diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h
index dce21daaf330..67b1a2f8397f 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/reg.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h
@@ -6784,12 +6784,14 @@ static inline void mlxsw_reg_ritr_counter_pack(char *payload, u32 index,
 		set_type = MLXSW_REG_RITR_COUNTER_SET_TYPE_BASIC;
 	else
 		set_type = MLXSW_REG_RITR_COUNTER_SET_TYPE_NO_COUNT;
-	mlxsw_reg_ritr_egress_counter_set_type_set(payload, set_type);
 
-	if (egress)
+	if (egress) {
+		mlxsw_reg_ritr_egress_counter_set_type_set(payload, set_type);
 		mlxsw_reg_ritr_egress_counter_index_set(payload, index);
-	else
+	} else {
+		mlxsw_reg_ritr_ingress_counter_set_type_set(payload, set_type);
 		mlxsw_reg_ritr_ingress_counter_index_set(payload, index);
+	}
 }
 
 static inline void mlxsw_reg_ritr_rif_pack(char *payload, u16 rif)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
index 4880521b11a7..7b7b17183d10 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c
@@ -4823,6 +4823,22 @@ static int mlxsw_sp_netdevice_vxlan_event(struct mlxsw_sp *mlxsw_sp,
 	return 0;
 }
 
+static bool mlxsw_sp_netdevice_event_is_router(unsigned long event)
+{
+	switch (event) {
+	case NETDEV_PRE_CHANGEADDR:
+	case NETDEV_CHANGEADDR:
+	case NETDEV_CHANGEMTU:
+	case NETDEV_OFFLOAD_XSTATS_ENABLE:
+	case NETDEV_OFFLOAD_XSTATS_DISABLE:
+	case NETDEV_OFFLOAD_XSTATS_REPORT_USED:
+	case NETDEV_OFFLOAD_XSTATS_REPORT_DELTA:
+		return true;
+	default:
+		return false;
+	}
+}
+
 static int mlxsw_sp_netdevice_event(struct notifier_block *nb,
 				    unsigned long event, void *ptr)
 {
@@ -4847,9 +4863,7 @@ static int mlxsw_sp_netdevice_event(struct notifier_block *nb,
 	else if (mlxsw_sp_netdev_is_ipip_ul(mlxsw_sp, dev))
 		err = mlxsw_sp_netdevice_ipip_ul_event(mlxsw_sp, dev,
 						       event, ptr);
-	else if (event == NETDEV_PRE_CHANGEADDR ||
-		 event == NETDEV_CHANGEADDR ||
-		 event == NETDEV_CHANGEMTU)
+	else if (mlxsw_sp_netdevice_event_is_router(event))
 		err = mlxsw_sp_netdevice_router_port_event(dev, event, ptr);
 	else if (mlxsw_sp_is_vrf_event(event, ptr))
 		err = mlxsw_sp_netdevice_vrf_event(dev, event, ptr);
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.c
index 1a2fef2a5379..5d494fabf93d 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dpipe.c
@@ -266,10 +266,10 @@ static int mlxsw_sp_dpipe_table_erif_counters_update(void *priv, bool enable)
 		if (!rif)
 			continue;
 		if (enable)
-			mlxsw_sp_rif_counter_alloc(mlxsw_sp, rif,
+			mlxsw_sp_rif_counter_alloc(rif,
 						   MLXSW_SP_RIF_COUNTER_EGRESS);
 		else
-			mlxsw_sp_rif_counter_free(mlxsw_sp, rif,
+			mlxsw_sp_rif_counter_free(rif,
 						  MLXSW_SP_RIF_COUNTER_EGRESS);
 	}
 	mutex_unlock(&mlxsw_sp->router->lock);
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
index d40762cfc453..79deb19e3a19 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c
@@ -225,6 +225,64 @@ int mlxsw_sp_rif_counter_value_get(struct mlxsw_sp *mlxsw_sp,
 	return 0;
 }
 
