As rte_write64() has an IO barrier, there's no need to have a barrier
before the call.
Signed-off-by: Yongseok Koh <ys...@mellanox.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 3c716b960..ca6cd3ae6 100644
--- a/drivers/net/mlx5/mlx5_rxq.c
+++ b/drivers/net/mlx5/mlx5_rxq.c
@@ -517,7 +517,6 @@ mlx5_arm_cq(struct mlx5_rxq_data *rxq, int sq_n_rxq)
doorbell = (uint64_t)doorbell_hi << 32;
doorbell |= rxq->cqn;
rxq->cq_db[MLX5_CQ_ARM_DB] = rte_cpu_to_be_32(doorbell_hi);
- rte_wmb();
rte_write64(rte_cpu_to_be_64(doorbell), cq_db_reg);
}
--
2.11.0
