net/mlx5: Parallelize vhca event handling

At present, mlx5 driver have a general purpose
event handler which not only handles vhca event
but also many other events. This incurs a huge
bottleneck because the event handler is
implemented by single threaded workqueue and all
events are forced to be handled in serial manner
even though application tries to create multiple
SFs simultaneously.

Introduce a dedicated vhca event handler which
manages SFs parallel creation.

Signed-off-by: Wei Zhang <weizhang@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Reviewed-by: Shay Drory <shayd@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>

1 files changed, 54 insertions, 3 deletions

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/sf/vhca_event.c b/drivers/net/ethernet/mellanox/mlx5/core/sf/vhca_event.c
index d908fba968f0..c6fd729de8b2 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/sf/vhca_event.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/sf/vhca_event.c
@@ -21,6 +21,15 @@ struct mlx5_vhca_event_work {
 	struct mlx5_vhca_state_event event;
 };
 
+struct mlx5_vhca_event_handler {
+	struct workqueue_struct *wq;
+};
+
+struct mlx5_vhca_events {
+	struct mlx5_core_dev *dev;
+	struct mlx5_vhca_event_handler handler[MLX5_DEV_MAX_WQS];
+};
+
 int mlx5_cmd_query_vhca_state(struct mlx5_core_dev *dev, u16 function_id, u32 *out, u32 outlen)
 {
 	u32 in[MLX5_ST_SZ_DW(query_vhca_state_in)] = {};
@@ -99,6 +108,12 @@ static void mlx5_vhca_state_work_handler(struct work_struct *_work)
 	kfree(work);
 }
 
+static void
+mlx5_vhca_events_work_enqueue(struct mlx5_core_dev *dev, int idx, struct work_struct *work)
+{
+	queue_work(dev->priv.vhca_events->handler[idx].wq, work);
+}
+
 static int
 mlx5_vhca_state_change_notifier(struct notifier_block *nb, unsigned long type, void *data)
 {
@@ -106,6 +121,7 @@ mlx5_vhca_state_change_notifier(struct notifier_block *nb, unsigned long type, v
 				mlx5_nb_cof(nb, struct mlx5_vhca_state_notifier, nb);
 	struct mlx5_vhca_event_work *work;
 	struct mlx5_eqe *eqe = data;
+	int wq_idx;
 
 	work = kzalloc(sizeof(*work), GFP_ATOMIC);
 	if (!work)
@@ -113,7 +129,8 @@ mlx5_vhca_state_change_notifier(struct notifier_block *nb, unsigned long type, v
 	INIT_WORK(&work->work, &mlx5_vhca_state_work_handler);
 	work->notifier = notifier;
 	work->event.function_id = be16_to_cpu(eqe->data.vhca_state.function_id);
-	mlx5_events_work_enqueue(notifier->dev, &work->work);
+	wq_idx = work->event.function_id % MLX5_DEV_MAX_WQS;
+	mlx5_vhca_events_work_enqueue(notifier->dev, wq_idx, &work->work);
 	return NOTIFY_OK;
 }
 
@@ -132,28 +149,62 @@ void mlx5_vhca_state_cap_handle(struct mlx5_core_dev *dev, void *set_hca_cap)
 int mlx5_vhca_event_init(struct mlx5_core_dev *dev)
 {
 	struct mlx5_vhca_state_notifier *notifier;
+	char wq_name[MLX5_CMD_WQ_MAX_NAME];
+	struct mlx5_vhca_events *events;
+	int err, i;
 
 	if (!mlx5_vhca_event_supported(dev))
 		return 0;
 
-	notifier = kzalloc(sizeof(*notifier), GFP_KERNEL);
-	if (!notifier)
+	events = kzalloc(sizeof(*events), GFP_KERNEL);
+	if (!events)
 		return -ENOMEM;
 
+	events->dev = dev;
+	dev->priv.vhca_events = events;
+	for (i = 0; i < MLX5_DEV_MAX_WQS; i++) {
+		snprintf(wq_name, MLX5_CMD_WQ_MAX_NAME, "mlx5_vhca_event%d", i);
+		events->handler[i].wq = create_singlethread_workqueue(wq_name);
+		if (!events->handler[i].wq) {
+			err = -ENOMEM;
+			goto err_create_wq;
+		}
+	}
+
+	notifier = kzalloc(sizeof(*notifier), GFP_KERNEL);
+	if (!notifier) {
+		err = -ENOMEM;
+		goto err_notifier;
+	}
+
 	dev->priv.vhca_state_notifier = notifier;
 	notifier->dev = dev;
 	BLOCKING_INIT_NOTIFIER_HEAD(&notifier->n_head);
 	MLX5_NB_INIT(&notifier->nb, mlx5_vhca_state_change_notifier, VHCA_STATE_CHANGE);
 	return 0;
+
+err_notifier:
+err_create_wq:
+	for (--i; i >= 0; i--)
+		destroy_workqueue(events->handler[i].wq);
+	kfree(events);
+	return err;
 }
 
 void mlx5_vhca_event_cleanup(struct mlx5_core_dev *dev)
 {
+	struct mlx5_vhca_events *vhca_events;
+	int i;
+
 	if (!mlx5_vhca_event_supported(dev))
 		return;
 
 	kfree(dev->priv.vhca_state_notifier);
 	dev->priv.vhca_state_notifier = NULL;
+	vhca_events = dev->priv.vhca_events;
+	for (i = 0; i < MLX5_DEV_MAX_WQS; i++)
+		destroy_workqueue(vhca_events->handler[i].wq);
+	kvfree(vhca_events);
 }
 
 void mlx5_vhca_event_start(struct mlx5_core_dev *dev)