Merge branch 'mlx4-vf-counters'

Or Gerlitz says:

====================
mlx4 driver update (+ new VF ndo)

This series from Eran and Hadar is further dealing with traffic
counters in the mlx4 driver, this time mostly around SRIOV.

We added a new ndo to read the VF counters through the PF netdev
netlink infrastructure plus mlx4 implementation for that ndo.

changes from V0:
  - applied feedback from John to use nested netlink encoding
    for the VF counters so we can extend it later
  - add handling of single ported VFs in the mlx4_en driver new ndo
  - avoid chopping the FW counters from 64 to 32 bits in mlx4_en PF flow
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/infiniband/hw/mlx4/mad.c

@@ -64,14 +64,6 @@ enum {
 #define GUID_TBL_BLK_NUM_ENTRIES 8
 #define GUID_TBL_BLK_SIZE (GUID_TBL_ENTRY_SIZE * GUID_TBL_BLK_NUM_ENTRIES)
 
-/* Counters should be saturate once they reach their maximum value */
-#define ASSIGN_32BIT_COUNTER(counter, value) do {\
-	if ((value) > U32_MAX)			 \
-		counter = cpu_to_be32(U32_MAX); \
-	else					 \
-		counter = cpu_to_be32(value);	 \
-} while (0)
-
 struct mlx4_mad_rcv_buf {
 	struct ib_grh grh;
 	u8 payload[256];
@@ -828,31 +820,25 @@ static int iboe_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
 			struct ib_wc *in_wc, struct ib_grh *in_grh,
 			struct ib_mad *in_mad, struct ib_mad *out_mad)
 {
-	struct mlx4_cmd_mailbox *mailbox;
+	struct mlx4_counter counter_stats;
 	struct mlx4_ib_dev *dev = to_mdev(ibdev);
 	int err;
-	u32 inmod = dev->counters[port_num - 1] & 0xffff;
-	u8 mode;
 
 	if (in_mad->mad_hdr.mgmt_class != IB_MGMT_CLASS_PERF_MGMT)
 		return -EINVAL;
 
-	mailbox = mlx4_alloc_cmd_mailbox(dev->dev);
-	if (IS_ERR(mailbox))
-		return IB_MAD_RESULT_FAILURE;
-
-	err = mlx4_cmd_box(dev->dev, 0, mailbox->dma, inmod, 0,
-			   MLX4_CMD_QUERY_IF_STAT, MLX4_CMD_TIME_CLASS_C,
-			   MLX4_CMD_WRAPPED);
+	memset(&counter_stats, 0, sizeof(counter_stats));
+	err = mlx4_get_counter_stats(dev->dev,
+				     dev->counters[port_num - 1].index,
+				     &counter_stats, 0);
 	if (err)
 		err = IB_MAD_RESULT_FAILURE;
 	else {
 		memset(out_mad->data, 0, sizeof out_mad->data);
-		mode = ((struct mlx4_counter *)mailbox->buf)->counter_mode;
-		switch (mode & 0xf) {
+		switch (counter_stats.counter_mode & 0xf) {
 		case 0:
-			edit_counter(mailbox->buf,
-						(void *)(out_mad->data + 40));
+			edit_counter(&counter_stats,
+				     (void *)(out_mad->data + 40));
 			err = IB_MAD_RESULT_SUCCESS | IB_MAD_RESULT_REPLY;
 			break;
 		default:
@@ -860,8 +846,6 @@ static int iboe_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
 		}
 	}
 
-	mlx4_free_cmd_mailbox(dev->dev, mailbox);
-
 	return err;
 }
 
@@ -869,10 +853,12 @@ int mlx4_ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num,
 			struct ib_wc *in_wc, struct ib_grh *in_grh,
 			struct ib_mad *in_mad, struct ib_mad *out_mad)
 {
+	struct mlx4_ib_dev *dev = to_mdev(ibdev);
 	switch (rdma_port_get_link_layer(ibdev, port_num)) {
 	case IB_LINK_LAYER_INFINIBAND:
-		return ib_process_mad(ibdev, mad_flags, port_num, in_wc,
-				      in_grh, in_mad, out_mad);
+		if (!mlx4_is_slave(dev->dev))
+			return ib_process_mad(ibdev, mad_flags, port_num, in_wc,
+					      in_grh, in_mad, out_mad);
 	case IB_LINK_LAYER_ETHERNET:
 		return iboe_process_mad(ibdev, mad_flags, port_num, in_wc,
 					  in_grh, in_mad, out_mad);

drivers/infiniband/hw/mlx4/main.c

@@ -2098,6 +2098,8 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 	struct mlx4_ib_iboe *iboe;
 	int ib_num_ports = 0;
 	int num_req_counters;
+	int allocated;
+	u32 counter_index;
 
 	pr_info_once("%s", mlx4_ib_version);
 
@@ -2263,19 +2265,31 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 	num_req_counters = mlx4_is_bonded(dev) ? 1 : ibdev->num_ports;
 	for (i = 0; i < num_req_counters; ++i) {
 		mutex_init(&ibdev->qp1_proxy_lock[i]);
+		allocated = 0;
 		if (mlx4_ib_port_link_layer(&ibdev->ib_dev, i + 1) ==
 						IB_LINK_LAYER_ETHERNET) {
-			err = mlx4_counter_alloc(ibdev->dev, &ibdev->counters[i]);
+			err = mlx4_counter_alloc(ibdev->dev, &counter_index);
+			/* if failed to allocate a new counter, use default */
 			if (err)
-				ibdev->counters[i] = -1;
-		} else {
-			ibdev->counters[i] = -1;
+				counter_index =
+					mlx4_get_default_counter_index(dev,
+								       i + 1);
+			else
+				allocated = 1;
+		} else { /* IB_LINK_LAYER_INFINIBAND use the default counter */
+			counter_index = mlx4_get_default_counter_index(dev,
+								       i + 1);
 		}
+		ibdev->counters[i].index = counter_index;
+		ibdev->counters[i].allocated = allocated;
+		pr_info("counter index %d for port %d allocated %d\n",
+			counter_index, i + 1, allocated);
 	}
 	if (mlx4_is_bonded(dev))
-		for (i = 1; i < ibdev->num_ports ; ++i)
-			ibdev->counters[i] = ibdev->counters[0];
-
+		for (i = 1; i < ibdev->num_ports ; ++i) {
+			ibdev->counters[i].index = ibdev->counters[0].index;
+			ibdev->counters[i].allocated = 0;
+		}
 
 	mlx4_foreach_port(i, dev, MLX4_PORT_TYPE_IB)
 		ib_num_ports++;
@@ -2415,10 +2429,12 @@ err_steer_qp_release:
 		mlx4_qp_release_range(dev, ibdev->steer_qpn_base,
 				      ibdev->steer_qpn_count);
 err_counter:
-	for (; i; --i)
-		if (ibdev->counters[i - 1] != -1)
-			mlx4_counter_free(ibdev->dev, ibdev->counters[i - 1]);
-
+	for (i = 0; i < ibdev->num_ports; ++i) {
+		if (ibdev->counters[i].index != -1 &&
+		    ibdev->counters[i].allocated)
+			mlx4_counter_free(ibdev->dev,
+					  ibdev->counters[i].index);
+	}
 err_map:
 	iounmap(ibdev->uar_map);
 
