp_flow_attr *ibv_attr,
+ struct mlx5_flow_action *action,
+ struct rte_flow_error *error)
+{
+ struct rxq_ctrl *rxq;
+ struct rte_flow *rte_flow;
+
+ assert(priv->pd);
+ assert(priv->ctx);
+ rte_flow = rte_calloc(__func__, 1, sizeof(*rte_flow), 0);
+ if (!rte_flow) {
+ rte_flow_error_set(error, ENOMEM, RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "cannot allocate flow memory");
+ return NULL;
+ }
+ if (action->drop) {
+ rte_flow->cq =
+ ibv_exp_create_cq(priv->ctx, 1, NULL, NULL, 0,
+ &(struct ibv_exp_cq_init_attr){
+ .comp_mask = 0,
+ });
+ if (!rte_flow->cq) {
+ rte_flow_error_set(error, ENOMEM,
+ RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "cannot allocate CQ");
+ goto error;
+ }
+ rte_flow->wq = ibv_exp_create_wq(priv->ctx,
+ &(struct ibv_exp_wq_init_attr){
+ .wq_type = IBV_EXP_WQT_RQ,
+ .max_recv_wr = 1,
+ .max_recv_sge = 1,
+ .pd = priv->pd,
+ .cq = rte_flow->cq,
+ });
+ } else {
+ rxq = container_of((*priv->rxqs)[action->queue_id],
+ struct rxq_ctrl, rxq);
+ rte_flow->rxq = &rxq->rxq;
+ rte_flow->wq = rxq->wq;
+ }
+ rte_flow->ibv_attr = ibv_attr;
+ rte_flow->ind_table = ibv_exp_create_rwq_ind_table(
+ priv->ctx,
+ &(struct ibv_exp_rwq_ind_table_init_attr){
+ .pd = priv->pd,
+ .log_ind_tbl_size = 0,
+ .ind_tbl = &rte_flow->wq,
+ .comp_mask = 0,
+ });
+ if (!rte_flow->ind_table) {
+ rte_flow_error_set(error, ENOMEM, RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "cannot allocate indirection table");
+ goto error;
+ }
+ rte_flow->qp = ibv_exp_create_qp(
+ priv->ctx,
+ &(struct ibv_exp_qp_init_attr){
+ .qp_type = IBV_QPT_RAW_PACKET,
+ .comp_mask =
+ IBV_EXP_QP_INIT_ATTR_PD |
+ IBV_EXP_QP_INIT_ATTR_PORT |
+ IBV_EXP_QP_INIT_ATTR_RX_HASH,
+ .pd = priv->pd,
+ .rx_hash_conf = &(struct ibv_exp_rx_hash_conf){
+ .rx_hash_function =
+ IBV_EXP_RX_HASH_FUNC_TOEPLITZ,
+ .rx_hash_key_len = rss_hash_default_key_len,
+ .rx_hash_key = rss_hash_default_key,
+ .rx_hash_fields_mask = 0,
+ .rwq_ind_tbl = rte_flow->ind_table,
+ },
+ .port_num = priv->port,
+ });
+ if (!rte_flow->qp) {
+ rte_flow_error_set(error, ENOMEM, RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "cannot allocate QP");
+ goto error;
+ }
+ rte_flow->ibv_flow = ibv_exp_create_flow(rte_flow->qp,
+ rte_flow->ibv_attr);
+ if (!rte_flow->ibv_flow) {
+ rte_flow_error_set(error, ENOMEM, RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "flow rule creation failure");
+ goto error;
+ }
+ return rte_flow;
+error:
+ assert(rte_flow);
+ if (rte_flow->qp)
+ ibv_destroy_qp(rte_flow->qp);
+ if (rte_flow->ind_table)
+ ibv_exp_destroy_rwq_ind_table(rte_flow->ind_table);
+ if (!rte_flow->rxq && rte_flow->wq)
+ ibv_exp_destroy_wq(rte_flow->wq);
+ if (!rte_flow->rxq && rte_flow->cq)
+ ibv_destroy_cq(rte_flow->cq);
+ rte_free(rte_flow->ibv_attr);
+ rte_free(rte_flow);
+ return NULL;
+}
+
+/**
+ * Convert a flow.
