octeon_ep: update tx/rx stats locally for persistence

Update tx/rx stats locally, so that ndo_get_stats64()
can use that and not rely on per queue resources to obtain statistics.
The latter used to cause race conditions when the device stopped.

Signed-off-by: Shinas Rasheed <[email protected]>
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>

Author: shinas-marvell

drivers/net/ethernet/marvell/octeon_ep/octep_ethtool.c

@@ -150,17 +150,14 @@ octep_get_ethtool_stats(struct net_device *netdev,
 				    iface_rx_stats,
 				    iface_tx_stats);
 
-	for (q = 0; q < oct->num_oqs; q++) {
-		struct octep_iq *iq = oct->iq[q];
-		struct octep_oq *oq = oct->oq[q];
-
-		tx_packets += iq->stats.instr_completed;
-		tx_bytes += iq->stats.bytes_sent;
-		tx_busy_errors += iq->stats.tx_busy;
-
-		rx_packets += oq->stats.packets;
-		rx_bytes += oq->stats.bytes;
-		rx_alloc_errors += oq->stats.alloc_failures;
+	for (q = 0; q < OCTEP_MAX_QUEUES; q++) {
+		tx_packets += oct->stats_iq[q].instr_completed;
+		tx_bytes += oct->stats_iq[q].bytes_sent;
+		tx_busy_errors += oct->stats_iq[q].tx_busy;
+
+		rx_packets += oct->stats_oq[q].packets;
+		rx_bytes += oct->stats_oq[q].bytes;
+		rx_alloc_errors += oct->stats_oq[q].alloc_failures;
 	}
 	i = 0;
 	data[i++] = rx_packets;
@@ -198,22 +195,18 @@ octep_get_ethtool_stats(struct net_device *netdev,
 	data[i++] = iface_rx_stats->err_pkts;
 
 	/* Per Tx Queue stats */
-	for (q = 0; q < oct->num_iqs; q++) {
-		struct octep_iq *iq = oct->iq[q];
-
-		data[i++] = iq->stats.instr_posted;
-		data[i++] = iq->stats.instr_completed;
-		data[i++] = iq->stats.bytes_sent;
-		data[i++] = iq->stats.tx_busy;
+	for (q = 0; q < OCTEP_MAX_QUEUES; q++) {
+		data[i++] = oct->stats_iq[q].instr_posted;
+		data[i++] = oct->stats_iq[q].instr_completed;
+		data[i++] = oct->stats_iq[q].bytes_sent;
+		data[i++] = oct->stats_iq[q].tx_busy;
 	}
 
 	/* Per Rx Queue stats */
-	for (q = 0; q < oct->num_oqs; q++) {
-		struct octep_oq *oq = oct->oq[q];
-
-		data[i++] = oq->stats.packets;
-		data[i++] = oq->stats.bytes;
-		data[i++] = oq->stats.alloc_failures;
+	for (q = 0; q < OCTEP_MAX_QUEUES; q++) {
+		data[i++] = oct->stats_oq[q].packets;
+		data[i++] = oct->stats_oq[q].bytes;
+		data[i++] = oct->stats_oq[q].alloc_failures;
 	}
 }
 

drivers/net/ethernet/marvell/octeon_ep/octep_main.c

@@ -822,7 +822,7 @@ static inline int octep_iq_full_check(struct octep_iq *iq)
 	if (unlikely(IQ_INSTR_SPACE(iq) >
 		     OCTEP_WAKE_QUEUE_THRESHOLD)) {
 		netif_start_subqueue(iq->netdev, iq->q_no);
-		iq->stats.restart_cnt++;
+		iq->stats->restart_cnt++;
 		return 0;
 	}
 
@@ -960,7 +960,7 @@ static netdev_tx_t octep_start_xmit(struct sk_buff *skb,
 	wmb();
 	/* Ring Doorbell to notify the NIC of new packets */
 	writel(iq->fill_cnt, iq->doorbell_reg);
-	iq->stats.instr_posted += iq->fill_cnt;
+	iq->stats->instr_posted += iq->fill_cnt;
 	iq->fill_cnt = 0;
 	return NETDEV_TX_OK;
 
@@ -991,22 +991,19 @@ static netdev_tx_t octep_start_xmit(struct sk_buff *skb,
 static void octep_get_stats64(struct net_device *netdev,
 			      struct rtnl_link_stats64 *stats)
 {
-	u64 tx_packets, tx_bytes, rx_packets, rx_bytes;
 	struct octep_device *oct = netdev_priv(netdev);
+	u64 tx_packets, tx_bytes, rx_packets, rx_bytes;
 	int q;
 
 	tx_packets = 0;
 	tx_bytes = 0;
 	rx_packets = 0;
 	rx_bytes = 0;
-	for (q = 0; q < oct->num_oqs; q++) {
-		struct octep_iq *iq = oct->iq[q];
-		struct octep_oq *oq = oct->oq[q];
-
-		tx_packets += iq->stats.instr_completed;
-		tx_bytes += iq->stats.bytes_sent;
-		rx_packets += oq->stats.packets;
-		rx_bytes += oq->stats.bytes;
+	for (q = 0; q < OCTEP_MAX_QUEUES; q++) {
+		tx_packets += oct->stats_iq[q].instr_completed;
+		tx_bytes += oct->stats_iq[q].bytes_sent;
+		rx_packets += oct->stats_oq[q].packets;
+		rx_bytes += oct->stats_oq[q].bytes;
 	}
 	stats->tx_packets = tx_packets;
 	stats->tx_bytes = tx_bytes;

drivers/net/ethernet/marvell/octeon_ep/octep_main.h

@@ -257,11 +257,17 @@ struct octep_device {
 	/* Pointers to Octeon Tx queues */
 	struct octep_iq *iq[OCTEP_MAX_IQ];
 
