Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <cdd8953e-2187-32f7-bb3c-aaf54581775d@redhat.com>
Date:   Wed, 31 May 2023 11:38:01 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.10.0
Subject: Re: [PATCH net v2] net/mlx5: Fix setting of irq->map.index for static
 IRQ case
Content-Language: en-US
To:     Niklas Schnelle <schnelle@linux.ibm.com>,
        Shay Drory <shayd@nvidia.com>,
        Saeed Mahameed <saeedm@nvidia.com>,
        Eli Cohen <elic@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>
Cc:     netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
        linux-s390@vger.kernel.org, Mark Brown <broonie@kernel.org>,
        Simon Horman <simon.horman@corigine.com>,
        linux-rdma@vger.kernel.org
References: <20230531084856.2091666-1-schnelle@linux.ibm.com>
From:   =?UTF-8?Q?C=c3=a9dric_Le_Goater?= <clegoate@redhat.com>
In-Reply-To: <20230531084856.2091666-1-schnelle@linux.ibm.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

On 5/31/23 10:48, Niklas Schnelle wrote:
> When dynamic IRQ allocation is not supported all IRQs are allocated up
> front in mlx5_irq_table_create() instead of dynamically as part of
> mlx5_irq_alloc(). In the latter dynamic case irq->map.index is set
> via the mapping returned by pci_msix_alloc_irq_at(). In the static case
> and prior to commit 1da438c0ae02 ("net/mlx5: Fix indexing of mlx5_irq")
> irq->map.index was set in mlx5_irq_alloc() twice once initially to 0 and
> then to the requested index before storing in the xarray. After this
> commit it is only set to 0 which breaks all other IRQ mappings.
> 
> Fix this by setting irq->map.index to the requested index together with
> irq->map.virq and improve the related comment to make it clearer which
> cases it deals with.
> 
> Tested-by: Mark Brown <broonie@kernel.org>
> Reviewed-by: Mark Brown <broonie@kernel.org>
> Reviewed-by: Simon Horman <simon.horman@corigine.com>
> Reviewed-by: Eli Cohen <elic@nvidia.com>
> Fixes: 1da438c0ae02 ("net/mlx5: Fix indexing of mlx5_irq")
> Signed-off-by: Niklas Schnelle <schnelle@linux.ibm.com>

I was seeing the issue on a zLPAR with a mlx5 VF device. The patch fixes it.

Tested-by: Cédric Le Goater <clg@redhat.com>

Thanks,

C.


> ---
> v1 -> v2:
> - Added R-bs/Acks
> - Fixed typos in commit message
> 
>   drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c | 9 +++++----
>   1 file changed, 5 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c b/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c
> index db5687d9fec9..fd5b43e8f3bb 100644
> --- a/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c
> +++ b/drivers/net/ethernet/mellanox/mlx5/core/pci_irq.c
> @@ -232,12 +232,13 @@ struct mlx5_irq *mlx5_irq_alloc(struct mlx5_irq_pool *pool, int i,
>   	if (!irq)
>   		return ERR_PTR(-ENOMEM);
>   	if (!i || !pci_msix_can_alloc_dyn(dev->pdev)) {
> -		/* The vector at index 0 was already allocated.
> -		 * Just get the irq number. If dynamic irq is not supported
> -		 * vectors have also been allocated.
> +		/* The vector at index 0 is always statically allocated. If
> +		 * dynamic irq is not supported all vectors are statically
> +		 * allocated. In both cases just get the irq number and set
> +		 * the index.
>   		 */
>   		irq->map.virq = pci_irq_vector(dev->pdev, i);
> -		irq->map.index = 0;
> +		irq->map.index = i;
>   	} else {
>   		irq->map = pci_msix_alloc_irq_at(dev->pdev, MSI_ANY_INDEX, af_desc);
>   		if (!irq->map.virq) {