Add per-device Q counters support

index 6cdd98ec9c1e..4cfb55ac0ed7 100644
@@ -5466,21 +5466,21 @@ static struct rdma_hw_stats *mlx5_ib_alloc_hw_stats(struct ib_device *ibdev,
 						    u8 port_num)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
-	struct mlx5_ib_port *port = &dev->port[port_num - 1];
+	const struct mlx5_ib_counters *cnts = &dev->port[port_num - 1].cnts;
 
 	/* We support only per port stats */
 	if (port_num == 0)
 		return NULL;
 
-	return rdma_alloc_hw_stats_struct(port->cnts.names,
-					  port->cnts.num_q_counters +
-					  port->cnts.num_cong_counters +
-					  port->cnts.num_ext_ppcnt_counters,
+	return rdma_alloc_hw_stats_struct(cnts->names,
+					  cnts->num_q_counters +
+					  cnts->num_cong_counters +
+					  cnts->num_ext_ppcnt_counters,
 					  RDMA_HW_STATS_DEFAULT_LIFESPAN);
 }
 
 static int mlx5_ib_query_q_counters(struct mlx5_core_dev *mdev,
-				    struct mlx5_ib_port *port,
+				    const struct mlx5_ib_counters *cnts,
 				    struct rdma_hw_stats *stats,
 				    u16 set_id)
 {

@@ -5497,8 +5497,8 @@ static int mlx5_ib_query_q_counters(struct mlx5_core_dev *mdev,
 	if (ret)
 		goto free;
 
-	for (i = 0; i < port->cnts.num_q_counters; i++) {
-		val = *(__be32 *)(out + port->cnts.offsets[i]);
+	for (i = 0; i < cnts->num_q_counters; i++) {
+		val = *(__be32 *)(out + cnts->offsets[i]);
 		stats->value[i] = (u64)be32_to_cpu(val);
 	}
 

@@ -5508,10 +5508,10 @@ static int mlx5_ib_query_q_counters(struct mlx5_core_dev *mdev,
 }
 
 static int mlx5_ib_query_ext_ppcnt_counters(struct mlx5_ib_dev *dev,
-					  struct mlx5_ib_port *port,
-					  struct rdma_hw_stats *stats)
+					    const struct mlx5_ib_counters *cnts,
+					    struct rdma_hw_stats *stats)
 {
-	int offset = port->cnts.num_q_counters + port->cnts.num_cong_counters;
+	int offset = cnts->num_q_counters + cnts->num_cong_counters;
 	int sz = MLX5_ST_SZ_BYTES(ppcnt_reg);
 	int ret, i;
 	void *out;

@@ -5524,12 +5524,10 @@ static int mlx5_ib_query_ext_ppcnt_counters(struct mlx5_ib_dev *dev,
 	if (ret)
 		goto free;
 
-	for (i = 0; i < port->cnts.num_ext_ppcnt_counters; i++) {
+	for (i = 0; i < cnts->num_ext_ppcnt_counters; i++)
 		stats->value[i + offset] =
 			be64_to_cpup((__be64 *)(out +
-				    port->cnts.offsets[i + offset]));
-	}
-
+				    cnts->offsets[i + offset]));
 free:
 	kvfree(out);
 	return ret;

@@ -5540,7 +5538,7 @@ static int mlx5_ib_get_hw_stats(struct ib_device *ibdev,
 				u8 port_num, int index)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
-	struct mlx5_ib_port *port = &dev->port[port_num - 1];
+	struct mlx5_ib_counters *cnts = &dev->port[port_num - 1].cnts;
 	struct mlx5_core_dev *mdev;
 	int ret, num_counters;
 	u8 mdev_port_num;

@@ -5548,18 +5546,17 @@ static int mlx5_ib_get_hw_stats(struct ib_device *ibdev,
 	if (!stats)
 		return -EINVAL;
 
-	num_counters = port->cnts.num_q_counters +
-		       port->cnts.num_cong_counters +
-		       port->cnts.num_ext_ppcnt_counters;
+	num_counters = cnts->num_q_counters +
+		       cnts->num_cong_counters +
+		       cnts->num_ext_ppcnt_counters;
 
