net/mlx5: refactor TC-flow infrastructure

This commit refactors tc_flow as a preparation to coming commits
that sends different type of messages and expect differ type of replies
while still using the same underlying routines.

Signed-off-by: Moti Haimovsky <motih@mellanox.com>

drivers/net/mlx5/mlx5.c

@@ -286,8 +286,8 @@ mlx5_dev_close(struct rte_eth_dev *dev)
 		close(priv->nl_socket_route);
 	if (priv->nl_socket_rdma >= 0)
 		close(priv->nl_socket_rdma);
-	if (priv->mnl_socket)
+	if (priv->tcf_context)
-		mlx5_flow_tcf_socket_destroy(priv->mnl_socket);
+		mlx5_flow_tcf_context_destroy(priv->tcf_context);
 	ret = mlx5_hrxq_ibv_verify(dev);
 	if (ret)
 		DRV_LOG(WARNING, "port %u some hash Rx queue still remain",
@@ -1138,8 +1138,8 @@ mlx5_dev_spawn(struct rte_device *dpdk_dev,
 	claim_zero(mlx5_mac_addr_add(eth_dev, &mac, 0, 0));
 	if (vf && config.vf_nl_en)
 		mlx5_nl_mac_addr_sync(eth_dev);
-	priv->mnl_socket = mlx5_flow_tcf_socket_create();
+	priv->tcf_context = mlx5_flow_tcf_context_create();
-	if (!priv->mnl_socket) {
+	if (!priv->tcf_context) {
 		err = -rte_errno;
 		DRV_LOG(WARNING,
 			"flow rules relying on switch offloads will not be"
@@ -1154,16 +1154,16 @@ mlx5_dev_spawn(struct rte_device *dpdk_dev,
 			error.message =
 				"cannot retrieve network interface index";
 		} else {
-			err = mlx5_flow_tcf_init(priv->mnl_socket, ifindex,
+			err = mlx5_flow_tcf_init(priv->tcf_context,
-						&error);
+						 ifindex, &error);
 		}
 		if (err) {
 			DRV_LOG(WARNING,
 				"flow rules relying on switch offloads will"
 				" not be supported: %s: %s",
 				error.message, strerror(rte_errno));
-			mlx5_flow_tcf_socket_destroy(priv->mnl_socket);
+			mlx5_flow_tcf_context_destroy(priv->tcf_context);
-			priv->mnl_socket = NULL;
+			priv->tcf_context = NULL;
 		}
 	}
 	TAILQ_INIT(&priv->flows);
@@ -1218,8 +1218,8 @@ mlx5_dev_spawn(struct rte_device *dpdk_dev,
 			close(priv->nl_socket_route);
 		if (priv->nl_socket_rdma >= 0)
 			close(priv->nl_socket_rdma);
-		if (priv->mnl_socket)
+		if (priv->tcf_context)
-			mlx5_flow_tcf_socket_destroy(priv->mnl_socket);
+			mlx5_flow_tcf_context_destroy(priv->tcf_context);
 		if (own_domain_id)
 			claim_zero(rte_eth_switch_domain_free(priv->domain_id));
 		rte_free(priv);

drivers/net/mlx5/mlx5.h

@@ -169,7 +169,7 @@ struct mlx5_drop {
 	struct mlx5_rxq_ibv *rxq; /* Verbs Rx queue. */
 };
 
-struct mnl_socket;
+struct mlx5_flow_tcf_context;
 
 struct priv {
 	LIST_ENTRY(priv) mem_event_cb; /* Called by memory event callback. */
@@ -236,7 +236,7 @@ struct priv {
 	rte_spinlock_t uar_lock[MLX5_UAR_PAGE_NUM_MAX];
 	/* UAR same-page access control required in 32bit implementations. */
 #endif
-	struct mnl_socket *mnl_socket; /* Libmnl socket. */
+	struct mlx5_flow_tcf_context *tcf_context; /* TC flower context. */
 };
 
