sfc: complete the next packet when we receive a timestamp

We now ignore the "completion" event when using tx queue timestamping,
and only pay attention to the two (high and low) timestamp events. The
NIC will send a pair of timestamp events for every packet transmitted.
The current firmware may merge the completion events, and it is possible
that future versions may reorder the completion and timestamp events.
As such the completion event is not useful.

Without this patch in place a merged completion event on a queue with
timestamping will cause a "spurious TX completion" error. This affects
SFN8000-series adapters.

Signed-off-by: Tom Zhao <[email protected]>
Acked-by: Martin Habets <[email protected]>
Signed-off-by: David S. Miller <[email protected]>

Author: tomz-xilinx

drivers/net/ethernet/sfc/ef10.c

@@ -2853,29 +2853,32 @@ efx_ef10_handle_tx_event(struct efx_channel *channel, efx_qword_t *event)
 	}
 
 	/* Transmit timestamps are only available for 8XXX series. They result
-	 * in three events per packet. These occur in order, and are:
-	 *  - the normal completion event
+	 * in up to three events per packet. These occur in order, and are:
+	 *  - the normal completion event (may be omitted)
 	 *  - the low part of the timestamp
 	 *  - the high part of the timestamp
 	 *
+	 * It's possible for multiple completion events to appear before the
+	 * corresponding timestamps. So we can for example get:
+	 *  COMP N
+	 *  COMP N+1
+	 *  TS_LO N
+	 *  TS_HI N
+	 *  TS_LO N+1
+	 *  TS_HI N+1
+	 *
+	 * In addition it's also possible for the adjacent completions to be
+	 * merged, so we may not see COMP N above. As such, the completion
+	 * events are not very useful here.
+	 *
 	 * Each part of the timestamp is itself split across two 16 bit
 	 * fields in the event.
 	 */
 	tx_ev_type = EFX_QWORD_FIELD(*event, ESF_EZ_TX_SOFT1);
 
 	switch (tx_ev_type) {
 	case TX_TIMESTAMP_EVENT_TX_EV_COMPLETION:
-		/* In case of Queue flush or FLR, we might have received
-		 * the previous TX completion event but not the Timestamp
-		 * events.
-		 */
-		if (tx_queue->completed_desc_ptr != tx_queue->ptr_mask)
-			efx_xmit_done(tx_queue, tx_queue->completed_desc_ptr);
-
-		tx_ev_desc_ptr = EFX_QWORD_FIELD(*event,
-						 ESF_DZ_TX_DESCR_INDX);
-		tx_queue->completed_desc_ptr =
-					tx_ev_desc_ptr & tx_queue->ptr_mask;
+		/* Ignore this event - see above. */
 		break;
 
 	case TX_TIMESTAMP_EVENT_TX_EV_TSTAMP_LO:
@@ -2887,8 +2890,7 @@ efx_ef10_handle_tx_event(struct efx_channel *channel, efx_qword_t *event)
 		ts_part = efx_ef10_extract_event_ts(event);
 		tx_queue->completed_timestamp_major = ts_part;
 
-		efx_xmit_done(tx_queue, tx_queue->completed_desc_ptr);
-		tx_queue->completed_desc_ptr = tx_queue->ptr_mask;
+		efx_xmit_done_single(tx_queue);
 		break;
 
 	default:

drivers/net/ethernet/sfc/efx.h

@@ -20,6 +20,7 @@ netdev_tx_t efx_hard_start_xmit(struct sk_buff *skb,
 				struct net_device *net_dev);
 netdev_tx_t efx_enqueue_skb(struct efx_tx_queue *tx_queue, struct sk_buff *skb);
 void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index);
+void efx_xmit_done_single(struct efx_tx_queue *tx_queue);
 int efx_setup_tc(struct net_device *net_dev, enum tc_setup_type type,
 		 void *type_data);
 extern unsigned int efx_piobuf_size;

drivers/net/ethernet/sfc/net_driver.h

@@ -208,8 +208,6 @@ struct efx_tx_buffer {
  *	avoid cache-line ping-pong between the xmit path and the
  *	completion path.
  * @merge_events: Number of TX merged completion events
- * @completed_desc_ptr: Most recent completed pointer - only used with
- *      timestamping.
  * @completed_timestamp_major: Top part of the most recent tx timestamp.
  * @completed_timestamp_minor: Low part of the most recent tx timestamp.
  * @insert_count: Current insert pointer
@@ -269,7 +267,6 @@ struct efx_tx_queue {
 	unsigned int merge_events;
 	unsigned int bytes_compl;
 	unsigned int pkts_compl;
-	unsigned int completed_desc_ptr;
 	u32 completed_timestamp_major;
 	u32 completed_timestamp_minor;
 

drivers/net/ethernet/sfc/tx.c

@@ -535,6 +535,44 @@ netdev_tx_t efx_hard_start_xmit(struct sk_buff *skb,
 	return efx_enqueue_skb(tx_queue, skb);
 }
 