+ *
+ * @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.
+ *
+ * @return
+ * A flow on success, NULL otherwise.
+ */
+static struct rte_flow *
+priv_flow_create(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 rte_flow *rte_flow;
+ struct mlx5_flow_action action;
+ struct mlx5_flow flow = { .offset = sizeof(struct ibv_exp_flow_attr), };
+ int err;
+
+ err = priv_flow_validate(priv, attr, items, actions, error, &flow);
+ if (err)
+ goto exit;
+ flow.ibv_attr = rte_malloc(__func__, flow.offset, 0);
+ flow.offset = sizeof(struct ibv_exp_flow_attr);
+ if (!flow.ibv_attr) {
+ rte_flow_error_set(error, ENOMEM, RTE_FLOW_ERROR_TYPE_HANDLE,
+ NULL, "cannot allocate ibv_attr memory");
+ goto exit;
+ }
+ *flow.ibv_attr = (struct ibv_exp_flow_attr){
+ .type = IBV_EXP_FLOW_ATTR_NORMAL,
+ .size = sizeof(struct ibv_exp_flow_attr),
+ .priority = attr->priority,
+ .num_of_specs = 0,
+ .port = 0,
+ .flags = 0,
+ .reserved = 0,
+ };
+ claim_zero(priv_flow_validate(priv, attr, items, actions,
+ error, &flow));
+ action = (struct mlx5_flow_action){
+ .queue = 0,
+ .drop = 0,
+ };
+ 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_QUEUE) {
+ action.queue = 1;
+ action.queue_id =
+ ((const struct rte_flow_action_queue *)
+ actions->conf)->index;
+ } else if (actions->type == RTE_FLOW_ACTION_TYPE_DROP) {
+ action.drop = 1;
+ } else {
+ rte_flow_error_set(error, ENOTSUP,
+ RTE_FLOW_ERROR_TYPE_ACTION,
+ actions, "unsupported action");
+ goto exit;
+ }
+ }
+ rte_flow = priv_flow_create_action_queue(priv, flow.ibv_attr,
+ &action, error);
+ return rte_flow;
+exit:
+ rte_free(flow.ibv_attr);
+ return NULL;
}
/**
@@ -74,15 +844,46 @@ mlx5_flow_create(struct rte_eth_dev *dev,
const struct rte_flow_action actions[],
struct rte_flow_error *error)
{
- (void)dev;
- (void)attr;
- (void)items;
- (void)actions;
- (void)error;
- rte_flow_error_set(error, ENOTSUP,
- RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
- NULL, "not implemented yet");
- return NULL;
+ struct priv *priv = dev->data->dev_private;
+ struct rte_flow *flow;
+
+ priv_lock(priv);
+ flow = priv_flow_create(priv, attr, items, actions, error);
+ if (flow) {
+ LIST_INSERT_HEAD(&priv->flows, flow, next);
+ DEBUG("Flow created %p", (void *)flow);
+ }
+ priv_unlock(priv);
+ return flow;
+}
+
+/**
+ * Destroy a flow.
+ *
+ * @param priv
+ * Pointer to private structure.
+ * @param[in] flow
+ * Flow to destroy.
