netdev - [PATCH net-next] net: netsec: add xdp tx return bulking support

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 17 Nov 2020 10:35:28 +0100
From:   Lorenzo Bianconi <lorenzo@...nel.org>
To:     netdev@...r.kernel.org
Cc:     lorenzo.bianconi@...hat.com, davem@...emloft.net, kuba@...nel.org,
        brouer@...hat.com, ilias.apalodimas@...aro.org
Subject: [PATCH net-next] net: netsec: add xdp tx return bulking support

Convert netsec driver to xdp_return_frame_bulk APIs.
Rely on xdp_return_frame_rx_napi for XDP_TX in order to try to recycle
the page in the "in-irq" page_pool cache.

Co-developed-by: Jesper Dangaard Brouer <brouer@...hat.com>
Signed-off-by: Jesper Dangaard Brouer <brouer@...hat.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@...nel.org>
---
This patch is just compile tested, I have not carried out any run test
---
 drivers/net/ethernet/socionext/netsec.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/socionext/netsec.c b/drivers/net/ethernet/socionext/netsec.c
index 1503cc9ec6e2..536aa8961dc6 100644
--- a/drivers/net/ethernet/socionext/netsec.c
+++ b/drivers/net/ethernet/socionext/netsec.c
@@ -631,6 +631,7 @@ static void netsec_set_rx_de(struct netsec_priv *priv,
 static bool netsec_clean_tx_dring(struct netsec_priv *priv)
 {
 	struct netsec_desc_ring *dring = &priv->desc_ring[NETSEC_RING_TX];
+	struct xdp_frame_bulk bq;
 	struct netsec_de *entry;
 	int tail = dring->tail;
 	unsigned int bytes;
@@ -639,8 +640,11 @@ static bool netsec_clean_tx_dring(struct netsec_priv *priv)
 	spin_lock(&dring->lock);
 
 	bytes = 0;
+	xdp_frame_bulk_init(&bq);
 	entry = dring->vaddr + DESC_SZ * tail;
 
+	rcu_read_lock(); /* need for xdp_return_frame_bulk */
+
 	while (!(entry->attr & (1U << NETSEC_TX_SHIFT_OWN_FIELD)) &&
 	       cnt < DESC_NUM) {
 		struct netsec_desc *desc;
@@ -665,7 +669,10 @@ static bool netsec_clean_tx_dring(struct netsec_priv *priv)
 			dev_kfree_skb(desc->skb);
 		} else {
 			bytes += desc->xdpf->len;
-			xdp_return_frame(desc->xdpf);
+			if (desc->buf_type == TYPE_NETSEC_XDP_TX)
+				xdp_return_frame_rx_napi(desc->xdpf);
+			else
+				xdp_return_frame_bulk(desc->xdpf, &bq);
 		}
 next:
 		/* clean up so netsec_uninit_pkt_dring() won't free the skb
@@ -684,6 +691,9 @@ static bool netsec_clean_tx_dring(struct netsec_priv *priv)
 		entry = dring->vaddr + DESC_SZ * tail;
 		cnt++;
 	}
+	xdp_flush_frame_bulk(&bq);
+
+	rcu_read_unlock();
 
 	spin_unlock(&dring->lock);
 
-- 
2.26.2