 #define PORT_ID(priv) ((priv)->dev_data->port_id)

drivers/net/mlx5/mlx5_flow.h

@@ -243,7 +243,6 @@ struct rte_flow {
 	struct rte_flow_action_rss rss;/**< RSS context. */
 	uint8_t key[MLX5_RSS_HASH_KEY_LEN]; /**< RSS hash key. */
 	uint16_t (*queue)[]; /**< Destination queues to redirect traffic to. */
-	void *nl_flow; /**< Netlink flow buffer if relevant. */
 	LIST_HEAD(dev_flows, mlx5_flow) dev_flows;
 	/**< Device flows that are part of the flow. */
 	uint32_t actions; /**< Bit-fields which mark all detected actions. */
@@ -350,9 +349,9 @@ int mlx5_flow_validate_item_vxlan_gpe(const struct rte_flow_item *item,
 
 /* mlx5_flow_tcf.c */
 
-int mlx5_flow_tcf_init(struct mnl_socket *nl, unsigned int ifindex,
+int mlx5_flow_tcf_init(struct mlx5_flow_tcf_context *ctx,
-		       struct rte_flow_error *error);
+		       unsigned int ifindex, struct rte_flow_error *error);
-struct mnl_socket *mlx5_flow_tcf_socket_create(void);
+struct mlx5_flow_tcf_context *mlx5_flow_tcf_context_create(void);
-void mlx5_flow_tcf_socket_destroy(struct mnl_socket *nl);
+void mlx5_flow_tcf_context_destroy(struct mlx5_flow_tcf_context *ctx);
 
 #endif /* RTE_PMD_MLX5_FLOW_H_ */

drivers/net/mlx5/mlx5_flow_tcf.c

@@ -235,6 +235,19 @@ struct tc_pedit_sel {
 #define TTL_LEN 1
 #endif
 
+/**
+ * Structure for holding netlink context.
+ * Note the size of the message buffer which is MNL_SOCKET_BUFFER_SIZE.
+ * Using this (8KB) buffer size ensures that netlink messages will never be
+ * truncated.
+ */
+struct mlx5_flow_tcf_context {
+	struct mnl_socket *nl; /* NETLINK_ROUTE libmnl socket. */
+	uint32_t seq; /* Message sequence number. */
+	uint32_t buf_size; /* Message buffer size. */
+	uint8_t *buf; /* Message buffer. */
+};
+
 /** Empty masks for known item types. */
 static const union {
 	struct rte_flow_item_port_id port_id;
@@ -2153,7 +2166,7 @@ flow_tcf_apply(struct rte_eth_dev *dev, struct rte_flow *flow,
 	       struct rte_flow_error *error)
 {
 	struct priv *priv = dev->data->dev_private;
-	struct mnl_socket *nl = priv->mnl_socket;
+	struct mnl_socket *nl = priv->tcf_context->nl;
 	struct mlx5_flow *dev_flow;
 	struct nlmsghdr *nlh;
 
@@ -2182,7 +2195,7 @@ static void
 flow_tcf_remove(struct rte_eth_dev *dev, struct rte_flow *flow)
 {
 	struct priv *priv = dev->data->dev_private;
-	struct mnl_socket *nl = priv->mnl_socket;
+	struct mnl_socket *nl = priv->tcf_context->nl;
 	struct mlx5_flow *dev_flow;
 	struct nlmsghdr *nlh;
 
@@ -2234,10 +2247,47 @@ const struct mlx5_flow_driver_ops mlx5_flow_tcf_drv_ops = {
 };
 