@@ -2535,8 +2551,9 @@ static void mlx4_ib_remove(struct mlx4_dev *dev, void *ibdev_ptr)
 
 	iounmap(ibdev->uar_map);
 	for (p = 0; p < ibdev->num_ports; ++p)
-		if (ibdev->counters[p] != -1)
-			mlx4_counter_free(ibdev->dev, ibdev->counters[p]);
+		if (ibdev->counters[p].index != -1 &&
+		    ibdev->counters[p].allocated)
+			mlx4_counter_free(ibdev->dev, ibdev->counters[p].index);
 	mlx4_foreach_port(p, dev, MLX4_PORT_TYPE_IB)
 		mlx4_CLOSE_PORT(dev, p);
 

drivers/infiniband/hw/mlx4/mlx4_ib.h

@@ -503,6 +503,11 @@ struct mlx4_ib_iov_port {
 	struct mlx4_ib_iov_sysfs_attr mcg_dentry;
 };
 
+struct counter_index {
+	u32		index;
+	u8		allocated;
+};
+
 struct mlx4_ib_dev {
 	struct ib_device	ib_dev;
 	struct mlx4_dev	       *dev;
@@ -521,7 +526,7 @@ struct mlx4_ib_dev {
 	struct mutex		cap_mask_mutex;
 	bool			ib_active;
 	struct mlx4_ib_iboe	iboe;
-	int			counters[MLX4_MAX_PORTS];
+	struct counter_index    counters[MLX4_MAX_PORTS];
 	int		       *eq_table;
 	struct kobject	       *iov_parent;
 	struct kobject	       *ports_parent;

drivers/infiniband/hw/mlx4/qp.c

@@ -1539,12 +1539,13 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
 	}
 
 	if (cur_state == IB_QPS_INIT && new_state == IB_QPS_RTR) {
-		if (dev->counters[qp->port - 1] != -1) {
+		if (dev->counters[qp->port - 1].index != -1) {
 			context->pri_path.counter_index =
-						dev->counters[qp->port - 1];
+					dev->counters[qp->port - 1].index;
 			optpar |= MLX4_QP_OPTPAR_COUNTER_INDEX;
 		} else
-			context->pri_path.counter_index = 0xff;
+			context->pri_path.counter_index =
+				MLX4_SINK_COUNTER_INDEX(dev->dev);
 
 		if (qp->flags & MLX4_IB_QP_NETIF) {
 			mlx4_ib_steer_qp_reg(dev, qp, 1);

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

@@ -49,6 +49,7 @@
 #include "mlx4.h"
 #include "fw.h"
 #include "fw_qos.h"
+#include "mlx4_stats.h"
 
 #define CMD_POLL_TOKEN 0xffff
 #define INBOX_MASK	0xffffffffffffff00ULL
@@ -3166,6 +3167,92 @@ int mlx4_set_vf_link_state(struct mlx4_dev *dev, int port, int vf, int link_stat
 }
 EXPORT_SYMBOL_GPL(mlx4_set_vf_link_state);
 
+int mlx4_get_counter_stats(struct mlx4_dev *dev, int counter_index,
+			   struct mlx4_counter *counter_stats, int reset)
+{
+	struct mlx4_cmd_mailbox *mailbox = NULL;
+	struct mlx4_counter *tmp_counter;
+	int err;
+	u32 if_stat_in_mod;
+
+	if (!counter_stats)
+		return -EINVAL;
+
+	if (counter_index == MLX4_SINK_COUNTER_INDEX(dev))
+		return 0;
+
+	mailbox = mlx4_alloc_cmd_mailbox(dev);
+	if (IS_ERR(mailbox))
+		return PTR_ERR(mailbox);
+
+	memset(mailbox->buf, 0, sizeof(struct mlx4_counter));
+	if_stat_in_mod = counter_index;
+	if (reset)
+		if_stat_in_mod |= MLX4_QUERY_IF_STAT_RESET;
+	err = mlx4_cmd_box(dev, 0, mailbox->dma,
+			   if_stat_in_mod, 0,
+			   MLX4_CMD_QUERY_IF_STAT,
+			   MLX4_CMD_TIME_CLASS_C,
+			   MLX4_CMD_NATIVE);
+	if (err) {
+		mlx4_dbg(dev, "%s: failed to read statistics for counter index %d\n",
+			 __func__, counter_index);
+		goto if_stat_out;
+	}
+	tmp_counter = (struct mlx4_counter *)mailbox->buf;
+	counter_stats->counter_mode = tmp_counter->counter_mode;
+	if (counter_stats->counter_mode == 0) {
+		counter_stats->rx_frames =
+			cpu_to_be64(be64_to_cpu(counter_stats->rx_frames) +
+				    be64_to_cpu(tmp_counter->rx_frames));
+		counter_stats->tx_frames =
+			cpu_to_be64(be64_to_cpu(counter_stats->tx_frames) +
+				    be64_to_cpu(tmp_counter->tx_frames));
+		counter_stats->rx_bytes =
+			cpu_to_be64(be64_to_cpu(counter_stats->rx_bytes) +
+				    be64_to_cpu(tmp_counter->rx_bytes));
+		counter_stats->tx_bytes =
+			cpu_to_be64(be64_to_cpu(counter_stats->tx_bytes) +
+				    be64_to_cpu(tmp_counter->tx_bytes));
+	}
+
+if_stat_out:
+	mlx4_free_cmd_mailbox(dev, mailbox);
+
+	return err;
+}
+EXPORT_SYMBOL_GPL(mlx4_get_counter_stats);
+
+int mlx4_get_vf_stats(struct mlx4_dev *dev, int port, int vf_idx,
+		      struct ifla_vf_stats *vf_stats)
+{
+	struct mlx4_counter tmp_vf_stats;
+	int slave;
+	int err = 0;
+
+	if (!vf_stats)
+		return -EINVAL;
+
+	if (!mlx4_is_master(dev))
+		return -EPROTONOSUPPORT;
+
+	slave = mlx4_get_slave_indx(dev, vf_idx);
+	if (slave < 0)
+		return -EINVAL;
+
+	port = mlx4_slaves_closest_port(dev, slave, port);
+	err = mlx4_calc_vf_counters(dev, slave, port, &tmp_vf_stats);
+	if (!err && tmp_vf_stats.counter_mode == 0) {
+		vf_stats->rx_packets = be64_to_cpu(tmp_vf_stats.rx_frames);
+		vf_stats->tx_packets = be64_to_cpu(tmp_vf_stats.tx_frames);
+		vf_stats->rx_bytes = be64_to_cpu(tmp_vf_stats.rx_bytes);
+		vf_stats->tx_bytes = be64_to_cpu(tmp_vf_stats.tx_bytes);
+	}
+
+	return err;
+}
+EXPORT_SYMBOL_GPL(mlx4_get_vf_stats);
+
 int mlx4_vf_smi_enabled(struct mlx4_dev *dev, int slave, int port)
 {
 	struct mlx4_priv *priv = mlx4_priv(dev);

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

@@ -119,6 +119,12 @@ static const char main_strings[][ETH_GSTRING_LEN] = {
 	"queue_stopped", "wake_queue", "tx_timeout", "rx_alloc_failed",
 	"rx_csum_good", "rx_csum_none", "rx_csum_complete", "tx_chksum_offload",
 