+struct mlxsw_sp_rif_counter_set_basic {
+	u64 good_unicast_packets;
+	u64 good_multicast_packets;
+	u64 good_broadcast_packets;
+	u64 good_unicast_bytes;
+	u64 good_multicast_bytes;
+	u64 good_broadcast_bytes;
+	u64 error_packets;
+	u64 discard_packets;
+	u64 error_bytes;
+	u64 discard_bytes;
+};
+
+static int
+mlxsw_sp_rif_counter_fetch_clear(struct mlxsw_sp_rif *rif,
+				 enum mlxsw_sp_rif_counter_dir dir,
+				 struct mlxsw_sp_rif_counter_set_basic *set)
+{
+	struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp;
+	char ricnt_pl[MLXSW_REG_RICNT_LEN];
+	unsigned int *p_counter_index;
+	int err;
+
+	if (!mlxsw_sp_rif_counter_valid_get(rif, dir))
+		return -EINVAL;
+
+	p_counter_index = mlxsw_sp_rif_p_counter_get(rif, dir);
+	if (!p_counter_index)
+		return -EINVAL;
+
+	mlxsw_reg_ricnt_pack(ricnt_pl, *p_counter_index,
+			     MLXSW_REG_RICNT_OPCODE_CLEAR);
+	err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ricnt), ricnt_pl);
+	if (err)
+		return err;
+
+	if (!set)
+		return 0;
+
+#define MLXSW_SP_RIF_COUNTER_EXTRACT(NAME)				\
+		(set->NAME = mlxsw_reg_ricnt_ ## NAME ## _get(ricnt_pl))
+
+	MLXSW_SP_RIF_COUNTER_EXTRACT(good_unicast_packets);
+	MLXSW_SP_RIF_COUNTER_EXTRACT(good_multicast_packets);
+	MLXSW_SP_RIF_COUNTER_EXTRACT(good_broadcast_packets);
+	MLXSW_SP_RIF_COUNTER_EXTRACT(good_unicast_bytes);
+	MLXSW_SP_RIF_COUNTER_EXTRACT(good_multicast_bytes);
+	MLXSW_SP_RIF_COUNTER_EXTRACT(good_broadcast_bytes);
+	MLXSW_SP_RIF_COUNTER_EXTRACT(error_packets);
+	MLXSW_SP_RIF_COUNTER_EXTRACT(discard_packets);
+	MLXSW_SP_RIF_COUNTER_EXTRACT(error_bytes);
+	MLXSW_SP_RIF_COUNTER_EXTRACT(discard_bytes);
+
+#undef MLXSW_SP_RIF_COUNTER_EXTRACT
+
+	return 0;
+}
+
 static int mlxsw_sp_rif_counter_clear(struct mlxsw_sp *mlxsw_sp,
 				      unsigned int counter_index)
 {
@@ -235,16 +293,20 @@ static int mlxsw_sp_rif_counter_clear(struct mlxsw_sp *mlxsw_sp,
 	return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ricnt), ricnt_pl);
 }
 
-int mlxsw_sp_rif_counter_alloc(struct mlxsw_sp *mlxsw_sp,
-			       struct mlxsw_sp_rif *rif,
+int mlxsw_sp_rif_counter_alloc(struct mlxsw_sp_rif *rif,
 			       enum mlxsw_sp_rif_counter_dir dir)
 {
+	struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp;
 	unsigned int *p_counter_index;
 	int err;
 
+	if (mlxsw_sp_rif_counter_valid_get(rif, dir))
+		return 0;
+
 	p_counter_index = mlxsw_sp_rif_p_counter_get(rif, dir);
 	if (!p_counter_index)
 		return -EINVAL;
+
 	err = mlxsw_sp_counter_alloc(mlxsw_sp, MLXSW_SP_COUNTER_SUB_POOL_RIF,
 				     p_counter_index);
 	if (err)
@@ -268,10 +330,10 @@ err_counter_clear:
 	return err;
 }
 
-void mlxsw_sp_rif_counter_free(struct mlxsw_sp *mlxsw_sp,
-			       struct mlxsw_sp_rif *rif,
+void mlxsw_sp_rif_counter_free(struct mlxsw_sp_rif *rif,
 			       enum mlxsw_sp_rif_counter_dir dir)
 {
+	struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp;
 	unsigned int *p_counter_index;
 
 	if (!mlxsw_sp_rif_counter_valid_get(rif, dir))
@@ -296,14 +358,12 @@ static void mlxsw_sp_rif_counters_alloc(struct mlxsw_sp_rif *rif)
 	if (!devlink_dpipe_table_counter_enabled(devlink,
 						 MLXSW_SP_DPIPE_TABLE_NAME_ERIF))
 		return;
-	mlxsw_sp_rif_counter_alloc(mlxsw_sp, rif, MLXSW_SP_RIF_COUNTER_EGRESS);
+	mlxsw_sp_rif_counter_alloc(rif, MLXSW_SP_RIF_COUNTER_EGRESS);
 }
 
 static void mlxsw_sp_rif_counters_free(struct mlxsw_sp_rif *rif)
 {
-	struct mlxsw_sp *mlxsw_sp = rif->mlxsw_sp;
-
-	mlxsw_sp_rif_counter_free(mlxsw_sp, rif, MLXSW_SP_RIF_COUNTER_EGRESS);
+	mlxsw_sp_rif_counter_free(rif, MLXSW_SP_RIF_COUNTER_EGRESS);
 }
 
 #define MLXSW_SP_PREFIX_COUNT (sizeof(struct in6_addr) * BITS_PER_BYTE + 1)
@@ -8148,6 +8208,166 @@ u16 mlxsw_sp_ipip_lb_ul_rif_id(const struct mlxsw_sp_rif_ipip_lb *lb_rif)
 	return lb_rif->ul_rif_id;
 }
 