 /**
- * Initialize ingress qdisc of a given network interface.
+ * Create and configure a libmnl socket for Netlink flow rules.
+ *
+ * @return
+ *   A valid libmnl socket object pointer on success, NULL otherwise and
+ *   rte_errno is set.
+ */
+static struct mnl_socket *
+flow_tcf_mnl_socket_create(void)
+{
+	struct mnl_socket *nl = mnl_socket_open(NETLINK_ROUTE);
+
+	if (nl) {
+		mnl_socket_setsockopt(nl, NETLINK_CAP_ACK, &(int){ 1 },
+				      sizeof(int));
+		if (!mnl_socket_bind(nl, 0, MNL_SOCKET_AUTOPID))
+			return nl;
+	}
+	rte_errno = errno;
+	if (nl)
+		mnl_socket_close(nl);
+	return NULL;
+}
+
+/**
+ * Destroy a libmnl socket.
  *
  * @param nl
  *   Libmnl socket of the @p NETLINK_ROUTE kind.
+ */
+static void
+flow_tcf_mnl_socket_destroy(struct mnl_socket *nl)
+{
+	if (nl)
+		mnl_socket_close(nl);
+}
+
+/**
+ * Initialize ingress qdisc of a given network interface.
+ *
+ * @param ctx
+ *   Pointer to tc-flower context to use.
  * @param ifindex
  *   Index of network interface to initialize.
  * @param[out] error
@@ -2247,11 +2297,12 @@ const struct mlx5_flow_driver_ops mlx5_flow_tcf_drv_ops = {
  *   0 on success, a negative errno value otherwise and rte_errno is set.
  */
 int
-mlx5_flow_tcf_init(struct mnl_socket *nl, unsigned int ifindex,
+mlx5_flow_tcf_init(struct mlx5_flow_tcf_context *ctx,
-		   struct rte_flow_error *error)
+		   unsigned int ifindex, struct rte_flow_error *error)
 {
 	struct nlmsghdr *nlh;
 	struct tcmsg *tcm;
+	struct mnl_socket *nl = ctx->nl;
 	alignas(struct nlmsghdr)
 	uint8_t buf[mnl_nlmsg_size(sizeof(*tcm) + 128)];
 
@@ -2290,37 +2341,47 @@ mlx5_flow_tcf_init(struct mnl_socket *nl, unsigned int ifindex,
 }
 
 /**
- * Create and configure a libmnl socket for Netlink flow rules.
+ * Create libmnl context for Netlink flow rules.
  *
  * @return
  *   A valid libmnl socket object pointer on success, NULL otherwise and
  *   rte_errno is set.
  */
-struct mnl_socket *
+struct mlx5_flow_tcf_context *
-mlx5_flow_tcf_socket_create(void)
+mlx5_flow_tcf_context_create(void)
 {
-	struct mnl_socket *nl = mnl_socket_open(NETLINK_ROUTE);
+	struct mlx5_flow_tcf_context *ctx = rte_zmalloc(__func__,
-
+							sizeof(*ctx),
-	if (nl) {
+							sizeof(uint32_t));
-		mnl_socket_setsockopt(nl, NETLINK_CAP_ACK, &(int){ 1 },
+	if (!ctx)
-				      sizeof(int));
+		goto error;
-		if (!mnl_socket_bind(nl, 0, MNL_SOCKET_AUTOPID))
+	ctx->nl = flow_tcf_mnl_socket_create();
-			return nl;
+	if (!ctx->nl)
-	}
+		goto error;
-	rte_errno = errno;
+	ctx->buf_size = MNL_SOCKET_BUFFER_SIZE;
-	if (nl)
+	ctx->buf = rte_zmalloc(__func__,
-		mnl_socket_close(nl);
+			       ctx->buf_size, sizeof(uint32_t));
+	if (!ctx->buf)
+		goto error;
+	ctx->seq = random();
+	return ctx;
+error:
+	mlx5_flow_tcf_context_destroy(ctx);
 	return NULL;
 }
 
 /**
- * Destroy a libmnl socket.
+ * Destroy a libmnl context.
  *
  * @param nl
  *   Libmnl socket of the @p NETLINK_ROUTE kind.
  */
 void
-mlx5_flow_tcf_socket_destroy(struct mnl_socket *nl)
+mlx5_flow_tcf_context_destroy(struct mlx5_flow_tcf_context *ctx)
 {
-	mnl_socket_close(nl);
+	if (!ctx)
+		return;
+	flow_tcf_mnl_socket_destroy(ctx->nl);
+	rte_free(ctx->buf);
+	rte_free(ctx);
 }