+	/* pf statistics */
+	"pf_rx_packets",
+	"pf_rx_bytes",
+	"pf_tx_packets",
+	"pf_tx_bytes",
+
 	/* priority flow control statistics rx */
 	"rx_pause_prio_0", "rx_pause_duration_prio_0",
 	"rx_pause_transition_prio_0",
@@ -368,6 +374,11 @@ static void mlx4_en_get_ethtool_stats(struct net_device *dev,
 		if (bitmap_iterator_test(&it))
 			data[index++] = ((unsigned long *)&priv->port_stats)[i];
 
+	for (i = 0; i < NUM_PF_STATS; i++, bitmap_iterator_inc(&it))
+		if (bitmap_iterator_test(&it))
+			data[index++] =
+				((unsigned long *)&priv->pf_stats)[i];
+
 	for (i = 0; i < NUM_FLOW_PRIORITY_STATS_RX;
 	     i++, bitmap_iterator_inc(&it))
 		if (bitmap_iterator_test(&it))
@@ -448,6 +459,12 @@ static void mlx4_en_get_strings(struct net_device *dev,
 				strcpy(data + (index++) * ETH_GSTRING_LEN,
 				       main_strings[strings]);
 
+		for (i = 0; i < NUM_PF_STATS; i++, strings++,
+		     bitmap_iterator_inc(&it))
+			if (bitmap_iterator_test(&it))
+				strcpy(data + (index++) * ETH_GSTRING_LEN,
+				       main_strings[strings]);
+
 		for (i = 0; i < NUM_FLOW_STATS; i++, strings++,
 		     bitmap_iterator_inc(&it))
 			if (bitmap_iterator_test(&it))

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

@@ -1597,6 +1597,9 @@ int mlx4_en_start_port(struct net_device *dev)
 	}
 	mdev->mac_removed[priv->port] = 0;
 
+	priv->counter_index =
+			mlx4_get_default_counter_index(mdev->dev, priv->port);
+
 	err = mlx4_en_config_rss_steer(priv);
 	if (err) {
 		en_err(priv, "Failed configuring rss steering\n");
@@ -1755,6 +1758,7 @@ void mlx4_en_stop_port(struct net_device *dev, int detach)
 
 	/* Set port as not active */
 	priv->port_up = false;
+	priv->counter_index = MLX4_SINK_COUNTER_INDEX(mdev->dev);
 
 	/* Promsicuous mode */
 	if (mdev->dev->caps.steering_mode ==
@@ -1891,6 +1895,7 @@ static void mlx4_en_clear_stats(struct net_device *dev)
 	       sizeof(priv->rx_priority_flowstats));
 	memset(&priv->tx_priority_flowstats, 0,
 	       sizeof(priv->tx_priority_flowstats));
+	memset(&priv->pf_stats, 0, sizeof(priv->pf_stats));
 
 	for (i = 0; i < priv->tx_ring_num; i++) {
 		priv->tx_ring[i]->bytes = 0;
@@ -2287,6 +2292,15 @@ static int mlx4_en_set_vf_link_state(struct net_device *dev, int vf, int link_st
 	return mlx4_set_vf_link_state(mdev->dev, en_priv->port, vf, link_state);
 }
 
+static int mlx4_en_get_vf_stats(struct net_device *dev, int vf,
+				struct ifla_vf_stats *vf_stats)
+{
+	struct mlx4_en_priv *en_priv = netdev_priv(dev);
+	struct mlx4_en_dev *mdev = en_priv->mdev;
+
+	return mlx4_get_vf_stats(mdev->dev, en_priv->port, vf, vf_stats);
+}
+
 #define PORT_ID_BYTE_LEN 8
 static int mlx4_en_get_phys_port_id(struct net_device *dev,
 				    struct netdev_phys_item_id *ppid)
@@ -2484,6 +2498,7 @@ static const struct net_device_ops mlx4_netdev_ops_master = {
 	.ndo_set_vf_rate	= mlx4_en_set_vf_rate,
 	.ndo_set_vf_spoofchk	= mlx4_en_set_vf_spoofchk,
 	.ndo_set_vf_link_state	= mlx4_en_set_vf_link_state,
+	.ndo_get_vf_stats       = mlx4_en_get_vf_stats,
 	.ndo_get_vf_config	= mlx4_en_get_vf_config,
 #ifdef CONFIG_NET_POLL_CONTROLLER
 	.ndo_poll_controller	= mlx4_en_netpoll,
@@ -2681,7 +2696,7 @@ void mlx4_en_update_pfc_stats_bitmap(struct mlx4_dev *dev,
 				     u8 rx_ppp, u8 rx_pause,
 				     u8 tx_ppp, u8 tx_pause)
 {
-	int last_i = NUM_MAIN_STATS + NUM_PORT_STATS;
+	int last_i = NUM_MAIN_STATS + NUM_PORT_STATS + NUM_PF_STATS;
 
 	if (!mlx4_is_slave(dev) &&
 	    (dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_FLOWSTATS_EN)) {
@@ -2743,6 +2758,11 @@ void mlx4_en_set_stats_bitmap(struct mlx4_dev *dev,
 	bitmap_set(stats_bitmap->bitmap, last_i, NUM_PORT_STATS);
 	last_i += NUM_PORT_STATS;
 
+	if (mlx4_is_master(dev))
+		bitmap_set(stats_bitmap->bitmap, last_i,
+			   NUM_PF_STATS);
+	last_i += NUM_PF_STATS;
+
 	mlx4_en_update_pfc_stats_bitmap(dev, stats_bitmap,
 					rx_ppp, rx_pause,
 					tx_ppp, tx_pause);
@@ -2778,6 +2798,7 @@ int mlx4_en_init_netdev(struct mlx4_en_dev *mdev, int port,
 
 	priv = netdev_priv(dev);
 	memset(priv, 0, sizeof(struct mlx4_en_priv));
+	priv->counter_index = MLX4_SINK_COUNTER_INDEX(mdev->dev);
 	spin_lock_init(&priv->stats_lock);
 	INIT_WORK(&priv->rx_mode_task, mlx4_en_do_set_rx_mode);
 	INIT_WORK(&priv->watchdog_task, mlx4_en_restart);

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

@@ -149,6 +149,7 @@ static unsigned long en_stats_adder(__be64 *start, __be64 *next, int num)
 
 int mlx4_en_DUMP_ETH_STATS(struct mlx4_en_dev *mdev, u8 port, u8 reset)
 {
+	struct mlx4_counter tmp_counter_stats;
 	struct mlx4_en_stat_out_mbox *mlx4_en_stats;
 	struct mlx4_en_stat_out_flow_control_mbox *flowstats;
 	struct mlx4_en_priv *priv = netdev_priv(mdev->pndev[port]);
@@ -156,7 +157,7 @@ int mlx4_en_DUMP_ETH_STATS(struct mlx4_en_dev *mdev, u8 port, u8 reset)
 	struct mlx4_cmd_mailbox *mailbox;
 	u64 in_mod = reset << 8 | port;
 	int err;
-	int i;
+	int i, counter_index;
 
 	mailbox = mlx4_alloc_cmd_mailbox(mdev->dev);
 	if (IS_ERR(mailbox))
@@ -296,6 +297,11 @@ int mlx4_en_DUMP_ETH_STATS(struct mlx4_en_dev *mdev, u8 port, u8 reset)
 
 	spin_unlock_bh(&priv->stats_lock);
 