+ */
+static void
+priv_flow_destroy(struct priv *priv,
+ struct rte_flow *flow)
+{
+ (void)priv;
+ LIST_REMOVE(flow, next);
+ if (flow->ibv_flow)
+ claim_zero(ibv_exp_destroy_flow(flow->ibv_flow));
+ if (flow->qp)
+ claim_zero(ibv_destroy_qp(flow->qp));
+ if (flow->ind_table)
+ claim_zero(ibv_exp_destroy_rwq_ind_table(flow->ind_table));
+ if (!flow->rxq && flow->wq)
+ claim_zero(ibv_exp_destroy_wq(flow->wq));
+ if (!flow->rxq && flow->cq)
+ claim_zero(ibv_destroy_cq(flow->cq));
+ rte_free(flow->ibv_attr);
+ DEBUG("Flow destroyed %p", (void *)flow);
+ rte_free(flow);
}
/**
@@ -96,13 +897,30 @@ mlx5_flow_destroy(struct rte_eth_dev *dev,
struct rte_flow *flow,
struct rte_flow_error *error)
{
- (void)dev;
- (void)flow;
+ struct priv *priv = dev->data->dev_private;
+
(void)error;
- rte_flow_error_set(error, ENOTSUP,
- RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
- NULL, "not implemented yet");
- return -rte_errno;
+ priv_lock(priv);
+ priv_flow_destroy(priv, flow);
+ priv_unlock(priv);
+ return 0;
+}
+
+/**
+ * Destroy all flows.
+ *
+ * @param priv
+ * Pointer to private structure.
+ */
+static void
+priv_flow_flush(struct priv *priv)
+{
+ while (!LIST_EMPTY(&priv->flows)) {
+ struct rte_flow *flow;
+
+ flow = LIST_FIRST(&priv->flows);
+ priv_flow_destroy(priv, flow);
+ }
}
/**
@@ -115,10 +933,62 @@ int
mlx5_flow_flush(struct rte_eth_dev *dev,
struct rte_flow_error *error)
{
- (void)dev;
+ struct priv *priv = dev->data->dev_private;
+
(void)error;
- rte_flow_error_set(error, ENOTSUP,
- RTE_FLOW_ERROR_TYPE_UNSPECIFIED,
- NULL, "not implemented yet");
- return -rte_errno;
+ priv_lock(priv);
+ priv_flow_flush(priv);
+ priv_unlock(priv);
+ return 0;
+}
+
+/**
+ * Remove all flows.
+ *
+ * Called by dev_stop() to remove all flows.
+ *
+ * @param priv
+ * Pointer to private structure.
+ */
+void
+priv_flow_stop(struct priv *priv)
+{
+ struct rte_flow *flow;
+
+ for (flow = LIST_FIRST(&priv->flows);
+ flow;
+ flow = LIST_NEXT(flow, next)) {
+ claim_zero(ibv_exp_destroy_flow(flow->ibv_flow));
+ flow->ibv_flow = NULL;
+ DEBUG("Flow %p removed", (void *)flow);
+ }
+}
+
+/**
+ * Add all flows.
+ *
+ * @param priv
+ * Pointer to private structure.
+ *
+ * @return
+ * 0 on success, a errno value otherwise and rte_errno is set.
+ */
+int
+priv_flow_start(struct priv *priv)
+{
+ struct rte_flow *flow;
+
+ for (flow = LIST_FIRST(&priv->flows);
+ flow;
+ flow = LIST_NEXT(flow, next)) {
+ flow->ibv_flow = ibv_exp_create_flow(flow->qp,
+ flow->ibv_attr);
+ if (!flow->ibv_flow) {
+ DEBUG("Flow %p cannot be applied", (void *)flow);
+ rte_errno = EINVAL;
+ return rte_errno;
+ }
+ DEBUG("Flow %p applied", (void *)flow);
+ }
+ return 0;
}
@@ -90,6 +90,7 @@ mlx5_dev_start(struct rte_eth_dev *dev)
if (dev->data->dev_conf.fdir_conf.mode != RTE_FDIR_MODE_NONE)
priv_fdir_enable(priv);
priv_dev_interrupt_handler_install(priv, dev);
+ err = priv_flow_start(priv);
priv_unlock(priv);
return -err;
}
@@ -120,6 +121,7 @@ mlx5_dev_stop(struct rte_eth_dev *dev)
priv_mac_addrs_disable(priv);
priv_destroy_hash_rxqs(priv);
priv_fdir_disable(priv);
+ priv_flow_stop(priv);
priv_dev_interrupt_handler_uninstall(priv, dev);
priv->started = 0;
priv_unlock(priv);