sfc: implement per-queue TSO (hw_gso) stats

Use our existing TSO stats, which count enqueued TSO TXes.
Users may expect them to count completions, as tx-packets and
tx-bytes do; however, these are the counters we have, and the
qstats documentation doesn't actually specify.

Reviewed-by: Jacob Keller <[e-mail]>
Signed-off-by: Edward Cree <[e-mail]>
Signed-off-by: David S. Miller <[e-mail]>

drivers/net/ethernet/sfc/efx.c

···
 	channel = efx_get_tx_channel(efx, idx);
 	stats->packets = 0;
 	stats->bytes = 0;
+	stats->hw_gso_packets = 0;
+	stats->hw_gso_wire_packets = 0;
 	efx_for_each_channel_tx_queue(tx_queue, channel) {
 		stats->packets += tx_queue->complete_packets -
 				  tx_queue->old_complete_packets;
 		stats->bytes += tx_queue->complete_bytes -
 				tx_queue->old_complete_bytes;
+		/* Note that, unlike stats->packets and stats->bytes,
+		 * these count TXes enqueued, rather than completed,
+		 * which may not be what users expect.
+		 */
+		stats->hw_gso_packets += tx_queue->tso_bursts -
+					 tx_queue->old_tso_bursts;
+		stats->hw_gso_wire_packets += tx_queue->tso_packets -
+					      tx_queue->old_tso_packets;
 	}
 }
 
···
 	rx->hw_drop_overruns = 0;
 	tx->packets = 0;
 	tx->bytes = 0;
+	tx->hw_gso_packets = 0;
+	tx->hw_gso_wire_packets = 0;
 
 	/* Count all packets on non-core queues, and packets before last
 	 * datapath start on core queues.
···
 						net_dev->real_num_tx_queues) {
 				tx->packets += tx_queue->complete_packets;
 				tx->bytes += tx_queue->complete_bytes;
+				tx->hw_gso_packets += tx_queue->tso_bursts;
+				tx->hw_gso_wire_packets += tx_queue->tso_packets;
 			} else {
 				tx->packets += tx_queue->old_complete_packets;
 				tx->bytes += tx_queue->old_complete_bytes;
+				tx->hw_gso_packets += tx_queue->old_tso_bursts;
+				tx->hw_gso_wire_packets += tx_queue->old_tso_packets;
 			}
 			/* Include XDP TX in device-wide stats */
 			tx->packets += tx_queue->complete_xdp_packets;

drivers/net/ethernet/sfc/net_driver.h

···
  *	efx_init_tx_queue()
  * @old_complete_bytes: Value of @complete_bytes as of last
  *	efx_init_tx_queue()
+ * @old_tso_bursts: Value of @tso_bursts as of last efx_init_tx_queue()
+ * @old_tso_packets: Value of @tso_packets as of last efx_init_tx_queue()
  * @read_count: Current read pointer.
  *	This is the number of buffers that have been removed from both rings.
  * @old_write_count: The value of @write_count when last checked.
···
 	bool xdp_tx;
 	unsigned long old_complete_packets;
 	unsigned long old_complete_bytes;
+	unsigned int old_tso_bursts;
+	unsigned int old_tso_packets;
 
 	/* Members used mainly on the completion path */
 	unsigned int read_count ____cacheline_aligned_in_smp;

drivers/net/ethernet/sfc/tx_common.c

···
 
 	tx_queue->old_complete_packets = tx_queue->complete_packets;
 	tx_queue->old_complete_bytes = tx_queue->complete_bytes;
+	tx_queue->old_tso_bursts = tx_queue->tso_bursts;
+	tx_queue->old_tso_packets = tx_queue->tso_packets;
 
 	tx_queue->xdp_tx = efx_channel_is_xdp_tx(tx_queue->channel);
 	tx_queue->tso_version = 0;