+	memset(&tmp_counter_stats, 0, sizeof(tmp_counter_stats));
+	counter_index = mlx4_get_default_counter_index(mdev->dev, port);
+	err = mlx4_get_counter_stats(mdev->dev, counter_index,
+				     &tmp_counter_stats, reset);
+
 	/* 0xffs indicates invalid value */
 	memset(mailbox->buf, 0xff, sizeof(*flowstats) * MLX4_NUM_PRIORITIES);
 
@@ -314,6 +320,13 @@ int mlx4_en_DUMP_ETH_STATS(struct mlx4_en_dev *mdev, u8 port, u8 reset)
 
 	spin_lock_bh(&priv->stats_lock);
 
+	if (tmp_counter_stats.counter_mode == 0) {
+		priv->pf_stats.rx_bytes   = be64_to_cpu(tmp_counter_stats.rx_bytes);
+		priv->pf_stats.tx_bytes   = be64_to_cpu(tmp_counter_stats.tx_bytes);
+		priv->pf_stats.rx_packets = be64_to_cpu(tmp_counter_stats.rx_frames);
+		priv->pf_stats.tx_packets = be64_to_cpu(tmp_counter_stats.tx_frames);
+	}
+
 	for (i = 0; i < MLX4_NUM_PRIORITIES; i++)	{
 		priv->rx_priority_flowstats[i].rx_pause =
 			be64_to_cpu(flowstats[i].rx_pause);

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

@@ -66,7 +66,7 @@ void mlx4_en_fill_qp_context(struct mlx4_en_priv *priv, int size, int stride,
 		context->pri_path.sched_queue |= user_prio << 3;
 		context->pri_path.feup = MLX4_FEUP_FORCE_ETH_UP;
 	}
-	context->pri_path.counter_index = 0xff;
+	context->pri_path.counter_index = priv->counter_index;
 	context->cqn_send = cpu_to_be32(cqn);
 	context->cqn_recv = cpu_to_be32(cqn);
 	context->db_rec_addr = cpu_to_be64(priv->res.db.dma << 2);

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

@@ -479,7 +479,7 @@ static int mlx4_dev_cap(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
 		}
 	}
 
-	dev->caps.max_counters = 1 << ilog2(dev_cap->max_counters);
+	dev->caps.max_counters = dev_cap->max_counters;
 
 	dev->caps.reserved_qps_cnt[MLX4_QP_REGION_FW] = dev_cap->reserved_qps;
 	dev->caps.reserved_qps_cnt[MLX4_QP_REGION_ETH_ADDR] =
@@ -2193,20 +2193,73 @@ err_free_icm:
 static int mlx4_init_counters_table(struct mlx4_dev *dev)
 {
 	struct mlx4_priv *priv = mlx4_priv(dev);
-	int nent;
+	int nent_pow2;
 
 	if (!(dev->caps.flags & MLX4_DEV_CAP_FLAG_COUNTERS))
 		return -ENOENT;
 
-	nent = dev->caps.max_counters;
-	return mlx4_bitmap_init(&priv->counters_bitmap, nent, nent - 1, 0, 0);
+	if (!dev->caps.max_counters)
+		return -ENOSPC;
+
+	nent_pow2 = roundup_pow_of_two(dev->caps.max_counters);
+	/* reserve last counter index for sink counter */
+	return mlx4_bitmap_init(&priv->counters_bitmap, nent_pow2,
+				nent_pow2 - 1, 0,
+				nent_pow2 - dev->caps.max_counters + 1);
 }
 
 static void mlx4_cleanup_counters_table(struct mlx4_dev *dev)
 {
+	if (!(dev->caps.flags & MLX4_DEV_CAP_FLAG_COUNTERS))
+		return;
+
+	if (!dev->caps.max_counters)
+		return;
+
 	mlx4_bitmap_cleanup(&mlx4_priv(dev)->counters_bitmap);
 }
 
+static void mlx4_cleanup_default_counters(struct mlx4_dev *dev)
+{
+	struct mlx4_priv *priv = mlx4_priv(dev);
+	int port;
+
+	for (port = 0; port < dev->caps.num_ports; port++)
+		if (priv->def_counter[port] != -1)
+			mlx4_counter_free(dev,  priv->def_counter[port]);
+}
+
+static int mlx4_allocate_default_counters(struct mlx4_dev *dev)
+{
+	struct mlx4_priv *priv = mlx4_priv(dev);
+	int port, err = 0;
+	u32 idx;
+
+	for (port = 0; port < dev->caps.num_ports; port++)
+		priv->def_counter[port] = -1;
+
+	for (port = 0; port < dev->caps.num_ports; port++) {
+		err = mlx4_counter_alloc(dev, &idx);
+
+		if (!err || err == -ENOSPC) {
+			priv->def_counter[port] = idx;
+		} else if (err == -ENOENT) {
+			err = 0;
+			continue;
+		} else {
+			mlx4_err(dev, "%s: failed to allocate default counter port %d err %d\n",
+				 __func__, port + 1, err);
+			mlx4_cleanup_default_counters(dev);
+			return err;
+		}
+
+		mlx4_dbg(dev, "%s: default counter index %d for port %d\n",
+			 __func__, priv->def_counter[port], port + 1);
+	}
+
+	return err;
+}
+
 int __mlx4_counter_alloc(struct mlx4_dev *dev, u32 *idx)
 {
 	struct mlx4_priv *priv = mlx4_priv(dev);
@@ -2215,8 +2268,10 @@ int __mlx4_counter_alloc(struct mlx4_dev *dev, u32 *idx)
 		return -ENOENT;
 
 	*idx = mlx4_bitmap_alloc(&priv->counters_bitmap);
-	if (*idx == -1)
-		return -ENOMEM;
+	if (*idx == -1) {
+		*idx = MLX4_SINK_COUNTER_INDEX(dev);
+		return -ENOSPC;
+	}
 
 	return 0;
 }
@@ -2239,8 +2294,35 @@ int mlx4_counter_alloc(struct mlx4_dev *dev, u32 *idx)
 }
 EXPORT_SYMBOL_GPL(mlx4_counter_alloc);
 
+static int __mlx4_clear_if_stat(struct mlx4_dev *dev,
+				u8 counter_index)
+{
+	struct mlx4_cmd_mailbox *if_stat_mailbox;
+	int err;
+	u32 if_stat_in_mod = (counter_index & 0xff) | MLX4_QUERY_IF_STAT_RESET;
+
+	if_stat_mailbox = mlx4_alloc_cmd_mailbox(dev);
+	if (IS_ERR(if_stat_mailbox))
+		return PTR_ERR(if_stat_mailbox);
+
+	err = mlx4_cmd_box(dev, 0, if_stat_mailbox->dma, if_stat_in_mod, 0,
+			   MLX4_CMD_QUERY_IF_STAT, MLX4_CMD_TIME_CLASS_C,
+			   MLX4_CMD_NATIVE);
+
+	mlx4_free_cmd_mailbox(dev, if_stat_mailbox);
+	return err;
+}
+
 void __mlx4_counter_free(struct mlx4_dev *dev, u32 idx)
 {
+	if (!(dev->caps.flags & MLX4_DEV_CAP_FLAG_COUNTERS))
+		return;
+
+	if (idx == MLX4_SINK_COUNTER_INDEX(dev))
+		return;
+
+	__mlx4_clear_if_stat(dev, idx);
+
 	mlx4_bitmap_free(&mlx4_priv(dev)->counters_bitmap, idx, MLX4_USE_RR);
 	return;
 }
@@ -2260,6 +2342,14 @@ void mlx4_counter_free(struct mlx4_dev *dev, u32 idx)
 }
 EXPORT_SYMBOL_GPL(mlx4_counter_free);
 