+static bool
+mlxsw_sp_router_port_l3_stats_enabled(struct mlxsw_sp_rif *rif)
+{
+	return mlxsw_sp_rif_counter_valid_get(rif,
+					      MLXSW_SP_RIF_COUNTER_EGRESS) &&
+	       mlxsw_sp_rif_counter_valid_get(rif,
+					      MLXSW_SP_RIF_COUNTER_INGRESS);
+}
+
+static int
+mlxsw_sp_router_port_l3_stats_enable(struct mlxsw_sp_rif *rif)
+{
+	int err;
+
+	err = mlxsw_sp_rif_counter_alloc(rif, MLXSW_SP_RIF_COUNTER_INGRESS);
+	if (err)
+		return err;
+
+	/* Clear stale data. */
+	err = mlxsw_sp_rif_counter_fetch_clear(rif,
+					       MLXSW_SP_RIF_COUNTER_INGRESS,
+					       NULL);
+	if (err)
+		goto err_clear_ingress;
+
+	err = mlxsw_sp_rif_counter_alloc(rif, MLXSW_SP_RIF_COUNTER_EGRESS);
+	if (err)
+		goto err_alloc_egress;
+
+	/* Clear stale data. */
+	err = mlxsw_sp_rif_counter_fetch_clear(rif,
+					       MLXSW_SP_RIF_COUNTER_EGRESS,
+					       NULL);
+	if (err)
+		goto err_clear_egress;
+
+	return 0;
+
+err_clear_egress:
+	mlxsw_sp_rif_counter_free(rif, MLXSW_SP_RIF_COUNTER_EGRESS);
+err_alloc_egress:
+err_clear_ingress:
+	mlxsw_sp_rif_counter_free(rif, MLXSW_SP_RIF_COUNTER_INGRESS);
+	return err;
+}
+
+static void
+mlxsw_sp_router_port_l3_stats_disable(struct mlxsw_sp_rif *rif)
+{
+	mlxsw_sp_rif_counter_free(rif, MLXSW_SP_RIF_COUNTER_EGRESS);
+	mlxsw_sp_rif_counter_free(rif, MLXSW_SP_RIF_COUNTER_INGRESS);
+}
+
+static void
+mlxsw_sp_router_port_l3_stats_report_used(struct mlxsw_sp_rif *rif,
+					  struct netdev_notifier_offload_xstats_info *info)
+{
+	if (!mlxsw_sp_router_port_l3_stats_enabled(rif))
+		return;
+	netdev_offload_xstats_report_used(info->report_used);
+}
+
+static int
+mlxsw_sp_router_port_l3_stats_fetch(struct mlxsw_sp_rif *rif,
+				    struct rtnl_hw_stats64 *p_stats)
+{
+	struct mlxsw_sp_rif_counter_set_basic ingress;
+	struct mlxsw_sp_rif_counter_set_basic egress;
+	int err;
+
+	err = mlxsw_sp_rif_counter_fetch_clear(rif,
+					       MLXSW_SP_RIF_COUNTER_INGRESS,
+					       &ingress);
+	if (err)
+		return err;
+
+	err = mlxsw_sp_rif_counter_fetch_clear(rif,
+					       MLXSW_SP_RIF_COUNTER_EGRESS,
+					       &egress);
+	if (err)
+		return err;
+
+#define MLXSW_SP_ROUTER_ALL_GOOD(SET, SFX)		\
+		((SET.good_unicast_ ## SFX) +		\
+		 (SET.good_multicast_ ## SFX) +		\
+		 (SET.good_broadcast_ ## SFX))
+
+	p_stats->rx_packets = MLXSW_SP_ROUTER_ALL_GOOD(ingress, packets);
+	p_stats->tx_packets = MLXSW_SP_ROUTER_ALL_GOOD(egress, packets);
+	p_stats->rx_bytes = MLXSW_SP_ROUTER_ALL_GOOD(ingress, bytes);
+	p_stats->tx_bytes = MLXSW_SP_ROUTER_ALL_GOOD(egress, bytes);
+	p_stats->rx_errors = ingress.error_packets;
+	p_stats->tx_errors = egress.error_packets;
+	p_stats->rx_dropped = ingress.discard_packets;
+	p_stats->tx_dropped = egress.discard_packets;
+	p_stats->multicast = ingress.good_multicast_packets +
+			     ingress.good_broadcast_packets;
+
+#undef MLXSW_SP_ROUTER_ALL_GOOD
+
+	return 0;
+}
+
+static int
+mlxsw_sp_router_port_l3_stats_report_delta(struct mlxsw_sp_rif *rif,
+					   struct netdev_notifier_offload_xstats_info *info)
+{
+	struct rtnl_hw_stats64 stats = {};
+	int err;
+
+	if (!mlxsw_sp_router_port_l3_stats_enabled(rif))
+		return 0;
+
+	err = mlxsw_sp_router_port_l3_stats_fetch(rif, &stats);
+	if (err)
+		return err;
+
+	netdev_offload_xstats_report_delta(info->report_delta, &stats);
+	return 0;
+}
+
+struct mlxsw_sp_router_hwstats_notify_work {
+	struct work_struct work;
+	struct net_device *dev;
+};
+
+static void mlxsw_sp_router_hwstats_notify_work(struct work_struct *work)
+{
+	struct mlxsw_sp_router_hwstats_notify_work *hws_work =
+		container_of(work, struct mlxsw_sp_router_hwstats_notify_work,
+			     work);
+
+	rtnl_lock();
+	rtnl_offload_xstats_notify(hws_work->dev);
+	rtnl_unlock();
+	dev_put(hws_work->dev);
+	kfree(hws_work);
+}
+
+static void
+mlxsw_sp_router_hwstats_notify_schedule(struct net_device *dev)
+{
+	struct mlxsw_sp_router_hwstats_notify_work *hws_work;
+
+	/* To collect notification payload, the core ends up sending another
+	 * notifier block message, which would deadlock on the attempt to
+	 * acquire the router lock again. Just postpone the notification until
+	 * later.
+	 */
+
+	hws_work = kzalloc(sizeof(*hws_work), GFP_KERNEL);
+	if (!hws_work)
+		return;
+
+	INIT_WORK(&hws_work->work, mlxsw_sp_router_hwstats_notify_work);
+	dev_hold(dev);
+	hws_work->dev = dev;
+	mlxsw_core_schedule_work(&hws_work->work);
+}
+
 int mlxsw_sp_rif_dev_ifindex(const struct mlxsw_sp_rif *rif)
 {
 	return rif->dev->ifindex;
@@ -8158,6 +8378,16 @@ const struct net_device *mlxsw_sp_rif_dev(const struct mlxsw_sp_rif *rif)
 	return rif->dev;
 }
 
