net: hns3: optimize the barrier using when cleaning tx BD

driver inclusion
category: bugfix
bugzilla: NA
CVE: NA

Currently, a barrier is used when cleaning each tx BD, which may
cause performance degradation.

This patch optimize it to use one barrier when cleaning tx BD
each round.

Feature or Bugfix:Bugfix
Signed-off-by: NYunsheng Lin <linyunsheng@huawei.com>
Reviewed-by: Nlipeng <lipeng321@huawei.com>
Reviewed-by: NXie XiuQi <xiexiuqi@huawei.com>
Signed-off-by: NYang Yingliang <yangyingliang@huawei.com>

drivers/net/ethernet/hisilicon/hns3/hns3_enet.c

@@ -2287,20 +2287,25 @@ static void hns3_reuse_buffer(struct hns3_enet_ring *ring, int i)
 	ring->desc[i].rx.bd_base_info = 0;
 }
 
-static void hns3_nic_reclaim_one_desc(struct hns3_enet_ring *ring, int *bytes,
-				      int *pkts)
+static void hns3_nic_reclaim_desc(struct hns3_enet_ring *ring, int head,
+				  int *bytes, int *pkts)
 {
 	int ntc = ring->next_to_clean;
 	struct hns3_desc_cb *desc_cb;
 
-	desc_cb = &ring->desc_cb[ntc];
-	(*pkts) += (desc_cb->type == DESC_TYPE_SKB);
-	(*bytes) += desc_cb->length;
-	/* desc_cb will be cleaned, after hnae3_free_buffer_detach */
-	hns3_free_buffer_detach(ring, ntc);
+	while (head != ntc) {
+		desc_cb = &ring->desc_cb[ntc];
+		(*pkts) += (desc_cb->type == DESC_TYPE_SKB);
+		(*bytes) += desc_cb->length;
+		/* desc_cb will be cleaned, after hnae3_free_buffer_detach */
+		hns3_free_buffer_detach(ring, ntc);
 
-	if (++ntc == ring->desc_num)
-		ntc = 0;
+		if (++ntc == ring->desc_num)
+			ntc = 0;
+
+		/* Issue prefetch for next Tx descriptor */
+		prefetch(&ring->desc_cb[ntc]);
+	}
 
 	/* This smp_store_release() pairs with smp_load_acquire() in
 	 * ring_space called by hns3_nic_net_xmit.
@@ -2345,11 +2350,7 @@ void hns3_clean_tx_ring(struct hns3_enet_ring *ring)
 
 	bytes = 0;
 	pkts = 0;
-	while (head != ring->next_to_clean) {
-		hns3_nic_reclaim_one_desc(ring, &bytes, &pkts);
-		/* Issue prefetch for next Tx descriptor */
-		prefetch(&ring->desc_cb[ring->next_to_clean]);
-	}
+	hns3_nic_reclaim_desc(ring, head, &bytes, &pkts);
 
 	ring->tqp_vector->tx_group.total_bytes += bytes;
 	ring->tqp_vector->tx_group.total_packets += pkts;