+int mlx4_get_default_counter_index(struct mlx4_dev *dev, int port)
+{
+	struct mlx4_priv *priv = mlx4_priv(dev);
+
+	return priv->def_counter[port - 1];
+}
+EXPORT_SYMBOL_GPL(mlx4_get_default_counter_index);
+
 void mlx4_set_admin_guid(struct mlx4_dev *dev, __be64 guid, int entry, int port)
 {
 	struct mlx4_priv *priv = mlx4_priv(dev);
@@ -2395,10 +2485,18 @@ static int mlx4_setup_hca(struct mlx4_dev *dev)
 		goto err_srq_table_free;
 	}
 
-	err = mlx4_init_counters_table(dev);
-	if (err && err != -ENOENT) {
-		mlx4_err(dev, "Failed to initialize counters table, aborting\n");
-		goto err_qp_table_free;
+	if (!mlx4_is_slave(dev)) {
+		err = mlx4_init_counters_table(dev);
+		if (err && err != -ENOENT) {
+			mlx4_err(dev, "Failed to initialize counters table, aborting\n");
+			goto err_qp_table_free;
+		}
+	}
+
+	err = mlx4_allocate_default_counters(dev);
+	if (err) {
+		mlx4_err(dev, "Failed to allocate default counters, aborting\n");
+		goto err_counters_table_free;
 	}
 
 	if (!mlx4_is_slave(dev)) {
@@ -2432,15 +2530,19 @@ static int mlx4_setup_hca(struct mlx4_dev *dev)
 			if (err) {
 				mlx4_err(dev, "Failed to set port %d, aborting\n",
 					 port);
-				goto err_counters_table_free;
+				goto err_default_countes_free;
 			}
 		}
 	}
 
 	return 0;
 
+err_default_countes_free:
+	mlx4_cleanup_default_counters(dev);
+
 err_counters_table_free:
-	mlx4_cleanup_counters_table(dev);
+	if (!mlx4_is_slave(dev))
+		mlx4_cleanup_counters_table(dev);
 
 err_qp_table_free:
 	mlx4_cleanup_qp_table(dev);
@@ -3173,7 +3275,9 @@ err_port:
 	for (--port; port >= 1; --port)
 		mlx4_cleanup_port_info(&priv->port[port]);
 
-	mlx4_cleanup_counters_table(dev);
+	mlx4_cleanup_default_counters(dev);
+	if (!mlx4_is_slave(dev))
+		mlx4_cleanup_counters_table(dev);
 	mlx4_cleanup_qp_table(dev);
 	mlx4_cleanup_srq_table(dev);
 	mlx4_cleanup_cq_table(dev);
@@ -3471,7 +3575,9 @@ static void mlx4_unload_one(struct pci_dev *pdev)
 		mlx4_free_resource_tracker(dev,
 					   RES_TR_FREE_SLAVES_ONLY);
 
-	mlx4_cleanup_counters_table(dev);
+	mlx4_cleanup_default_counters(dev);
+	if (!mlx4_is_slave(dev))
+		mlx4_cleanup_counters_table(dev);
 	mlx4_cleanup_qp_table(dev);
 	mlx4_cleanup_srq_table(dev);
 	mlx4_cleanup_cq_table(dev);

drivers/net/ethernet/mellanox/mlx4/mlx4.h

@@ -65,6 +65,8 @@
 
 #define INIT_HCA_TPT_MW_ENABLE          (1 << 7)
 
+#define MLX4_QUERY_IF_STAT_RESET	BIT(31)
+
 enum {
 	MLX4_HCR_BASE		= 0x80680,
 	MLX4_HCR_SIZE		= 0x0001c,
@@ -874,6 +876,7 @@ struct mlx4_priv {
 	struct mlx4_qp_table	qp_table;
 	struct mlx4_mcg_table	mcg_table;
 	struct mlx4_bitmap	counters_bitmap;
+	int			def_counter[MLX4_MAX_PORTS];
 
 	struct mlx4_catas_err	catas_err;
 
@@ -1007,6 +1010,8 @@ int __mlx4_write_mtt(struct mlx4_dev *dev, struct mlx4_mtt *mtt,
 		     int start_index, int npages, u64 *page_list);
 int __mlx4_counter_alloc(struct mlx4_dev *dev, u32 *idx);
 void __mlx4_counter_free(struct mlx4_dev *dev, u32 idx);
+int mlx4_calc_vf_counters(struct mlx4_dev *dev, int slave, int port,
+			  struct mlx4_counter *data);
 int __mlx4_xrcd_alloc(struct mlx4_dev *dev, u32 *xrcdn);
 void __mlx4_xrcd_free(struct mlx4_dev *dev, u32 xrcdn);
 

drivers/net/ethernet/mellanox/mlx4/mlx4_en.h

@@ -566,6 +566,7 @@ struct mlx4_en_priv {
 #endif
 	struct mlx4_en_perf_stats pstats;
 	struct mlx4_en_pkt_stats pkstats;
+	struct mlx4_en_counter_stats pf_stats;
 	struct mlx4_en_flow_stats_rx rx_priority_flowstats[MLX4_NUM_PRIORITIES];
 	struct mlx4_en_flow_stats_tx tx_priority_flowstats[MLX4_NUM_PRIORITIES];
 	struct mlx4_en_flow_stats_rx rx_flowstats;
@@ -582,6 +583,7 @@ struct mlx4_en_priv {
 	int base_tx_qpn;
 	struct hlist_head mac_hash[MLX4_EN_MAC_HASH_SIZE];
 	struct hwtstamp_config hwtstamp_config;
+	u32 counter_index;
 
 #ifdef CONFIG_MLX4_EN_DCB
 	struct ieee_ets ets;

drivers/net/ethernet/mellanox/mlx4/mlx4_stats.h

@@ -23,6 +23,14 @@ struct mlx4_en_pkt_stats {
 #define NUM_PKT_STATS		43
 };
 
+struct mlx4_en_counter_stats {
+	unsigned long rx_packets;
+	unsigned long rx_bytes;
+	unsigned long tx_packets;
+	unsigned long tx_bytes;
+#define NUM_PF_STATS      4
+};
+
 struct mlx4_en_port_stats {
 	unsigned long tso_packets;
 	unsigned long xmit_more;
@@ -71,7 +79,8 @@ struct mlx4_en_flow_stats_tx {
 
