net/mlx5e: Dynamically allocate DIM structure for SQs/RQs

Make it possible for the DIM structure to be torn down while an SQ or RQ is
still active. Changing the CQ period mode is an example where the previous
sampling done with the DIM structure would need to be invalidated.

Co-developed-by: Nabil S. Alramli <dev@nalramli.com>
Signed-off-by: Nabil S. Alramli <dev@nalramli.com>
Co-developed-by: Joe Damato <jdamato@fastly.com>
Signed-off-by: Joe Damato <jdamato@fastly.com>
Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://lore.kernel.org/r/20240419080445.417574-4-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Author: Binary-Eater

drivers/net/ethernet/mellanox/mlx5/core/en.h

@@ -430,7 +430,7 @@ struct mlx5e_txqsq {
 	u16                        cc;
 	u16                        skb_fifo_cc;
 	u32                        dma_fifo_cc;
-	struct dim                 dim; /* Adaptive Moderation */
+	struct dim                *dim; /* Adaptive Moderation */
 
 	/* dirtied @xmit */
 	u16                        pc ____cacheline_aligned_in_smp;
@@ -722,7 +722,7 @@ struct mlx5e_rq {
 	int                    ix;
 	unsigned int           hw_mtu;
 
-	struct dim         dim; /* Dynamic Interrupt Moderation */
+	struct dim            *dim; /* Dynamic Interrupt Moderation */
 
 	/* XDP */
 	struct bpf_prog __rcu *xdp_prog;

drivers/net/ethernet/mellanox/mlx5/core/en_dim.c

@@ -44,7 +44,7 @@ mlx5e_complete_dim_work(struct dim *dim, struct dim_cq_moder moder,
 void mlx5e_rx_dim_work(struct work_struct *work)
 {
 	struct dim *dim = container_of(work, struct dim, work);
-	struct mlx5e_rq *rq = container_of(dim, struct mlx5e_rq, dim);
+	struct mlx5e_rq *rq = dim->priv;
 	struct dim_cq_moder cur_moder =
 		net_dim_get_rx_moderation(dim->mode, dim->profile_ix);
 
@@ -54,7 +54,7 @@ void mlx5e_rx_dim_work(struct work_struct *work)
 void mlx5e_tx_dim_work(struct work_struct *work)
 {
 	struct dim *dim = container_of(work, struct dim, work);
-	struct mlx5e_txqsq *sq = container_of(dim, struct mlx5e_txqsq, dim);
+	struct mlx5e_txqsq *sq = dim->priv;
 	struct dim_cq_moder cur_moder =
 		net_dim_get_tx_moderation(dim->mode, dim->profile_ix);
 

drivers/net/ethernet/mellanox/mlx5/core/en_main.c

@@ -962,11 +962,20 @@ static int mlx5e_alloc_rq(struct mlx5e_params *params,
 		}
 	}
 
-	INIT_WORK(&rq->dim.work, mlx5e_rx_dim_work);
-	rq->dim.mode = params->rx_cq_moderation.cq_period_mode;
+	rq->dim = kvzalloc_node(sizeof(*rq->dim), GFP_KERNEL, node);
+	if (!rq->dim) {
+		err = -ENOMEM;
+		goto err_unreg_xdp_rxq_info;
+	}
+
+	rq->dim->priv = rq;
+	INIT_WORK(&rq->dim->work, mlx5e_rx_dim_work);
+	rq->dim->mode = params->rx_cq_moderation.cq_period_mode;
 
 	return 0;
 
+err_unreg_xdp_rxq_info:
+	xdp_rxq_info_unreg(&rq->xdp_rxq);
 err_destroy_page_pool:
 	page_pool_destroy(rq->page_pool);
 err_free_by_rq_type:
@@ -1014,6 +1023,7 @@ static void mlx5e_free_rq(struct mlx5e_rq *rq)
 		mlx5e_free_wqe_alloc_info(rq);
 	}
 
+	kvfree(rq->dim);
 	xdp_rxq_info_unreg(&rq->xdp_rxq);
 	page_pool_destroy(rq->page_pool);
 	mlx5_wq_destroy(&rq->wq_ctrl);
@@ -1341,7 +1351,7 @@ void mlx5e_deactivate_rq(struct mlx5e_rq *rq)
 
 void mlx5e_close_rq(struct mlx5e_rq *rq)
 {
-	cancel_work_sync(&rq->dim.work);
+	cancel_work_sync(&rq->dim->work);
 	cancel_work_sync(&rq->recover_work);
 	mlx5e_destroy_rq(rq);
 	mlx5e_free_rx_descs(rq);
@@ -1616,12 +1626,20 @@ static int mlx5e_alloc_txqsq(struct mlx5e_channel *c,
 	err = mlx5e_alloc_txqsq_db(sq, cpu_to_node(c->cpu));
 	if (err)
 		goto err_sq_wq_destroy;
+	sq->dim = kvzalloc_node(sizeof(*sq->dim), GFP_KERNEL, cpu_to_node(c->cpu));
+	if (!sq->dim) {
+		err = -ENOMEM;
+		goto err_free_txqsq_db;
+	}
 
-	INIT_WORK(&sq->dim.work, mlx5e_tx_dim_work);
-	sq->dim.mode = params->tx_cq_moderation.cq_period_mode;
+	sq->dim->priv = sq;
+	INIT_WORK(&sq->dim->work, mlx5e_tx_dim_work);
+	sq->dim->mode = params->tx_cq_moderation.cq_period_mode;
 
 	return 0;
 
+err_free_txqsq_db:
+	mlx5e_free_txqsq_db(sq);
 err_sq_wq_destroy:
 	mlx5_wq_destroy(&sq->wq_ctrl);
 
@@ -1630,6 +1648,7 @@ static int mlx5e_alloc_txqsq(struct mlx5e_channel *c,
 
 void mlx5e_free_txqsq(struct mlx5e_txqsq *sq)
 {
+	kvfree(sq->dim);
 	mlx5e_free_txqsq_db(sq);
 	mlx5_wq_destroy(&sq->wq_ctrl);
 }
@@ -1841,7 +1860,7 @@ void mlx5e_close_txqsq(struct mlx5e_txqsq *sq)
 	struct mlx5_core_dev *mdev = sq->mdev;
 	struct mlx5_rate_limit rl = {0};
 
-	cancel_work_sync(&sq->dim.work);
+	cancel_work_sync(&sq->dim->work);
 	cancel_work_sync(&sq->recover_work);
 	mlx5e_destroy_sq(mdev, sq->sqn);
 	if (sq->rate_limit) {

drivers/net/ethernet/mellanox/mlx5/core/en_txrx.c

@@ -55,7 +55,7 @@ static void mlx5e_handle_tx_dim(struct mlx5e_txqsq *sq)
 		return;
 
 	dim_update_sample(sq->cq.event_ctr, stats->packets, stats->bytes, &dim_sample);
-	net_dim(&sq->dim, dim_sample);
+	net_dim(sq->dim, dim_sample);
 }
 
 static void mlx5e_handle_rx_dim(struct mlx5e_rq *rq)
@@ -67,7 +67,7 @@ static void mlx5e_handle_rx_dim(struct mlx5e_rq *rq)
 		return;
 
 	dim_update_sample(rq->cq.event_ctr, stats->packets, stats->bytes, &dim_sample);
-	net_dim(&rq->dim, dim_sample);
+	net_dim(rq->dim, dim_sample);
 }
 
 void mlx5e_trigger_irq(struct mlx5e_icosq *sq)