+static void mlxsw_sp_rif_push_l3_stats(struct mlxsw_sp_rif *rif)
+{
+	struct rtnl_hw_stats64 stats = {};
+
+	if (!mlxsw_sp_router_port_l3_stats_fetch(rif, &stats))
+		netdev_offload_xstats_push_delta(rif->dev,
+						 NETDEV_OFFLOAD_XSTATS_TYPE_L3,
+						 &stats);
+}
+
 static struct mlxsw_sp_rif *
 mlxsw_sp_rif_create(struct mlxsw_sp *mlxsw_sp,
 		    const struct mlxsw_sp_rif_params *params,
@@ -8218,10 +8448,19 @@ mlxsw_sp_rif_create(struct mlxsw_sp *mlxsw_sp,
 			goto err_mr_rif_add;
 	}
 
-	mlxsw_sp_rif_counters_alloc(rif);
+	if (netdev_offload_xstats_enabled(rif->dev,
+					  NETDEV_OFFLOAD_XSTATS_TYPE_L3)) {
+		err = mlxsw_sp_router_port_l3_stats_enable(rif);
+		if (err)
+			goto err_stats_enable;
+		mlxsw_sp_router_hwstats_notify_schedule(rif->dev);
+	} else {
+		mlxsw_sp_rif_counters_alloc(rif);
+	}
 
 	return rif;
 
+err_stats_enable:
 err_mr_rif_add:
 	for (i--; i >= 0; i--)
 		mlxsw_sp_mr_rif_del(vr->mr_table[i], rif);
@@ -8251,7 +8490,15 @@ static void mlxsw_sp_rif_destroy(struct mlxsw_sp_rif *rif)
 	mlxsw_sp_router_rif_gone_sync(mlxsw_sp, rif);
 	vr = &mlxsw_sp->router->vrs[rif->vr_id];
 
-	mlxsw_sp_rif_counters_free(rif);
+	if (netdev_offload_xstats_enabled(rif->dev,
+					  NETDEV_OFFLOAD_XSTATS_TYPE_L3)) {
+		mlxsw_sp_rif_push_l3_stats(rif);
+		mlxsw_sp_router_port_l3_stats_disable(rif);
+		mlxsw_sp_router_hwstats_notify_schedule(rif->dev);
+	} else {
+		mlxsw_sp_rif_counters_free(rif);
+	}
+
 	for (i = 0; i < MLXSW_SP_L3_PROTO_MAX; i++)
 		mlxsw_sp_mr_rif_del(vr->mr_table[i], rif);
 	ops->deconfigure(rif);
@@ -9128,6 +9375,35 @@ static int mlxsw_sp_router_port_pre_changeaddr_event(struct mlxsw_sp_rif *rif,
 	return -ENOBUFS;
 }
 
+static int
+mlxsw_sp_router_port_offload_xstats_cmd(struct mlxsw_sp_rif *rif,
+					unsigned long event,
+					struct netdev_notifier_offload_xstats_info *info)
+{
+	switch (info->type) {
+	case NETDEV_OFFLOAD_XSTATS_TYPE_L3:
+		break;
+	default:
+		return 0;
+	}
+
+	switch (event) {
+	case NETDEV_OFFLOAD_XSTATS_ENABLE:
+		return mlxsw_sp_router_port_l3_stats_enable(rif);
+	case NETDEV_OFFLOAD_XSTATS_DISABLE:
+		mlxsw_sp_router_port_l3_stats_disable(rif);
+		return 0;
+	case NETDEV_OFFLOAD_XSTATS_REPORT_USED:
+		mlxsw_sp_router_port_l3_stats_report_used(rif, info);
+		return 0;
+	case NETDEV_OFFLOAD_XSTATS_REPORT_DELTA:
+		return mlxsw_sp_router_port_l3_stats_report_delta(rif, info);
+	}
+
+	WARN_ON_ONCE(1);
+	return 0;
+}
+
 int mlxsw_sp_netdevice_router_port_event(struct net_device *dev,
 					 unsigned long event, void *ptr)
 {
@@ -9153,6 +9429,15 @@ int mlxsw_sp_netdevice_router_port_event(struct net_device *dev,
 	case NETDEV_PRE_CHANGEADDR:
 		err = mlxsw_sp_router_port_pre_changeaddr_event(rif, ptr);
 		break;
+	case NETDEV_OFFLOAD_XSTATS_ENABLE:
+	case NETDEV_OFFLOAD_XSTATS_DISABLE:
+	case NETDEV_OFFLOAD_XSTATS_REPORT_USED:
+	case NETDEV_OFFLOAD_XSTATS_REPORT_DELTA:
+		err = mlxsw_sp_router_port_offload_xstats_cmd(rif, event, ptr);
+		break;
+	default:
+		WARN_ON_ONCE(1);
+		break;
 	}
 