 #define NUM_FLOW_STATS (NUM_FLOW_STATS_RX + NUM_FLOW_STATS_TX + \
 			NUM_FLOW_PRIORITY_STATS_TX + \
-			NUM_FLOW_PRIORITY_STATS_RX)
+			NUM_FLOW_PRIORITY_STATS_RX + \
+			NUM_PF_STATS)
 
 struct mlx4_en_stat_out_flow_control_mbox {
 	/* Total number of PAUSE frames received from the far-end port */

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

@@ -46,8 +46,11 @@
 
 #include "mlx4.h"
 #include "fw.h"
+#include "mlx4_stats.h"
 
 #define MLX4_MAC_VALID		(1ull << 63)
+#define MLX4_PF_COUNTERS_PER_PORT	2
+#define MLX4_VF_COUNTERS_PER_PORT	1
 
 struct mac_res {
 	struct list_head list;
@@ -459,11 +462,21 @@ void mlx4_init_quotas(struct mlx4_dev *dev)
 	dev->quotas.mpt =
 		priv->mfunc.master.res_tracker.res_alloc[RES_MPT].quota[pf];
 }
+
+static int get_max_gauranteed_vfs_counter(struct mlx4_dev *dev)
+{
+	/* reduce the sink counter */
+	return (dev->caps.max_counters - 1 -
+		(MLX4_PF_COUNTERS_PER_PORT * MLX4_MAX_PORTS))
+		/ MLX4_MAX_PORTS;
+}
+
 int mlx4_init_resource_tracker(struct mlx4_dev *dev)
 {
 	struct mlx4_priv *priv = mlx4_priv(dev);
 	int i, j;
 	int t;
+	int max_vfs_guarantee_counter = get_max_gauranteed_vfs_counter(dev);
 
 	priv->mfunc.master.res_tracker.slave_list =
 		kzalloc(dev->num_slaves * sizeof(struct slave_list),
@@ -499,6 +512,9 @@ int mlx4_init_resource_tracker(struct mlx4_dev *dev)
 			res_alloc->allocated = kzalloc((dev->persist->
 							num_vfs + 1) *
 						       sizeof(int), GFP_KERNEL);
+		/* Reduce the sink counter */
+		if (i == RES_COUNTER)
+			res_alloc->res_free = dev->caps.max_counters - 1;
 
 		if (!res_alloc->quota || !res_alloc->guaranteed ||
 		    !res_alloc->allocated)
@@ -577,9 +593,17 @@ int mlx4_init_resource_tracker(struct mlx4_dev *dev)
 				break;
 			case RES_COUNTER:
 				res_alloc->quota[t] = dev->caps.max_counters;
-				res_alloc->guaranteed[t] = 0;
 				if (t == mlx4_master_func_num(dev))
-					res_alloc->res_free = res_alloc->quota[t];
+					res_alloc->guaranteed[t] =
+						MLX4_PF_COUNTERS_PER_PORT *
+						MLX4_MAX_PORTS;
+				else if (t <= max_vfs_guarantee_counter)
+					res_alloc->guaranteed[t] =
+						MLX4_VF_COUNTERS_PER_PORT *
+						MLX4_MAX_PORTS;
+				else
+					res_alloc->guaranteed[t] = 0;
+				res_alloc->res_free -= res_alloc->guaranteed[t];
 				break;
 			default:
 				break;
@@ -700,6 +724,9 @@ static void update_gid(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *inbox,
 	}
 }
 
+static int handle_counter(struct mlx4_dev *dev, struct mlx4_qp_context *qpc,
+			  u8 slave, int port);
+
 static int update_vport_qp_param(struct mlx4_dev *dev,
 				 struct mlx4_cmd_mailbox *inbox,
 				 u8 slave, u32 qpn)
@@ -715,6 +742,10 @@ static int update_vport_qp_param(struct mlx4_dev *dev,
 	vp_oper = &priv->mfunc.master.vf_oper[slave].vport[port];
 	qp_type	= (be32_to_cpu(qpc->flags) >> 16) & 0xff;
 
+	err = handle_counter(dev, qpc, slave, port);
+	if (err)
+		goto out;
+
 	if (MLX4_VGT != vp_oper->state.default_vlan) {
 		/* the reserved QPs (special, proxy, tunnel)
 		 * do not operate over vlans
@@ -859,6 +890,83 @@ static void put_res(struct mlx4_dev *dev, int slave, u64 res_id,
 	spin_unlock_irq(mlx4_tlock(dev));
 }
 
+static int counter_alloc_res(struct mlx4_dev *dev, int slave, int op, int cmd,
+			     u64 in_param, u64 *out_param, int port);
+
+static int handle_existing_counter(struct mlx4_dev *dev, u8 slave, int port,
+				   int counter_index)
+{
+	struct res_common *r;
+	struct res_counter *counter;
+	int ret = 0;
+
+	if (counter_index == MLX4_SINK_COUNTER_INDEX(dev))
+		return ret;
+
+	spin_lock_irq(mlx4_tlock(dev));
+	r = find_res(dev, counter_index, RES_COUNTER);
+	if (!r || r->owner != slave)
+		ret = -EINVAL;
+	counter = container_of(r, struct res_counter, com);
+	if (!counter->port)
+		counter->port = port;
+
+	spin_unlock_irq(mlx4_tlock(dev));
+	return ret;
+}
+
+static int handle_unexisting_counter(struct mlx4_dev *dev,
+				     struct mlx4_qp_context *qpc, u8 slave,
+				     int port)
+{
+	struct mlx4_priv *priv = mlx4_priv(dev);
+	struct mlx4_resource_tracker *tracker = &priv->mfunc.master.res_tracker;
+	struct res_common *tmp;
+	struct res_counter *counter;
+	u64 counter_idx = MLX4_SINK_COUNTER_INDEX(dev);
+	int err = 0;
+
+	spin_lock_irq(mlx4_tlock(dev));
+	list_for_each_entry(tmp,
+			    &tracker->slave_list[slave].res_list[RES_COUNTER],
+			    list) {
+		counter = container_of(tmp, struct res_counter, com);
+		if (port == counter->port) {
+			qpc->pri_path.counter_index  = counter->com.res_id;
+			spin_unlock_irq(mlx4_tlock(dev));
+			return 0;
+		}
+	}
+	spin_unlock_irq(mlx4_tlock(dev));
+
+	/* No existing counter, need to allocate a new counter */
+	err = counter_alloc_res(dev, slave, RES_OP_RESERVE, 0, 0, &counter_idx,
+				port);
+	if (err == -ENOENT) {
+		err = 0;
+	} else if (err && err != -ENOSPC) {
+		mlx4_err(dev, "%s: failed to create new counter for slave %d err %d\n",
+			 __func__, slave, err);
+	} else {
+		qpc->pri_path.counter_index = counter_idx;
+		mlx4_dbg(dev, "%s: alloc new counter for slave %d index %d\n",
+			 __func__, slave, qpc->pri_path.counter_index);
+		err = 0;
+	}
+
+	return err;
+}
+
+static int handle_counter(struct mlx4_dev *dev, struct mlx4_qp_context *qpc,
+			  u8 slave, int port)
+{
+	if (qpc->pri_path.counter_index != MLX4_SINK_COUNTER_INDEX(dev))
+		return handle_existing_counter(dev, slave, port,
+					       qpc->pri_path.counter_index);
+
+	return handle_unexisting_counter(dev, qpc, slave, port);
+}
+
 static struct res_common *alloc_qp_tr(int id)
 {
 	struct res_qp *ret;
@@ -952,7 +1060,7 @@ static struct res_common *alloc_srq_tr(int id)
 	return &ret->com;
 }
 
