On Mon, Jul 8, 2024 at 9:31 AM <vanshika.shu...@nxp.com> wrote:
> @@ -118,6 +120,18 @@ dpaa_mbuf_create_pool(struct rte_mempool *mp)
>         rte_memcpy(bp_info, (void *)&rte_dpaa_bpid_info[bpid],
>                    sizeof(struct dpaa_bp_info));
>         mp->pool_data = (void *)bp_info;
> +       /* Update per core mempool cache threshold to optimal value which is
> +        * number of buffers that can be released to HW buffer pool in
> +        * a single API call.
> +        */
> +       for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
> +               cache = &mp->local_cache[lcore_id];
> +               DPAA_MEMPOOL_DEBUG("lCore %d: cache->flushthresh %d -> %d\n",

No \n please.


> +                       lcore_id, cache->flushthresh,
> +                       (uint32_t)(cache->size + DPAA_MBUF_MAX_ACQ_REL));
> +               if (cache->flushthresh)
> +                       cache->flushthresh = cache->size + 
> DPAA_MBUF_MAX_ACQ_REL;
> +       }
>
>         DPAA_MEMPOOL_INFO("BMAN pool created for bpid =%d", bpid);
>         return 0;


-- 
David Marchand

Reply via email to