[dpdk-dev] [PATCH] net/bnxt: disable receive end-of-packet padding

Lance Richardson lance.richardson at broadcom.com
Thu Dec 3 15:16:21 CET 2020


Testing has shown that the packet forwarding rate for packet sizes
that are not a multiple of the cache line size is reduced when the
DMA size is padded to a multiple of the cache line size. Improve
performance for these packet sizes by disabling EOP padding.

Fixes: f4253e97e7b0 ("net/bnxt: set padding flags in Rx descriptor")
Signed-off-by: Lance Richardson <lance.richardson at broadcom.com>
Reviewed-by: Ajit Kumar Khaparde <ajit.khaparde at broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur at broadcom.com>
---
 drivers/net/bnxt/bnxt_rxr.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/bnxt/bnxt_rxr.c b/drivers/net/bnxt/bnxt_rxr.c
index fdbe6f71e..af1774844 100644
--- a/drivers/net/bnxt/bnxt_rxr.c
+++ b/drivers/net/bnxt/bnxt_rxr.c
@@ -1113,7 +1113,7 @@ int bnxt_init_one_rx_ring(struct bnxt_rx_queue *rxq)
 	size = rte_pktmbuf_data_room_size(rxq->mb_pool) - RTE_PKTMBUF_HEADROOM;
 	size = RTE_MIN(BNXT_MAX_PKT_LEN, size);
 
-	type = RX_PROD_PKT_BD_TYPE_RX_PROD_PKT | RX_PROD_PKT_BD_FLAGS_EOP_PAD;
+	type = RX_PROD_PKT_BD_TYPE_RX_PROD_PKT;
 
 	rxr = rxq->rx_ring;
 	ring = rxr->rx_ring_struct;
-- 
2.25.1



More information about the dev mailing list