On Fri, Oct 9, 2020 at 9:36 AM Lance Richardson
<lance.richard...@broadcom.com> wrote:
>
> The fast mbuf free offload for non-vector mode requires
> additional checks in order to handle long tx buffer
> descriptors, so dedicated functions are needed for
> vector- and non-vector-modes.
>
> Fixes: 103169df2880 ("net/bnxt: support fast mbuf free")
> Signed-off-by: Lance Richardson <lance.richard...@broadcom.com>
> Reviewed-by: Ajit Kumar Khaparde <ajit.khapa...@broadcom.com>
Patch applied to dpdk-next-net-brcm/for-next-net with updated the commit id.
> ---
> drivers/net/bnxt/bnxt_rxtx_vec_common.h | 27 ++++++++++++++++++++
> drivers/net/bnxt/bnxt_rxtx_vec_neon.c | 2 +-
> drivers/net/bnxt/bnxt_rxtx_vec_sse.c | 2 +-
> drivers/net/bnxt/bnxt_txr.c | 34 +++++++++++++++++++++++++
> drivers/net/bnxt/bnxt_txr.h | 27 --------------------
> 5 files changed, 63 insertions(+), 29 deletions(-)
>
> diff --git a/drivers/net/bnxt/bnxt_rxtx_vec_common.h
> b/drivers/net/bnxt/bnxt_rxtx_vec_common.h
> index 33ac53568..d540e9eee 100644
> --- a/drivers/net/bnxt/bnxt_rxtx_vec_common.h
> +++ b/drivers/net/bnxt/bnxt_rxtx_vec_common.h
> @@ -95,6 +95,33 @@ bnxt_rxq_rearm(struct bnxt_rx_queue *rxq, struct
> bnxt_rx_ring_info *rxr)
> rxq->rxrearm_nb -= nb;
> }
>
> +/*
> + * Transmit completion function for use when DEV_TX_OFFLOAD_MBUF_FAST_FREE
> + * is enabled.
> + */
> +static inline void
> +bnxt_tx_cmp_vec_fast(struct bnxt_tx_queue *txq, int nr_pkts)
> +{
> + struct bnxt_tx_ring_info *txr = txq->tx_ring;
> + uint32_t ring_mask = txr->tx_ring_struct->ring_mask;
> + struct rte_mbuf **free = txq->free;
> + uint16_t cons = txr->tx_cons;
> + unsigned int blk = 0;
> +
> + while (nr_pkts--) {
> + struct bnxt_sw_tx_bd *tx_buf;
> +
> + tx_buf = &txr->tx_buf_ring[cons];
> + cons = (cons + 1) & ring_mask;
> + free[blk++] = tx_buf->mbuf;
> + tx_buf->mbuf = NULL;
> + }
> + if (blk)
> + rte_mempool_put_bulk(free[0]->pool, (void **)free, blk);
> +
> + txr->tx_cons = cons;
> +}
> +
> static inline void
> bnxt_tx_cmp_vec(struct bnxt_tx_queue *txq, int nr_pkts)
> {
> diff --git a/drivers/net/bnxt/bnxt_rxtx_vec_neon.c
> b/drivers/net/bnxt/bnxt_rxtx_vec_neon.c
> index 4c04cc43a..f49e29ccb 100644
> --- a/drivers/net/bnxt/bnxt_rxtx_vec_neon.c
> +++ b/drivers/net/bnxt/bnxt_rxtx_vec_neon.c
> @@ -369,7 +369,7 @@ bnxt_handle_tx_cp_vec(struct bnxt_tx_queue *txq)
> cpr->valid = !!(raw_cons & cp_ring_struct->ring_size);
> if (nb_tx_pkts) {
> if (txq->offloads & DEV_TX_OFFLOAD_MBUF_FAST_FREE)
> - bnxt_tx_cmp_fast(txq, nb_tx_pkts);
> + bnxt_tx_cmp_vec_fast(txq, nb_tx_pkts);
> else
> bnxt_tx_cmp_vec(txq, nb_tx_pkts);
> cpr->cp_raw_cons = raw_cons;
> diff --git a/drivers/net/bnxt/bnxt_rxtx_vec_sse.c
> b/drivers/net/bnxt/bnxt_rxtx_vec_sse.c
> index f71f46985..e4ba63551 100644
> --- a/drivers/net/bnxt/bnxt_rxtx_vec_sse.c
> +++ b/drivers/net/bnxt/bnxt_rxtx_vec_sse.c
> @@ -340,7 +340,7 @@ bnxt_handle_tx_cp_vec(struct bnxt_tx_queue *txq)
> cpr->valid = !!(raw_cons & cp_ring_struct->ring_size);
> if (nb_tx_pkts) {
> if (txq->offloads & DEV_TX_OFFLOAD_MBUF_FAST_FREE)
> - bnxt_tx_cmp_fast(txq, nb_tx_pkts);
> + bnxt_tx_cmp_vec_fast(txq, nb_tx_pkts);
> else
> bnxt_tx_cmp_vec(txq, nb_tx_pkts);
> cpr->cp_raw_cons = raw_cons;
> diff --git a/drivers/net/bnxt/bnxt_txr.c b/drivers/net/bnxt/bnxt_txr.c
> index 125ac8341..fb358d6f1 100644
> --- a/drivers/net/bnxt/bnxt_txr.c
> +++ b/drivers/net/bnxt/bnxt_txr.c
> @@ -337,6 +337,40 @@ static uint16_t bnxt_start_xmit(struct rte_mbuf *tx_pkt,
> return 0;
> }
>
> +/*
> + * Transmit completion function for use when DEV_TX_OFFLOAD_MBUF_FAST_FREE
> + * is enabled.