+void efx_xmit_done_single(struct efx_tx_queue *tx_queue)
+{
+	unsigned int pkts_compl = 0, bytes_compl = 0;
+	unsigned int read_ptr;
+	bool finished = false;
+
+	read_ptr = tx_queue->read_count & tx_queue->ptr_mask;
+
+	while (!finished) {
+		struct efx_tx_buffer *buffer = &tx_queue->buffer[read_ptr];
+
+		if (!efx_tx_buffer_in_use(buffer)) {
+			struct efx_nic *efx = tx_queue->efx;
+
+			netif_err(efx, hw, efx->net_dev,
+				  "TX queue %d spurious single TX completion\n",
+				  tx_queue->queue);
+			efx_schedule_reset(efx, RESET_TYPE_TX_SKIP);
+			return;
+		}
+
+		/* Need to check the flag before dequeueing. */
+		if (buffer->flags & EFX_TX_BUF_SKB)
+			finished = true;
+		efx_dequeue_buffer(tx_queue, buffer, &pkts_compl, &bytes_compl);
+
+		++tx_queue->read_count;
+		read_ptr = tx_queue->read_count & tx_queue->ptr_mask;
+	}
+
+	tx_queue->pkts_compl += pkts_compl;
+	tx_queue->bytes_compl += bytes_compl;
+
+	EFX_WARN_ON_PARANOID(pkts_compl != 1);
+
+	efx_xmit_done_check_empty(tx_queue);
+}
+
 void efx_init_tx_queue_core_txq(struct efx_tx_queue *tx_queue)
 {
 	struct efx_nic *efx = tx_queue->efx;

drivers/net/ethernet/sfc/tx_common.c

@@ -80,7 +80,6 @@ void efx_init_tx_queue(struct efx_tx_queue *tx_queue)
 	tx_queue->xmit_more_available = false;
 	tx_queue->timestamping = (efx_ptp_use_mac_tx_timestamps(efx) &&
 				  tx_queue->channel == efx_ptp_channel(efx));
-	tx_queue->completed_desc_ptr = tx_queue->ptr_mask;
 	tx_queue->completed_timestamp_major = 0;
 	tx_queue->completed_timestamp_minor = 0;
 
@@ -210,10 +209,9 @@ static void efx_dequeue_buffers(struct efx_tx_queue *tx_queue,
 	while (read_ptr != stop_index) {
 		struct efx_tx_buffer *buffer = &tx_queue->buffer[read_ptr];
 
-		if (!(buffer->flags & EFX_TX_BUF_OPTION) &&
-		    unlikely(buffer->len == 0)) {
+		if (!efx_tx_buffer_in_use(buffer)) {
 			netif_err(efx, tx_err, efx->net_dev,
-				  "TX queue %d spurious TX completion id %x\n",
+				  "TX queue %d spurious TX completion id %d\n",
 				  tx_queue->queue, read_ptr);
 			efx_schedule_reset(efx, RESET_TYPE_TX_SKIP);
 			return;
@@ -226,6 +224,19 @@ static void efx_dequeue_buffers(struct efx_tx_queue *tx_queue,
 	}
 }
 
+void efx_xmit_done_check_empty(struct efx_tx_queue *tx_queue)
+{
+	if ((int)(tx_queue->read_count - tx_queue->old_write_count) >= 0) {
+		tx_queue->old_write_count = READ_ONCE(tx_queue->write_count);
+		if (tx_queue->read_count == tx_queue->old_write_count) {
+			/* Ensure that read_count is flushed. */
+			smp_mb();
+			tx_queue->empty_read_count =
+				tx_queue->read_count | EFX_EMPTY_COUNT_VALID;
+		}
+	}
+}
+
 void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index)
 {
 	unsigned int fill_level, pkts_compl = 0, bytes_compl = 0;
@@ -256,15 +267,7 @@ void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index)
 			netif_tx_wake_queue(tx_queue->core_txq);
 	}
 
-	/* Check whether the hardware queue is now empty */
-	if ((int)(tx_queue->read_count - tx_queue->old_write_count) >= 0) {
-		tx_queue->old_write_count = READ_ONCE(tx_queue->write_count);
-		if (tx_queue->read_count == tx_queue->old_write_count) {
-			smp_mb();
-			tx_queue->empty_read_count =
-				tx_queue->read_count | EFX_EMPTY_COUNT_VALID;
-		}
-	}
+	efx_xmit_done_check_empty(tx_queue);
 }
 
 /* Remove buffers put into a tx_queue for the current packet.

drivers/net/ethernet/sfc/tx_common.h

@@ -21,6 +21,12 @@ void efx_dequeue_buffer(struct efx_tx_queue *tx_queue,
 			unsigned int *pkts_compl,
 			unsigned int *bytes_compl);
 
+static inline bool efx_tx_buffer_in_use(struct efx_tx_buffer *buffer)
+{
+	return buffer->len || (buffer->flags & EFX_TX_BUF_OPTION);
+}
+
+void efx_xmit_done_check_empty(struct efx_tx_queue *tx_queue);
 void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index);
 
 void efx_enqueue_unwind(struct efx_tx_queue *tx_queue,