This patch is not correct,
You are using old code. This was already solved.
Thanks,
Ori
> -----Original Message-----
> From: xiangxia.m....@gmail.com <xiangxia.m....@gmail.com>
> Sent: Wednesday, November 27, 2019 4:19 PM
> To: dev@dpdk.org
> Cc: Tonghao Zhang <xiangxia.m....@gmail.com>; Ori Kam
> <or...@mellanox.com>
> Subject: [PATCH dpdk-dev 3/3] net/mlx5: remove the assert for rxq cq
>
> From: Tonghao Zhang <xiangxia.m....@gmail.com>
>
> The mlx5_rxq_obj_hairpin_new doesn't alloc the
> cq anymore. When releasing it, don't check it.
>
> Fixes: e79c9be91515 ("net/mlx5: support Rx hairpin queues")
> Cc: Ori Kam <or...@mellanox.com>
> Signed-off-by: Tonghao Zhang <xiangxia.m....@gmail.com>
> ---
> drivers/net/mlx5/mlx5_rxq.c | 1 -
> 1 file changed, 1 deletion(-)
>
> diff --git a/drivers/net/mlx5/mlx5_rxq.c b/drivers/net/mlx5/mlx5_rxq.c
> index 2a12d5fa3436..582a2330e575 100644
> --- a/drivers/net/mlx5/mlx5_rxq.c
> +++ b/drivers/net/mlx5/mlx5_rxq.c
> @@ -673,7 +673,6 @@ mlx5_rxq_obj_release(struct mlx5_rxq_obj
> *rxq_obj)
> assert(rxq_obj);
> if (rxq_obj->type == MLX5_RXQ_OBJ_TYPE_IBV)
> assert(rxq_obj->wq);
> - assert(rxq_obj->cq);
> if (rte_atomic32_dec_and_test(&rxq_obj->refcnt)) {
> switch (rxq_obj->type) {
> case MLX5_RXQ_OBJ_TYPE_IBV:
> --
> 2.23.0
