bcm63xx_enet: switch to napi_build_skb() to reuse skbuff_heads

napi_build_skb() reuses NAPI skbuff_head cache in order to save some
cycles on freeing/allocating skbuff_heads on every new Rx or completed
Tx.
Use napi_consume_skb() to feed the cache with skbuff_heads of completed
Tx so it's never empty.

Signed-off-by: Sieng Piaw Liew <liew.s.piaw@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/broadcom/bcm63xx_enet.c

@@ -388,7 +388,7 @@ static int bcm_enet_receive_queue(struct net_device *dev, int budget)
 					 priv->rx_buf_size, DMA_FROM_DEVICE);
 			priv->rx_buf[desc_idx] = NULL;
 
-			skb = build_skb(buf, priv->rx_frag_size);
+			skb = napi_build_skb(buf, priv->rx_frag_size);
 			if (unlikely(!skb)) {
 				skb_free_frag(buf);
 				dev->stats.rx_dropped++;
@@ -468,7 +468,7 @@ static int bcm_enet_tx_reclaim(struct net_device *dev, int force)
 			dev->stats.tx_errors++;
 
 		bytes += skb->len;
-		dev_kfree_skb(skb);
+		napi_consume_skb(skb, !force);
 		released++;
 	}