 out:
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h
index 99e8371a82a5..fa829658a11b 100644
--- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h
+++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h
@@ -159,11 +159,9 @@ int mlxsw_sp_rif_counter_value_get(struct mlxsw_sp *mlxsw_sp,
 				   struct mlxsw_sp_rif *rif,
 				   enum mlxsw_sp_rif_counter_dir dir,
 				   u64 *cnt);
-void mlxsw_sp_rif_counter_free(struct mlxsw_sp *mlxsw_sp,
-			       struct mlxsw_sp_rif *rif,
+void mlxsw_sp_rif_counter_free(struct mlxsw_sp_rif *rif,
 			       enum mlxsw_sp_rif_counter_dir dir);
-int mlxsw_sp_rif_counter_alloc(struct mlxsw_sp *mlxsw_sp,
-			       struct mlxsw_sp_rif *rif,
+int mlxsw_sp_rif_counter_alloc(struct mlxsw_sp_rif *rif,
 			       enum mlxsw_sp_rif_counter_dir dir);
 struct mlxsw_sp_neigh_entry *
 mlxsw_sp_rif_neigh_next(struct mlxsw_sp_rif *rif,
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index c79ee2296296..19a27ac361ef 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -1950,6 +1950,7 @@ enum netdev_ml_priv_type {
  *	@watchdog_dev_tracker:	refcount tracker used by watchdog.
  *	@dev_registered_tracker:	tracker for reference held while
  *					registered
+ *	@offload_xstats_l3:	L3 HW stats for this netdevice.
  *
  *	FIXME: cleanup struct net_device such that network protocol info
  *	moves out.
@@ -2287,6 +2288,7 @@ struct net_device {
 	netdevice_tracker	linkwatch_dev_tracker;
 	netdevice_tracker	watchdog_dev_tracker;
 	netdevice_tracker	dev_registered_tracker;
+	struct rtnl_hw_stats64	*offload_xstats_l3;
 };
 #define to_net_dev(d) container_of(d, struct net_device, dev)
 
@@ -2727,6 +2729,10 @@ enum netdev_cmd {
 	NETDEV_CVLAN_FILTER_DROP_INFO,
 	NETDEV_SVLAN_FILTER_PUSH_INFO,
 	NETDEV_SVLAN_FILTER_DROP_INFO,
+	NETDEV_OFFLOAD_XSTATS_ENABLE,
+	NETDEV_OFFLOAD_XSTATS_DISABLE,
+	NETDEV_OFFLOAD_XSTATS_REPORT_USED,
+	NETDEV_OFFLOAD_XSTATS_REPORT_DELTA,
 };
 const char *netdev_cmd_to_name(enum netdev_cmd cmd);
 
@@ -2777,6 +2783,42 @@ struct netdev_notifier_pre_changeaddr_info {
 	const unsigned char *dev_addr;
 };
 
+enum netdev_offload_xstats_type {
+	NETDEV_OFFLOAD_XSTATS_TYPE_L3 = 1,
+};
+
+struct netdev_notifier_offload_xstats_info {
+	struct netdev_notifier_info info; /* must be first */
+	enum netdev_offload_xstats_type type;
+
+	union {
+		/* NETDEV_OFFLOAD_XSTATS_REPORT_DELTA */
+		struct netdev_notifier_offload_xstats_rd *report_delta;
+		/* NETDEV_OFFLOAD_XSTATS_REPORT_USED */
+		struct netdev_notifier_offload_xstats_ru *report_used;
+	};
+};
+
+int netdev_offload_xstats_enable(struct net_device *dev,
+				 enum netdev_offload_xstats_type type,
+				 struct netlink_ext_ack *extack);
+int netdev_offload_xstats_disable(struct net_device *dev,
+				  enum netdev_offload_xstats_type type);
+bool netdev_offload_xstats_enabled(const struct net_device *dev,
+				   enum netdev_offload_xstats_type type);
+int netdev_offload_xstats_get(struct net_device *dev,
+			      enum netdev_offload_xstats_type type,
+			      struct rtnl_hw_stats64 *stats, bool *used,
+			      struct netlink_ext_ack *extack);
+void
+netdev_offload_xstats_report_delta(struct netdev_notifier_offload_xstats_rd *rd,
+				   const struct rtnl_hw_stats64 *stats);
+void
+netdev_offload_xstats_report_used(struct netdev_notifier_offload_xstats_ru *ru);
+void netdev_offload_xstats_push_delta(struct net_device *dev,
+				      enum netdev_offload_xstats_type type,
+				      const struct rtnl_hw_stats64 *stats);
+
 static inline void netdev_notifier_info_init(struct netdev_notifier_info *info,
 					     struct net_device *dev)
 {
diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index bb9cb84114c1..7f970b16da3a 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -134,4 +134,7 @@ extern int ndo_dflt_bridge_getlink(struct sk_buff *skb, u32 pid, u32 seq,
 				   int (*vlan_fill)(struct sk_buff *skb,
 						    struct net_device *dev,
 						    u32 filter_mask));
+
+extern void rtnl_offload_xstats_notify(struct net_device *dev);
+
 #endif	/* __LINUX_RTNETLINK_H */
diff --git a/include/uapi/linux/if_link.h b/include/uapi/linux/if_link.h
index e315e53125f4..ddca20357e7e 100644
--- a/include/uapi/linux/if_link.h
+++ b/include/uapi/linux/if_link.h
@@ -245,6 +245,21 @@ struct rtnl_link_stats64 {
 	__u64	rx_nohandler;
 };
 