> + */
> +static void bnxt_tx_cmp_fast(struct bnxt_tx_queue *txq, int nr_pkts)
> +{
> + struct bnxt_tx_ring_info *txr = txq->tx_ring;
> + struct rte_mbuf **free = txq->free;
> + uint16_t cons = txr->tx_cons;
> + unsigned int blk = 0;
> + int i, j;
> +
> + for (i = 0; i < nr_pkts; i++) {
> + struct bnxt_sw_tx_bd *tx_buf;
> + unsigned short nr_bds;
> +
> + tx_buf = &txr->tx_buf_ring[cons];
> + nr_bds = tx_buf->nr_bds;
> + for (j = 0; j < nr_bds; j++) {
> + if (tx_buf->mbuf) {
> + /* Add mbuf to the bulk free array */
> + free[blk++] = tx_buf->mbuf;
> + tx_buf->mbuf = NULL;
> + }
> + cons = RING_NEXT(txr->tx_ring_struct, cons);
> + tx_buf = &txr->tx_buf_ring[cons];
> + }
> + }
> + if (blk)
> + rte_mempool_put_bulk(free[0]->pool, (void *)free, blk);
> +
> + txr->tx_cons = cons;
> +}
> +
> static void bnxt_tx_cmp(struct bnxt_tx_queue *txq, int nr_pkts)
> {
> struct bnxt_tx_ring_info *txr = txq->tx_ring;
> diff --git a/drivers/net/bnxt/bnxt_txr.h b/drivers/net/bnxt/bnxt_txr.h
> index c0b1fde22..d241227d4 100644
> --- a/drivers/net/bnxt/bnxt_txr.h
> +++ b/drivers/net/bnxt/bnxt_txr.h
> @@ -44,33 +44,6 @@ static inline uint32_t bnxt_tx_avail(struct bnxt_tx_queue
> *txq)
> bnxt_tx_bds_in_hw(txq)) - 1);
> }
>
> -/*
> - * Transmit completion function for use when DEV_TX_OFFLOAD_MBUF_FAST_FREE
> - * is enabled.
> - */
> -static inline void
> -bnxt_tx_cmp_fast(struct bnxt_tx_queue *txq, int nr_pkts)
> -{
> - struct bnxt_tx_ring_info *txr = txq->tx_ring;
> - uint32_t ring_mask = txr->tx_ring_struct->ring_mask;
> - struct rte_mbuf **free = txq->free;
> - uint16_t cons = txr->tx_cons;
> - unsigned int blk = 0;
> -
> - while (nr_pkts--) {
> - struct bnxt_sw_tx_bd *tx_buf;
> -
> - tx_buf = &txr->tx_buf_ring[cons];
> - cons = (cons + 1) & ring_mask;
> - free[blk++] = tx_buf->mbuf;
> - tx_buf->mbuf = NULL;
> - }
> - if (blk)
> - rte_mempool_put_bulk(free[0]->pool, (void **)free, blk);
> -
> - txr->tx_cons = cons;
> -}
> -
> void bnxt_free_tx_rings(struct bnxt *bp);
> int bnxt_init_one_tx_ring(struct bnxt_tx_queue *txq);
> int bnxt_init_tx_ring_struct(struct bnxt_tx_queue *txq, unsigned int
> socket_id);
> --
> 2.25.1
>
