Re: [PATCH rdma-next 2/2] RDMA/mlx5: Fix memory leak in case we fail to add an IB device

[Mark Bloch <markb@xxxxxxxxxxxx>]

On 2/11/2019 07:40, Leon Romanovsky wrote:
> From: Mark Bloch <markb@xxxxxxxxxxxx>
> 
> Make sure the IB device is freed on failure.
> 
> Fixes: b5ca15ad7e61 ("IB/mlx5: Add proper representors support")
> Signed-off-by: Mark Bloch <markb@xxxxxxxxxxxx>
> Reviewed-by: Bodong Wang <bodong@xxxxxxxxxxxx>
> Signed-off-by: Leon Romanovsky <leonro@xxxxxxxxxxxx>
> ---
>  drivers/infiniband/hw/mlx5/ib_rep.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/infiniband/hw/mlx5/ib_rep.c b/drivers/infiniband/hw/mlx5/ib_rep.c
> index 6d7b8bad4b61..ff663e0ec3c2 100644
> --- a/drivers/infiniband/hw/mlx5/ib_rep.c
> +++ b/drivers/infiniband/hw/mlx5/ib_rep.c
> @@ -78,8 +78,10 @@ mlx5_ib_vport_rep_load(struct mlx5_core_dev *dev, struct mlx5_eswitch_rep *rep)
>  	ibdev->mdev = dev;
>  	ibdev->num_ports = max(MLX5_CAP_GEN(dev, num_ports),
>  			       MLX5_CAP_GEN(dev, num_vhca_ports));
> -	if (!__mlx5_ib_add(ibdev, &rep_profile))
> +	if (!__mlx5_ib_add(ibdev, &rep_profile)) {
> +		ib_dealloc_device(&dev->ib_dev);
>  		return -EINVAL;
> +	}
>  

This is wrong :/

it should be:

diff --git a/drivers/infiniband/hw/mlx5/ib_rep.c b/drivers/infiniband/hw/mlx5/ib_rep.c
index 46a9ddc8ca56..07106bf394fa 100644
--- a/drivers/infiniband/hw/mlx5/ib_rep.c
+++ b/drivers/infiniband/hw/mlx5/ib_rep.c
@@ -78,8 +78,10 @@ mlx5_ib_vport_rep_load(struct mlx5_core_dev *dev, struct mlx5_eswitch_rep *rep)
        ibdev->mdev = dev;
        ibdev->num_ports = max(MLX5_CAP_GEN(dev, num_ports),
                               MLX5_CAP_GEN(dev, num_vhca_ports));
-       if (!__mlx5_ib_add(ibdev, &rep_profile))
+       if (!__mlx5_ib_add(ibdev, &rep_profile)) {
+               ib_dealloc_device(&ibdev->ib_dev);
                return -EINVAL;
+       }

        rep->rep_if[REP_IB].priv = ibdev;

>  	rep->rep_if[REP_IB].priv = ibdev;
>  
>