 	/* q_counters are per IB device, query the master mdev */
-	ret = mlx5_ib_query_q_counters(dev->mdev, port, stats,
-				       port->cnts.set_id);
+	ret = mlx5_ib_query_q_counters(dev->mdev, cnts, stats, cnts->set_id);
 	if (ret)
 		return ret;
 
 	if (MLX5_CAP_PCAM_FEATURE(dev->mdev, rx_icrc_encapsulated_counter)) {
-		ret =  mlx5_ib_query_ext_ppcnt_counters(dev, port, stats);
+		ret =  mlx5_ib_query_ext_ppcnt_counters(dev, cnts, stats);
 		if (ret)
 			return ret;
 	}

@@ -5576,10 +5573,10 @@ static int mlx5_ib_get_hw_stats(struct ib_device *ibdev,
 		}
 		ret = mlx5_lag_query_cong_counters(dev->mdev,
 						   stats->value +
-						   port->cnts.num_q_counters,
-						   port->cnts.num_cong_counters,
-						   port->cnts.offsets +
-						   port->cnts.num_q_counters);
+						   cnts->num_q_counters,
+						   cnts->num_cong_counters,
+						   cnts->offsets +
+						   cnts->num_q_counters);
 
 		mlx5_ib_put_native_port_mdev(dev, port_num);
 		if (ret)

@@ -5594,20 +5591,21 @@ static struct rdma_hw_stats *
 mlx5_ib_counter_alloc_stats(struct rdma_counter *counter)
 {
 	struct mlx5_ib_dev *dev = to_mdev(counter->device);
-	struct mlx5_ib_port *port = &dev->port[counter->port - 1];
+	const struct mlx5_ib_counters *cnts =
+			&dev->port[counter->port - 1].cnts;
 
 	/* Q counters are in the beginning of all counters */
-	return rdma_alloc_hw_stats_struct(port->cnts.names,
-					  port->cnts.num_q_counters,
+	return rdma_alloc_hw_stats_struct(cnts->names,
+					  cnts->num_q_counters,
 					  RDMA_HW_STATS_DEFAULT_LIFESPAN);
 }
 
 static int mlx5_ib_counter_update_stats(struct rdma_counter *counter)
 {
 	struct mlx5_ib_dev *dev = to_mdev(counter->device);
-	struct mlx5_ib_port *port = &dev->port[counter->port - 1];
+	struct mlx5_ib_counters *cnts = &dev->port[counter->port - 1].cnts;
 
-	return mlx5_ib_query_q_counters(dev->mdev, port,
+	return mlx5_ib_query_q_counters(dev->mdev, cnts,
 					counter->stats, counter->id);
 }
 

index 4cfb55ac0ed7..4a3d700cd783 100644
@@ -5322,11 +5322,21 @@ static const struct mlx5_ib_counter ext_ppcnt_cnts[] = {
 	INIT_EXT_PPCNT_COUNTER(rx_icrc_encapsulated),
 };
 
+static bool is_mdev_switchdev_mode(const struct mlx5_core_dev *mdev)
+{
+	return MLX5_ESWITCH_MANAGER(mdev) &&
+	       mlx5_ib_eswitch_mode(mdev->priv.eswitch) ==
+		       MLX5_ESWITCH_OFFLOADS;
+}
+
 static void mlx5_ib_dealloc_counters(struct mlx5_ib_dev *dev)
 {
+	int num_cnt_ports;
 	int i;
 
-	for (i = 0; i < dev->num_ports; i++) {
+	num_cnt_ports = is_mdev_switchdev_mode(dev->mdev) ? 1 : dev->num_ports;
+
+	for (i = 0; i < num_cnt_ports; i++) {
 		if (dev->port[i].cnts.set_id_valid)
 			mlx5_core_dealloc_q_counter(dev->mdev,
 						    dev->port[i].cnts.set_id);

@@ -5428,13 +5438,15 @@ static void mlx5_ib_fill_counters(struct mlx5_ib_dev *dev,
 
 static int mlx5_ib_alloc_counters(struct mlx5_ib_dev *dev)
 {
+	int num_cnt_ports;
 	int err = 0;
 	int i;
 	bool is_shared;
 
 	is_shared = MLX5_CAP_GEN(dev->mdev, log_max_uctx) != 0;
+	num_cnt_ports = is_mdev_switchdev_mode(dev->mdev) ? 1 : dev->num_ports;
 
