From: Tariq Toukan <tar...@mellanox.com>

A loaded XDP program might write to the xdp_frame data area,
prefetchw() it to avoid a potential cache miss.

Performance tests:
ConnectX-5, XDP_TX packet rate, single ring.
CPU: Intel(R) Xeon(R) CPU E5-2680 v3 @ 2.50GHz

Before: 13,172,976 pps
After:  13,456,248 pps
2% gain.

Fixes: 22f453988194 ("net/mlx5e: Support XDP over Striding RQ")
Signed-off-by: Tariq Toukan <tar...@mellanox.com>
Signed-off-by: Saeed Mahameed <sae...@mellanox.com>
---
 drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c 
b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
index e33ca03b2100..15d8ae28c040 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c
@@ -1099,6 +1099,7 @@ mlx5e_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq, 
struct mlx5e_mpw_info *wi,
 
        dma_sync_single_range_for_cpu(rq->pdev, di->addr, head_offset,
                                      frag_size, DMA_FROM_DEVICE);
+       prefetchw(va); /* xdp_frame data area */
        prefetch(data);
 
        rcu_read_lock();
-- 
2.17.0

Reply via email to