net/mlx4_en: fix spurious timestamping callbacks

commit fc96256c upstream.

When multiple skb are TX-completed in a row, we might incorrectly keep
a timestamp of a prior skb and cause extra work.

Fixes: ec693d47 ("net/mlx4_en: Add HW timestamping (TS) support")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Reviewed-by: Eran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>

drivers/net/ethernet/mellanox/mlx4/en_tx.c

@@ -371,7 +371,6 @@ static bool mlx4_en_process_tx_cq(struct net_device *dev,
 	u32 packets = 0;
 	u32 bytes = 0;
 	int factor = priv->cqe_factor;
-	u64 timestamp = 0;
 	int done = 0;
 	int budget = priv->tx_work_limit;
 
@@ -405,9 +404,12 @@ static bool mlx4_en_process_tx_cq(struct net_device *dev,
 		new_index = be16_to_cpu(cqe->wqe_index) & size_mask;
 
 		do {
+			u64 timestamp = 0;
+
 			txbbs_skipped += ring->last_nr_txbb;
 			ring_index = (ring_index + ring->last_nr_txbb) & size_mask;
-			if (ring->tx_info[ring_index].ts_requested)
+
+			if (unlikely(ring->tx_info[ring_index].ts_requested))
 				timestamp = mlx4_en_get_cqe_ts(cqe);
 
 			/* free next descriptor */