-static struct res_common *alloc_counter_tr(int id)
+static struct res_common *alloc_counter_tr(int id, int port)
 {
 	struct res_counter *ret;
 
@@ -962,6 +1070,7 @@ static struct res_common *alloc_counter_tr(int id)
 
 	ret->com.res_id = id;
 	ret->com.state = RES_COUNTER_ALLOCATED;
+	ret->port = port;
 
 	return &ret->com;
 }
@@ -1022,7 +1131,7 @@ static struct res_common *alloc_tr(u64 id, enum mlx4_resource type, int slave,
 		pr_err("implementation missing\n");
 		return NULL;
 	case RES_COUNTER:
-		ret = alloc_counter_tr(id);
+		ret = alloc_counter_tr(id, extra);
 		break;
 	case RES_XRCD:
 		ret = alloc_xrcdn_tr(id);
@@ -1039,6 +1148,53 @@ static struct res_common *alloc_tr(u64 id, enum mlx4_resource type, int slave,
 	return ret;
 }
 
+int mlx4_calc_vf_counters(struct mlx4_dev *dev, int slave, int port,
+			  struct mlx4_counter *data)
+{
+	struct mlx4_priv *priv = mlx4_priv(dev);
+	struct mlx4_resource_tracker *tracker = &priv->mfunc.master.res_tracker;
+	struct res_common *tmp;
+	struct res_counter *counter;
+	int *counters_arr;
+	int i = 0, err = 0;
+
+	memset(data, 0, sizeof(*data));
+
+	counters_arr = kmalloc_array(dev->caps.max_counters,
+				     sizeof(*counters_arr), GFP_KERNEL);
+	if (!counters_arr)
+		return -ENOMEM;
+
+	spin_lock_irq(mlx4_tlock(dev));
+	list_for_each_entry(tmp,
+			    &tracker->slave_list[slave].res_list[RES_COUNTER],
+			    list) {
+		counter = container_of(tmp, struct res_counter, com);
+		if (counter->port == port) {
+			counters_arr[i] = (int)tmp->res_id;
+			i++;
+		}
+	}
+	spin_unlock_irq(mlx4_tlock(dev));
+	counters_arr[i] = -1;
+
+	i = 0;
+
+	while (counters_arr[i] != -1) {
+		err = mlx4_get_counter_stats(dev, counters_arr[i], data,
+					     0);
+		if (err) {
+			memset(data, 0, sizeof(*data));
+			goto table_changed;
+		}
+		i++;
+	}
+
+table_changed:
+	kfree(counters_arr);
+	return 0;
+}
+
 static int add_res_range(struct mlx4_dev *dev, int slave, u64 base, int count,
 			 enum mlx4_resource type, int extra)
 {
@@ -2001,7 +2157,7 @@ static int vlan_alloc_res(struct mlx4_dev *dev, int slave, int op, int cmd,
 }
 
 static int counter_alloc_res(struct mlx4_dev *dev, int slave, int op, int cmd,
-			     u64 in_param, u64 *out_param)
+			     u64 in_param, u64 *out_param, int port)
 {
 	u32 index;
 	int err;
@@ -2019,7 +2175,7 @@ static int counter_alloc_res(struct mlx4_dev *dev, int slave, int op, int cmd,
 		return err;
 	}
 
-	err = add_res_range(dev, slave, index, 1, RES_COUNTER, 0);
+	err = add_res_range(dev, slave, index, 1, RES_COUNTER, port);
 	if (err) {
 		__mlx4_counter_free(dev, index);
 		mlx4_release_resource(dev, slave, RES_COUNTER, 1, 0);
@@ -2101,7 +2257,7 @@ int mlx4_ALLOC_RES_wrapper(struct mlx4_dev *dev, int slave,
 
 	case RES_COUNTER:
 		err = counter_alloc_res(dev, slave, vhcr->op_modifier, alop,
-					vhcr->in_param, &vhcr->out_param);
+					vhcr->in_param, &vhcr->out_param, 0);
 		break;
 
 	case RES_XRCD:
@@ -2335,6 +2491,9 @@ static int counter_free_res(struct mlx4_dev *dev, int slave, int op, int cmd,
 		return -EINVAL;
 
 	index = get_param_l(&in_param);
+	if (index == MLX4_SINK_COUNTER_INDEX(dev))
+		return 0;
+
 	err = rem_res_range(dev, slave, index, 1, RES_COUNTER, 0);
 	if (err)
 		return err;

include/linux/if_link.h

@@ -5,6 +5,15 @@
 
 
 /* We don't want this structure exposed to user space */
+struct ifla_vf_stats {
+	__u64 rx_packets;
+	__u64 tx_packets;
+	__u64 rx_bytes;
+	__u64 tx_bytes;
+	__u64 broadcast;
+	__u64 multicast;
+};
+
 struct ifla_vf_info {
 	__u32 vf;
 	__u8 mac[32];

include/linux/mlx4/cmd.h

@@ -35,6 +35,8 @@
 
 #include <linux/dma-mapping.h>
 #include <linux/if_link.h>
+#include <linux/mlx4/device.h>
+#include <linux/netdevice.h>
 
 enum {
 	/* initialization and general commands */
@@ -300,6 +302,10 @@ static inline int mlx4_cmd_imm(struct mlx4_dev *dev, u64 in_param, u64 *out_para
 struct mlx4_cmd_mailbox *mlx4_alloc_cmd_mailbox(struct mlx4_dev *dev);
 void mlx4_free_cmd_mailbox(struct mlx4_dev *dev, struct mlx4_cmd_mailbox *mailbox);
 
+int mlx4_get_counter_stats(struct mlx4_dev *dev, int counter_index,
+			   struct mlx4_counter *counter_stats, int reset);
+int mlx4_get_vf_stats(struct mlx4_dev *dev, int port, int vf_idx,
+		      struct ifla_vf_stats *vf_stats);
 u32 mlx4_comm_get_version(void);
 int mlx4_set_vf_mac(struct mlx4_dev *dev, int port, int vf, u64 mac);
 int mlx4_set_vf_vlan(struct mlx4_dev *dev, int port, int vf, u16 vlan, u8 qos);

include/linux/mlx4/device.h

@@ -771,6 +771,14 @@ union mlx4_ext_av {
 	struct mlx4_eth_av	eth;
 };
 
+/* Counters should be saturate once they reach their maximum value */
+#define ASSIGN_32BIT_COUNTER(counter, value) do {	\
+	if ((value) > U32_MAX)				\
+		counter = cpu_to_be32(U32_MAX);		\
+	else						\
+		counter = cpu_to_be32(value);		\
+} while (0)
+
 struct mlx4_counter {
 	u8	reserved1[3];
 	u8	counter_mode;
@@ -957,6 +965,7 @@ struct mlx4_mad_ifc {
 			((dev)->caps.flags & MLX4_DEV_CAP_FLAG_IBOE))
 
 #define MLX4_INVALID_SLAVE_ID	0xFF
+#define MLX4_SINK_COUNTER_INDEX(dev)	(dev->caps.max_counters - 1)
 
 void handle_port_mgmt_change_event(struct work_struct *work);
 