+/* Subset of link stats useful for in-HW collection. Meaning of the fields is as
+ * for struct rtnl_link_stats64.
+ */
+struct rtnl_hw_stats64 {
+	__u64	rx_packets;
+	__u64	tx_packets;
+	__u64	rx_bytes;
+	__u64	tx_bytes;
+	__u64	rx_errors;
+	__u64	tx_errors;
+	__u64	rx_dropped;
+	__u64	tx_dropped;
+	__u64	multicast;
+};
+
 /* The struct should be in sync with struct ifmap */
 struct rtnl_link_ifmap {
 	__u64	mem_start;
@@ -1207,6 +1222,17 @@ enum {
 
 #define IFLA_STATS_FILTER_BIT(ATTR)	(1 << (ATTR - 1))
 
+enum {
+	IFLA_STATS_GETSET_UNSPEC,
+	IFLA_STATS_GET_FILTERS, /* Nest of IFLA_STATS_LINK_xxx, each a u32 with
+				 * a filter mask for the corresponding group.
+				 */
+	IFLA_STATS_SET_OFFLOAD_XSTATS_L3_STATS, /* 0 or 1 as u8 */
+	__IFLA_STATS_GETSET_MAX,
+};
+
+#define IFLA_STATS_GETSET_MAX (__IFLA_STATS_GETSET_MAX - 1)
+
 /* These are embedded into IFLA_STATS_LINK_XSTATS:
  * [IFLA_STATS_LINK_XSTATS]
  * -> [LINK_XSTATS_TYPE_xxx]
@@ -1224,10 +1250,21 @@ enum {
 enum {
 	IFLA_OFFLOAD_XSTATS_UNSPEC,
 	IFLA_OFFLOAD_XSTATS_CPU_HIT, /* struct rtnl_link_stats64 */
+	IFLA_OFFLOAD_XSTATS_HW_S_INFO,	/* HW stats info. A nest */
+	IFLA_OFFLOAD_XSTATS_L3_STATS,	/* struct rtnl_hw_stats64 */
 	__IFLA_OFFLOAD_XSTATS_MAX
 };
 #define IFLA_OFFLOAD_XSTATS_MAX (__IFLA_OFFLOAD_XSTATS_MAX - 1)
 
+enum {
+	IFLA_OFFLOAD_XSTATS_HW_S_INFO_UNSPEC,
+	IFLA_OFFLOAD_XSTATS_HW_S_INFO_REQUEST,		/* u8 */
+	IFLA_OFFLOAD_XSTATS_HW_S_INFO_USED,		/* u8 */
+	__IFLA_OFFLOAD_XSTATS_HW_S_INFO_MAX,
+};
+#define IFLA_OFFLOAD_XSTATS_HW_S_INFO_MAX \
+	(__IFLA_OFFLOAD_XSTATS_HW_S_INFO_MAX - 1)
+
 /* XDP section */
 
 #define XDP_FLAGS_UPDATE_IF_NOEXIST	(1U << 0)
diff --git a/include/uapi/linux/rtnetlink.h b/include/uapi/linux/rtnetlink.h
index 0970cb4b1b88..51530aade46e 100644
--- a/include/uapi/linux/rtnetlink.h
+++ b/include/uapi/linux/rtnetlink.h
@@ -146,6 +146,8 @@ enum {
 #define RTM_NEWSTATS RTM_NEWSTATS
 	RTM_GETSTATS = 94,
 #define RTM_GETSTATS RTM_GETSTATS
+	RTM_SETSTATS,
+#define RTM_SETSTATS RTM_SETSTATS
 