-	for (i = 0; i < dev->num_ports; i++) {
+	for (i = 0; i < num_cnt_ports; i++) {
 		err = __mlx5_ib_alloc_counters(dev, &dev->port[i].cnts);
 		if (err)
 			goto err_alloc;

@@ -5454,7 +5466,6 @@ static int mlx5_ib_alloc_counters(struct mlx5_ib_dev *dev)
 		}
 		dev->port[i].cnts.set_id_valid = true;
 	}
-
 	return 0;
 
 err_alloc:

@@ -5462,16 +5473,41 @@ static int mlx5_ib_alloc_counters(struct mlx5_ib_dev *dev)
 	return err;
 }
 
+static const struct mlx5_ib_counters *get_counters(struct mlx5_ib_dev *dev,
+						   u8 port_num)
+{
+	return is_mdev_switchdev_mode(dev->mdev) ? &dev->port[0].cnts :
+						   &dev->port[port_num].cnts;
+}
+
+/**
+ * mlx5_ib_get_counters_id - Returns counters id to use for device+port
+ * @dev:	Pointer to mlx5 IB device
+ * @port_num:	Zero based port number
+ *
+ * mlx5_ib_get_counters_id() Returns counters set id to use for given
+ * device port combination in switchdev and non switchdev mode of the
+ * parent device.
+ */
+u16 mlx5_ib_get_counters_id(struct mlx5_ib_dev *dev, u8 port_num)
+{
+	const struct mlx5_ib_counters *cnts = get_counters(dev, port_num);
+
+	return cnts->set_id;
+}
+
 static struct rdma_hw_stats *mlx5_ib_alloc_hw_stats(struct ib_device *ibdev,
 						    u8 port_num)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
-	const struct mlx5_ib_counters *cnts = &dev->port[port_num - 1].cnts;
+	const struct mlx5_ib_counters *cnts;
+	bool is_switchdev = is_mdev_switchdev_mode(dev->mdev);
 
-	/* We support only per port stats */
-	if (port_num == 0)
+	if ((is_switchdev && port_num) || (!is_switchdev && !port_num))
 		return NULL;
 
+	cnts = get_counters(dev, port_num - 1);
+
 	return rdma_alloc_hw_stats_struct(cnts->names,
 					  cnts->num_q_counters +
 					  cnts->num_cong_counters +

@@ -5538,7 +5574,7 @@ static int mlx5_ib_get_hw_stats(struct ib_device *ibdev,
 				u8 port_num, int index)
 {
 	struct mlx5_ib_dev *dev = to_mdev(ibdev);
-	struct mlx5_ib_counters *cnts = &dev->port[port_num - 1].cnts;
+	const struct mlx5_ib_counters *cnts = get_counters(dev, port_num - 1);
 	struct mlx5_core_dev *mdev;
 	int ret, num_counters;
 	u8 mdev_port_num;

@@ -5592,7 +5628,7 @@ mlx5_ib_counter_alloc_stats(struct rdma_counter *counter)
 {
 	struct mlx5_ib_dev *dev = to_mdev(counter->device);
 	const struct mlx5_ib_counters *cnts =
-			&dev->port[counter->port - 1].cnts;
+		get_counters(dev, counter->port - 1);
 
 	/* Q counters are in the beginning of all counters */
 	return rdma_alloc_hw_stats_struct(cnts->names,

@@ -5603,7 +5639,8 @@ mlx5_ib_counter_alloc_stats(struct rdma_counter *counter)
 static int mlx5_ib_counter_update_stats(struct rdma_counter *counter)
 {
 	struct mlx5_ib_dev *dev = to_mdev(counter->device);
-	struct mlx5_ib_counters *cnts = &dev->port[counter->port - 1].cnts;
+	const struct mlx5_ib_counters *cnts =
+		get_counters(dev, counter->port - 1);
 
 	return mlx5_ib_query_q_counters(dev->mdev, cnts,
 					counter->stats, counter->id);

index f6a53455bf8b..cb41a7e6255a 100644
@@ -1475,4 +1475,5 @@ int bfregn_to_uar_index(struct mlx5_ib_dev *dev,
 			bool dyn_bfreg);
 
