On 11/23/2017 10:33 PM, Jingjing Wu wrote:
> Signed-off-by: Jingjing Wu <jingjing...@intel.com>
<...>
> @@ -233,6 +233,7 @@ CONFIG_RTE_LIBRTE_AVF_DEBUG_TX=n
> CONFIG_RTE_LIBRTE_AVF_DEBUG_TX_FREE=n
> CONFIG_RTE_LIBRTE_AVF_DEBUG_RX=n
> CONFIG_RTE_LIBRTE_AVF_16BYTE_RX_DESC=n
> +CONFIG_RTE_LIBRTE_AVF_INC_VECTOR=y
Can you please move this just below CONFIG_RTE_LIBRTE_AVF_PMD, since this
enable/disable vector PMD more important than debug configs.
<...>
> +#ifdef RTE_LIBRTE_AVF_INC_VECTOR
> +static inline bool
> +check_rx_vec_allow(struct avf_rx_queue *rxq)
> +{
> + if (rxq->rx_free_thresh >= AVF_VPMD_RX_MAX_BURST &&
> + rxq->nb_rx_desc % rxq->rx_free_thresh == 0) {
> + PMD_INIT_LOG(DEBUG, "Vector Rx"
> + " can be enabled on this rxq.");
> + return TRUE;
> + }
> +
> + PMD_INIT_LOG(DEBUG, "Vector Rx"
> + " cannot be enabled on this rxq.");
Can merge these two lines.
<...>
>
> -/* choose rx function*/
> +/* choose tx function*/
Can you please fix this when added in patch 4/14?
<...>
> + rte_mempool_put_bulk(free[0]->pool,
> + (void *)free,
Is void * cast required?
<...>