 	RTM_NEWCACHEREPORT = 96,
 #define RTM_NEWCACHEREPORT RTM_NEWCACHEREPORT
@@ -765,6 +767,8 @@ enum rtnetlink_groups {
 #define RTNLGRP_MCTP_IFADDR	RTNLGRP_MCTP_IFADDR
 	RTNLGRP_TUNNEL,
 #define RTNLGRP_TUNNEL		RTNLGRP_TUNNEL
+	RTNLGRP_STATS,
+#define RTNLGRP_STATS		RTNLGRP_STATS
 	__RTNLGRP_MAX
 };
 #define RTNLGRP_MAX	(__RTNLGRP_MAX - 1)
diff --git a/net/core/dev.c b/net/core/dev.c
index 2d6771075720..c9e54e5ad48d 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1622,7 +1622,8 @@ const char *netdev_cmd_to_name(enum netdev_cmd cmd)
 	N(UDP_TUNNEL_DROP_INFO) N(CHANGE_TX_QUEUE_LEN)
 	N(CVLAN_FILTER_PUSH_INFO) N(CVLAN_FILTER_DROP_INFO)
 	N(SVLAN_FILTER_PUSH_INFO) N(SVLAN_FILTER_DROP_INFO)
-	N(PRE_CHANGEADDR)
+	N(PRE_CHANGEADDR) N(OFFLOAD_XSTATS_ENABLE) N(OFFLOAD_XSTATS_DISABLE)
+	N(OFFLOAD_XSTATS_REPORT_USED) N(OFFLOAD_XSTATS_REPORT_DELTA)
 	}
 #undef N
 	return "UNKNOWN_NETDEV_EVENT";
@@ -1939,6 +1940,32 @@ static int call_netdevice_notifiers_info(unsigned long val,
 	return raw_notifier_call_chain(&netdev_chain, val, info);
 }
 
+/**
+ *	call_netdevice_notifiers_info_robust - call per-netns notifier blocks
+ *	                                       for and rollback on error
+ *	@val_up: value passed unmodified to notifier function
+ *	@val_down: value passed unmodified to the notifier function when
+ *	           recovering from an error on @val_up
+ *	@info: notifier information data
+ *
+ *	Call all per-netns network notifier blocks, but not notifier blocks on
+ *	the global notifier chain. Parameters and return value are as for
+ *	raw_notifier_call_chain_robust().
+ */
+
+static int
+call_netdevice_notifiers_info_robust(unsigned long val_up,
+				     unsigned long val_down,
+				     struct netdev_notifier_info *info)
+{
+	struct net *net = dev_net(info->dev);
+
+	ASSERT_RTNL();
+
+	return raw_notifier_call_chain_robust(&net->netdev_chain,
+					      val_up, val_down, info);
+}
+
 static int call_netdevice_notifiers_extack(unsigned long val,
 					   struct net_device *dev,
 					   struct netlink_ext_ack *extack)
@@ -7728,6 +7755,242 @@ void netdev_bonding_info_change(struct net_device *dev,
 }
 EXPORT_SYMBOL(netdev_bonding_info_change);
 