 int mlx5_ib_qp_set_counter(struct ib_qp *qp, struct rdma_counter *counter);
+u16 mlx5_ib_get_counters_id(struct mlx5_ib_dev *dev, u8 port_num);
 #endif /* MLX5_IB_H */

index 379328b2598f..e96dcdfd6c3a 100644
@@ -3386,19 +3386,16 @@ static int __mlx5_ib_qp_set_counter(struct ib_qp *qp,
 	struct mlx5_ib_dev *dev = to_mdev(qp->device);
 	struct mlx5_ib_qp *mqp = to_mqp(qp);
 	struct mlx5_qp_context context = {};
-	struct mlx5_ib_port *mibport = NULL;
 	struct mlx5_ib_qp_base *base;
 	u32 set_id;
 
 	if (!MLX5_CAP_GEN(dev->mdev, rts2rts_qp_counters_set_id))
 		return 0;
 
-	if (counter) {
+	if (counter)
 		set_id = counter->id;
-	} else {
-		mibport = &dev->port[mqp->port - 1];
-		set_id = mibport->cnts.set_id;
-	}
+	else
+		set_id = mlx5_ib_get_counters_id(dev, mqp->port - 1);
 
 	base = &mqp->trans_qp.base;
 	context.qp_counter_set_usr_page &= cpu_to_be32(0xffffff);

@@ -3459,7 +3456,6 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 	struct mlx5_ib_cq *send_cq, *recv_cq;
 	struct mlx5_qp_context *context;
 	struct mlx5_ib_pd *pd;
-	struct mlx5_ib_port *mibport = NULL;
 	enum mlx5_qp_state mlx5_cur, mlx5_new;
 	enum mlx5_qp_optpar optpar;
 	u32 set_id = 0;

@@ -3624,11 +3620,10 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
 		if (qp->flags & MLX5_IB_QP_UNDERLAY)
 			port_num = 0;
 
-		mibport = &dev->port[port_num];
 		if (ibqp->counter)
 			set_id = ibqp->counter->id;
 		else
-			set_id = mibport->cnts.set_id;
+			set_id = mlx5_ib_get_counters_id(dev, port_num);
 		context->qp_counter_set_usr_page |=
 			cpu_to_be32(set_id << 24);
 	}

@@ -3817,6 +3812,8 @@ static int mlx5_ib_modify_dct(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 
 	dctc = MLX5_ADDR_OF(create_dct_in, qp->dct.in, dct_context_entry);
 	if (cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT) {
+		u16 set_id;
+
 		required |= IB_QP_ACCESS_FLAGS | IB_QP_PKEY_INDEX | IB_QP_PORT;
 		if (!is_valid_mask(attr_mask, required, 0))
 			return -EINVAL;

@@ -3843,7 +3840,9 @@ static int mlx5_ib_modify_dct(struct ib_qp *ibqp, struct ib_qp_attr *attr,
 		}
 		MLX5_SET(dctc, dctc, pkey_index, attr->pkey_index);
 		MLX5_SET(dctc, dctc, port, attr->port_num);
-		MLX5_SET(dctc, dctc, counter_set_id, dev->port[attr->port_num - 1].cnts.set_id);
+
+		set_id = mlx5_ib_get_counters_id(dev, attr->port_num - 1);
+		MLX5_SET(dctc, dctc, counter_set_id, set_id);
 
 	} else if (cur_state == IB_QPS_INIT && new_state == IB_QPS_RTR) {
 		struct mlx5_ib_modify_qp_resp resp = {};

@@ -6331,11 +6330,13 @@ int mlx5_ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr,
 	}
 
 	if (curr_wq_state == IB_WQS_RESET && wq_state == IB_WQS_RDY) {
+		u16 set_id;
+
+		set_id = mlx5_ib_get_counters_id(dev, 0);
 		if (MLX5_CAP_GEN(dev->mdev, modify_rq_counter_set_id)) {
 			MLX5_SET64(modify_rq_in, in, modify_bitmask,
 				   MLX5_MODIFY_RQ_IN_MODIFY_BITMASK_RQ_COUNTER_SET_ID);
-			MLX5_SET(rqc, rqc, counter_set_id,
-				 dev->port->cnts.set_id);
+			MLX5_SET(rqc, rqc, counter_set_id, set_id);
 		} else
 			dev_info_once(
 				&dev->ib_dev.dev,