@@ -1347,6 +1356,7 @@ int mlx4_wol_write(struct mlx4_dev *dev, u64 config, int port);
 
 int mlx4_counter_alloc(struct mlx4_dev *dev, u32 *idx);
 void mlx4_counter_free(struct mlx4_dev *dev, u32 idx);
+int mlx4_get_default_counter_index(struct mlx4_dev *dev, int port);
 
 void mlx4_set_admin_guid(struct mlx4_dev *dev, __be64 guid, int entry,
 			 int port);

include/linux/netdevice.h

@@ -1100,6 +1100,10 @@ struct net_device_ops {
 						     struct ifla_vf_info *ivf);
 	int			(*ndo_set_vf_link_state)(struct net_device *dev,
 							 int vf, int link_state);
+	int			(*ndo_get_vf_stats)(struct net_device *dev,
+						    int vf,
+						    struct ifla_vf_stats
+						    *vf_stats);
 	int			(*ndo_set_vf_port)(struct net_device *dev,
 						   int vf,
 						   struct nlattr *port[]);

include/uapi/linux/if_link.h

@@ -484,6 +484,7 @@ enum {
 	IFLA_VF_RSS_QUERY_EN,	/* RSS Redirection Table and Hash Key query
 				 * on/off switch
 				 */
+	IFLA_VF_STATS,		/* network device statistics */
 	__IFLA_VF_MAX,
 };
 
@@ -533,6 +534,18 @@ struct ifla_vf_rss_query_en {
 	__u32 setting;
 };
 
+enum {
+	IFLA_VF_STATS_RX_PACKETS,
+	IFLA_VF_STATS_TX_PACKETS,
+	IFLA_VF_STATS_RX_BYTES,
+	IFLA_VF_STATS_TX_BYTES,
+	IFLA_VF_STATS_BROADCAST,
+	IFLA_VF_STATS_MULTICAST,
+	__IFLA_VF_STATS_MAX,
+};
+
+#define IFLA_VF_STATS_MAX (__IFLA_VF_STATS_MAX - 1)
+
 /* VF ports management section
  *
  *	Nested layout of set/get msg is:

net/core/rtnetlink.c

@@ -819,7 +819,19 @@ static inline int rtnl_vfinfo_size(const struct net_device *dev,
 			 nla_total_size(sizeof(struct ifla_vf_spoofchk)) +
 			 nla_total_size(sizeof(struct ifla_vf_rate)) +
 			 nla_total_size(sizeof(struct ifla_vf_link_state)) +
-			 nla_total_size(sizeof(struct ifla_vf_rss_query_en)));
+			 nla_total_size(sizeof(struct ifla_vf_rss_query_en)) +
+			 /* IFLA_VF_STATS_RX_PACKETS */
+			 nla_total_size(sizeof(__u64)) +
+			 /* IFLA_VF_STATS_TX_PACKETS */
+			 nla_total_size(sizeof(__u64)) +
+			 /* IFLA_VF_STATS_RX_BYTES */
+			 nla_total_size(sizeof(__u64)) +
+			 /* IFLA_VF_STATS_TX_BYTES */
+			 nla_total_size(sizeof(__u64)) +
+			 /* IFLA_VF_STATS_BROADCAST */
+			 nla_total_size(sizeof(__u64)) +
+			 /* IFLA_VF_STATS_MULTICAST */
+			 nla_total_size(sizeof(__u64)));
 		return size;
 	} else
 		return 0;
@@ -1123,7 +1135,7 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
 	    && (ext_filter_mask & RTEXT_FILTER_VF)) {
 		int i;
 
-		struct nlattr *vfinfo, *vf;
+		struct nlattr *vfinfo, *vf, *vfstats;
 		int num_vfs = dev_num_vf(dev->dev.parent);
 
 		vfinfo = nla_nest_start(skb, IFLA_VFINFO_LIST);
@@ -1138,6 +1150,7 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
 			struct ifla_vf_spoofchk vf_spoofchk;
 			struct ifla_vf_link_state vf_linkstate;
 			struct ifla_vf_rss_query_en vf_rss_query_en;
+			struct ifla_vf_stats vf_stats;
 
 			/*
 			 * Not all SR-IOV capable drivers support the
@@ -1190,6 +1203,30 @@ static int rtnl_fill_ifinfo(struct sk_buff *skb, struct net_device *dev,
 				    sizeof(vf_rss_query_en),
 				    &vf_rss_query_en))
 				goto nla_put_failure;
+			memset(&vf_stats, 0, sizeof(vf_stats));
+			if (dev->netdev_ops->ndo_get_vf_stats)
+				dev->netdev_ops->ndo_get_vf_stats(dev, i,
+								  &vf_stats);
+			vfstats = nla_nest_start(skb, IFLA_VF_STATS);
+			if (!vfstats) {
+				nla_nest_cancel(skb, vf);
+				nla_nest_cancel(skb, vfinfo);
+				goto nla_put_failure;
+			}
+			if (nla_put_u64(skb, IFLA_VF_STATS_RX_PACKETS,
+					vf_stats.rx_packets) ||
+			    nla_put_u64(skb, IFLA_VF_STATS_TX_PACKETS,
+					vf_stats.tx_packets) ||
+			    nla_put_u64(skb, IFLA_VF_STATS_RX_BYTES,
+					vf_stats.rx_bytes) ||
+			    nla_put_u64(skb, IFLA_VF_STATS_TX_BYTES,
+					vf_stats.tx_bytes) ||
+			    nla_put_u64(skb, IFLA_VF_STATS_BROADCAST,
+					vf_stats.broadcast) ||
+			    nla_put_u64(skb, IFLA_VF_STATS_MULTICAST,
+					vf_stats.multicast))
+				goto nla_put_failure;
+			nla_nest_end(skb, vfstats);
 			nla_nest_end(skb, vf);
 		}
 		nla_nest_end(skb, vfinfo);
@@ -1303,6 +1340,16 @@ static const struct nla_policy ifla_vf_policy[IFLA_VF_MAX+1] = {
 	[IFLA_VF_RATE]		= { .len = sizeof(struct ifla_vf_rate) },
 	[IFLA_VF_LINK_STATE]	= { .len = sizeof(struct ifla_vf_link_state) },
 	[IFLA_VF_RSS_QUERY_EN]	= { .len = sizeof(struct ifla_vf_rss_query_en) },
+	[IFLA_VF_STATS]		= { .type = NLA_NESTED },
+};
+
+static const struct nla_policy ifla_vf_stats_policy[IFLA_VF_STATS_MAX + 1] = {
+	[IFLA_VF_STATS_RX_PACKETS]	= { .type = NLA_U64 },
+	[IFLA_VF_STATS_TX_PACKETS]	= { .type = NLA_U64 },
+	[IFLA_VF_STATS_RX_BYTES]	= { .type = NLA_U64 },
+	[IFLA_VF_STATS_TX_BYTES]	= { .type = NLA_U64 },
+	[IFLA_VF_STATS_BROADCAST]	= { .type = NLA_U64 },
+	[IFLA_VF_STATS_MULTICAST]	= { .type = NLA_U64 },
 };
 
 static const struct nla_policy ifla_port_policy[IFLA_PORT_MAX+1] = {