Received-SPF: pass (google.com: domain of linux-kernel+bounces-54476-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.48.161 as permitted sender) client-ip=147.75.48.161;
Message-ID: <7e338c2a-6091-4093-8ca2-bb3b2af3e79d@gmail.com>
Date: Tue, 6 Feb 2024 10:11:28 +0200
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH net-next] eth: mlx5: link NAPI instances to queues and
 IRQs
Content-Language: en-US
To: Joe Damato <jdamato@fastly.com>, linux-kernel@vger.kernel.org,
 netdev@vger.kernel.org, Gal Pressman <gal@nvidia.com>
Cc: tariqt@nvidia.com, rrameshbabu@nvidia.com,
 Saeed Mahameed <saeedm@nvidia.com>, Leon Romanovsky <leon@kernel.org>,
 "David S. Miller" <davem@davemloft.net>, Eric Dumazet <edumazet@google.com>,
 Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
 "open list:MELLANOX MLX5 core VPI driver" <linux-rdma@vger.kernel.org>
References: <20240206010311.149103-1-jdamato@fastly.com>
From: Tariq Toukan <ttoukan.linux@gmail.com>
In-Reply-To: <20240206010311.149103-1-jdamato@fastly.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit


On 06/02/2024 3:03, Joe Damato wrote:
> Make mlx5 compatible with the newly added netlink queue GET APIs.
> 
> Signed-off-by: Joe Damato <jdamato@fastly.com>

+ Gal

Hi Joe,
Thanks for your patch.

We have already prepared a similar patch, and it's part of our internal 
submission queue, and planned to be submitted soon.

Please see my comments below, let us know if you're welling to respin a 
V2 or wait for our patch.

> ---
>   drivers/net/ethernet/mellanox/mlx5/core/en.h      | 1 +
>   drivers/net/ethernet/mellanox/mlx5/core/en_main.c | 8 ++++++++
>   2 files changed, 9 insertions(+)
> 
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h
> index 55c6ace0acd5..3f86ee1831a8 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h
> @@ -768,6 +768,7 @@ struct mlx5e_channel {
>   	u16                        qos_sqs_size;
>   	u8                         num_tc;
>   	u8                         lag_port;
> +	unsigned int		   irq;
>   
>   	/* XDP_REDIRECT */
>   	struct mlx5e_xdpsq         xdpsq;
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
> index c8e8f512803e..e1bfff1fb328 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c
> @@ -2473,6 +2473,9 @@ static void mlx5e_close_queues(struct mlx5e_channel *c)
>   	mlx5e_close_tx_cqs(c);
>   	mlx5e_close_cq(&c->icosq.cq);
>   	mlx5e_close_cq(&c->async_icosq.cq);
> +
> +	netif_queue_set_napi(c->netdev, c->ix, NETDEV_QUEUE_TYPE_TX, NULL);
> +	netif_queue_set_napi(c->netdev, c->ix, NETDEV_QUEUE_TYPE_RX, NULL);
>   }
>   
>   static u8 mlx5e_enumerate_lag_port(struct mlx5_core_dev *mdev, int ix)
> @@ -2558,6 +2561,7 @@ static int mlx5e_open_channel(struct mlx5e_priv *priv, int ix,
>   	c->stats    = &priv->channel_stats[ix]->ch;
>   	c->aff_mask = irq_get_effective_affinity_mask(irq);
>   	c->lag_port = mlx5e_enumerate_lag_port(priv->mdev, ix);
> +	c->irq		= irq;
>   
>   	netif_napi_add(netdev, &c->napi, mlx5e_napi_poll);
>   
> @@ -2602,6 +2606,10 @@ static void mlx5e_activate_channel(struct mlx5e_channel *c)
>   		mlx5e_activate_xsk(c);
>   	else
>   		mlx5e_activate_rq(&c->rq);
> +
> +	netif_napi_set_irq(&c->napi, c->irq);

Can be safely moved to mlx5e_open_channel without interfering with other 
existing mapping. This would save the new irq field in mlx5e_channel.

> +	netif_queue_set_napi(c->netdev, c->ix, NETDEV_QUEUE_TYPE_TX, &c->napi);

In some configurations we have multiple txqs per channel, so the txq 
indices are not 1-to-1 with their channel index.

This should be called per each txq, with the proper txq index.

It should be done also for feture-dedicated SQs (like QOS/HTB).

> +	netif_queue_set_napi(c->netdev, c->ix, NETDEV_QUEUE_TYPE_RX, &c->napi);

For consistency, I'd move this one as well, to match the TX handling.

>   }
>   
>   static void mlx5e_deactivate_channel(struct mlx5e_channel *c)