drivers/xen/*net*: use skb_is_gso()

Suggested-by: Paul Durrant <Paul.Durrant@xxxxxxxxxx>
Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>

--- a/drivers/xen/netback/netback.c
+++ b/drivers/xen/netback/netback.c
@@ -322,7 +322,7 @@ int netif_be_start_xmit(struct sk_buff *
 	}
 
 	netif->rx_req_cons_peek += skb_shinfo(skb)->nr_frags + 1 +
-				   !!skb_shinfo(skb)->gso_size;
+				   !!skb_is_gso(skb);
 	netif_get(netif);
 
 	if (netbk_can_queue(dev) && netbk_queue_full(netif)) {
--- a/drivers/xen/netfront/netfront.c
+++ b/drivers/xen/netfront/netfront.c
@@ -1019,7 +1019,7 @@ static int network_start_xmit(struct sk_
 #endif
 
 #if HAVE_TSO
-	if (skb_shinfo(skb)->gso_size) {
+	if (skb_is_gso(skb)) {
 		struct netif_extra_info *gso = (struct netif_extra_info *)
 			RING_GET_REQUEST(&np->tx, ++i);
 
--- a/drivers/xen/sfc_netfront/accel_vi.c
+++ b/drivers/xen/sfc_netfront/accel_vi.c
@@ -676,13 +676,11 @@ netfront_accel_vi_tx_post(netfront_accel
 	VPRINTK("%s: %d bytes, gso %d\n", __FUNCTION__, skb->len, 
 		skb_shinfo(skb)->gso_size);
 	
-	if (skb_shinfo(skb)->gso_size) {
+	if (skb_is_gso(skb))
 		return netfront_accel_enqueue_skb_tso(vnic, skb);
-	}
 
-	if (skb->len <= NETFRONT_ACCEL_TX_BUF_LENGTH) {
+	if (skb->len <= NETFRONT_ACCEL_TX_BUF_LENGTH)
 		return netfront_accel_enqueue_skb_single(vnic, skb);
-	}
 
 	return netfront_accel_enqueue_skb_multi(vnic, skb);
 }