+	/* Per iq stats */
+	struct octep_iq_stats stats_iq[OCTEP_MAX_IQ];
+
 	/* Rx queues (OQ: Output Queue) */
 	u16 num_oqs;
 	/* Pointers to Octeon Rx queues */
 	struct octep_oq *oq[OCTEP_MAX_OQ];
 
+	/* Per oq stats */
+	struct octep_oq_stats stats_oq[OCTEP_MAX_OQ];
+
 	/* Hardware port number of the PCIe interface */
 	u16 pcie_port;
 

drivers/net/ethernet/marvell/octeon_ep/octep_rx.c

@@ -87,7 +87,7 @@ static int octep_oq_refill(struct octep_device *oct, struct octep_oq *oq)
 		page = dev_alloc_page();
 		if (unlikely(!page)) {
 			dev_err(oq->dev, "refill: rx buffer alloc failed\n");
-			oq->stats.alloc_failures++;
+			oq->stats->alloc_failures++;
 			break;
 		}
 
@@ -98,7 +98,7 @@ static int octep_oq_refill(struct octep_device *oct, struct octep_oq *oq)
 				"OQ-%d buffer refill: DMA mapping error!\n",
 				oq->q_no);
 			put_page(page);
-			oq->stats.alloc_failures++;
+			oq->stats->alloc_failures++;
 			break;
 		}
 		oq->buff_info[refill_idx].page = page;
@@ -134,6 +134,7 @@ static int octep_setup_oq(struct octep_device *oct, int q_no)
 	oq->netdev = oct->netdev;
 	oq->dev = &oct->pdev->dev;
 	oq->q_no = q_no;
+	oq->stats = &oct->stats_oq[q_no];
 	oq->max_count = CFG_GET_OQ_NUM_DESC(oct->conf);
 	oq->ring_size_mask = oq->max_count - 1;
 	oq->buffer_size = CFG_GET_OQ_BUF_SIZE(oct->conf);
@@ -443,7 +444,7 @@ static int __octep_oq_process_rx(struct octep_device *oct,
 		if (!skb) {
 			octep_oq_drop_rx(oq, buff_info,
 					 &read_idx, &desc_used);
-			oq->stats.alloc_failures++;
+			oq->stats->alloc_failures++;
 			continue;
 		}
 		skb_reserve(skb, data_offset);
@@ -494,8 +495,8 @@ static int __octep_oq_process_rx(struct octep_device *oct,
 
 	oq->host_read_idx = read_idx;
 	oq->refill_count += desc_used;
-	oq->stats.packets += pkt;
-	oq->stats.bytes += rx_bytes;
+	oq->stats->packets += pkt;
+	oq->stats->bytes += rx_bytes;
 
 	return pkt;
 }

drivers/net/ethernet/marvell/octeon_ep/octep_rx.h

@@ -186,8 +186,8 @@ struct octep_oq {
 	 */
 	u8 __iomem *pkts_sent_reg;
 
-	/* Statistics for this OQ. */
-	struct octep_oq_stats stats;
+	/* Pointer to statistics for this OQ. */
+	struct octep_oq_stats *stats;
 
 	/* Packets pending to be processed */
 	u32 pkts_pending;

drivers/net/ethernet/marvell/octeon_ep/octep_tx.c

@@ -81,9 +81,9 @@ int octep_iq_process_completions(struct octep_iq *iq, u16 budget)
 	}
 
 	iq->pkts_processed += compl_pkts;
-	iq->stats.instr_completed += compl_pkts;
-	iq->stats.bytes_sent += compl_bytes;
-	iq->stats.sgentry_sent += compl_sg;
+	iq->stats->instr_completed += compl_pkts;
+	iq->stats->bytes_sent += compl_bytes;
+	iq->stats->sgentry_sent += compl_sg;
 	iq->flush_index = fi;
 
 	netdev_tx_completed_queue(iq->netdev_q, compl_pkts, compl_bytes);
@@ -187,6 +187,7 @@ static int octep_setup_iq(struct octep_device *oct, int q_no)
 	iq->netdev = oct->netdev;
 	iq->dev = &oct->pdev->dev;
 	iq->q_no = q_no;
+	iq->stats = &oct->stats_iq[q_no];
 	iq->max_count = CFG_GET_IQ_NUM_DESC(oct->conf);
 	iq->ring_size_mask = iq->max_count - 1;
 	iq->fill_threshold = CFG_GET_IQ_DB_MIN(oct->conf);

drivers/net/ethernet/marvell/octeon_ep/octep_tx.h

@@ -170,8 +170,8 @@ struct octep_iq {
 	 */
 	u16 flush_index;
 
-	/* Statistics for this input queue. */
-	struct octep_iq_stats stats;
+	/* Pointer to statistics for this input queue. */
+	struct octep_iq_stats *stats;
 
 	/* Pointer to the Virtual Base addr of the input ring. */
 	struct octep_tx_desc_hw *desc_ring;