+static int netdev_offload_xstats_enable_l3(struct net_device *dev,
+					   struct netlink_ext_ack *extack)
+{
+	struct netdev_notifier_offload_xstats_info info = {
+		.info.dev = dev,
+		.info.extack = extack,
+		.type = NETDEV_OFFLOAD_XSTATS_TYPE_L3,
+	};
+	int err;
+	int rc;
+
+	dev->offload_xstats_l3 = kzalloc(sizeof(*dev->offload_xstats_l3),
+					 GFP_KERNEL);
+	if (!dev->offload_xstats_l3)
+		return -ENOMEM;
+
+	rc = call_netdevice_notifiers_info_robust(NETDEV_OFFLOAD_XSTATS_ENABLE,
+						  NETDEV_OFFLOAD_XSTATS_DISABLE,
+						  &info.info);
+	err = notifier_to_errno(rc);
+	if (err)
+		goto free_stats;
+
+	return 0;
+
+free_stats:
+	kfree(dev->offload_xstats_l3);
+	dev->offload_xstats_l3 = NULL;
+	return err;
+}
+
+int netdev_offload_xstats_enable(struct net_device *dev,
+				 enum netdev_offload_xstats_type type,
+				 struct netlink_ext_ack *extack)
+{
+	ASSERT_RTNL();
+
+	if (netdev_offload_xstats_enabled(dev, type))
+		return -EALREADY;
+
+	switch (type) {
+	case NETDEV_OFFLOAD_XSTATS_TYPE_L3:
+		return netdev_offload_xstats_enable_l3(dev, extack);
+	}
+
+	WARN_ON(1);
+	return -EINVAL;
+}
+EXPORT_SYMBOL(netdev_offload_xstats_enable);
+
+static void netdev_offload_xstats_disable_l3(struct net_device *dev)
+{
+	struct netdev_notifier_offload_xstats_info info = {
+		.info.dev = dev,
+		.type = NETDEV_OFFLOAD_XSTATS_TYPE_L3,
+	};
+
+	call_netdevice_notifiers_info(NETDEV_OFFLOAD_XSTATS_DISABLE,
+				      &info.info);
+	kfree(dev->offload_xstats_l3);
+	dev->offload_xstats_l3 = NULL;
+}
+
+int netdev_offload_xstats_disable(struct net_device *dev,
+				  enum netdev_offload_xstats_type type)
+{
+	ASSERT_RTNL();
+
+	if (!netdev_offload_xstats_enabled(dev, type))
+		return -EALREADY;
+
+	switch (type) {
+	case NETDEV_OFFLOAD_XSTATS_TYPE_L3:
+		netdev_offload_xstats_disable_l3(dev);
+		return 0;
+	}
+
+	WARN_ON(1);
+	return -EINVAL;
+}
+EXPORT_SYMBOL(netdev_offload_xstats_disable);
+
+static void netdev_offload_xstats_disable_all(struct net_device *dev)
+{
+	netdev_offload_xstats_disable(dev, NETDEV_OFFLOAD_XSTATS_TYPE_L3);
+}
+
+static struct rtnl_hw_stats64 *
+netdev_offload_xstats_get_ptr(const struct net_device *dev,
+			      enum netdev_offload_xstats_type type)
+{
+	switch (type) {
+	case NETDEV_OFFLOAD_XSTATS_TYPE_L3:
+		return dev->offload_xstats_l3;
+	}
+
+	WARN_ON(1);
+	return NULL;
+}
+
+bool netdev_offload_xstats_enabled(const struct net_device *dev,
+				   enum netdev_offload_xstats_type type)
+{
+	ASSERT_RTNL();
+
+	return netdev_offload_xstats_get_ptr(dev, type);
+}
+EXPORT_SYMBOL(netdev_offload_xstats_enabled);
+
+struct netdev_notifier_offload_xstats_ru {
+	bool used;
+};
+
+struct netdev_notifier_offload_xstats_rd {
+	struct rtnl_hw_stats64 stats;
+	bool used;
+};
+
+static void netdev_hw_stats64_add(struct rtnl_hw_stats64 *dest,
+				  const struct rtnl_hw_stats64 *src)
+{
+	dest->rx_packets	  += src->rx_packets;
+	dest->tx_packets	  += src->tx_packets;
+	dest->rx_bytes		  += src->rx_bytes;
+	dest->tx_bytes		  += src->tx_bytes;
+	dest->rx_errors		  += src->rx_errors;
+	dest->tx_errors		  += src->tx_errors;
+	dest->rx_dropped	  += src->rx_dropped;
+	dest->tx_dropped	  += src->tx_dropped;
+	dest->multicast		  += src->multicast;
+}
+
+static int netdev_offload_xstats_get_used(struct net_device *dev,
+					  enum netdev_offload_xstats_type type,
+					  bool *p_used,
+					  struct netlink_ext_ack *extack)
+{
+	struct netdev_notifier_offload_xstats_ru report_used = {};
+	struct netdev_notifier_offload_xstats_info info = {
+		.info.dev = dev,
+		.info.extack = extack,
+		.type = type,
+		.report_used = &report_used,
+	};
+	int rc;
+
+	WARN_ON(!netdev_offload_xstats_enabled(dev, type));
+	rc = call_netdevice_notifiers_info(NETDEV_OFFLOAD_XSTATS_REPORT_USED,
+					   &info.info);
+	*p_used = report_used.used;
+	return notifier_to_errno(rc);
+}
+
+static int netdev_offload_xstats_get_stats(struct net_device *dev,
+					   enum netdev_offload_xstats_type type,
+					   struct rtnl_hw_stats64 *p_stats,
+					   bool *p_used,
+					   struct netlink_ext_ack *extack)
+{
+	struct netdev_notifier_offload_xstats_rd report_delta = {};
+	struct netdev_notifier_offload_xstats_info info = {
+		.info.dev = dev,
+		.info.extack = extack,
+		.type = type,
+		.report_delta = &report_delta,
+	};
+	struct rtnl_hw_stats64 *stats;
+	int rc;
+
+	stats = netdev_offload_xstats_get_ptr(dev, type);
+	if (WARN_ON(!stats))
+		return -EINVAL;
+
+	rc = call_netdevice_notifiers_info(NETDEV_OFFLOAD_XSTATS_REPORT_DELTA,
+					   &info.info);
+
+	/* Cache whatever we got, even if there was an error, otherwise the
+	 * successful stats retrievals would get lost.
+	 */
+	netdev_hw_stats64_add(stats, &report_delta.stats);
+
+	if (p_stats)
+		*p_stats = *stats;
+	*p_used = report_delta.used;
+
+	return notifier_to_errno(rc);
+}
+
+int netdev_offload_xstats_get(struct net_device *dev,
+			      enum netdev_offload_xstats_type type,
+			      struct rtnl_hw_stats64 *p_stats, bool *p_used,
+			      struct netlink_ext_ack *extack)
+{
+	ASSERT_RTNL();
+
+	if (p_stats)
+		return netdev_offload_xstats_get_stats(dev, type, p_stats,
+						       p_used, extack);
+	else
+		return netdev_offload_xstats_get_used(dev, type, p_used,
+						      extack);
+}
+EXPORT_SYMBOL(netdev_offload_xstats_get);
+
+void
+netdev_offload_xstats_report_delta(struct netdev_notifier_offload_xstats_rd *report_delta,
+				   const struct rtnl_hw_stats64 *stats)
+{
+	report_delta->used = true;
+	netdev_hw_stats64_add(&report_delta->stats, stats);
+}
+EXPORT_SYMBOL(netdev_offload_xstats_report_delta);
+
+void
+netdev_offload_xstats_report_used(struct netdev_notifier_offload_xstats_ru *report_used)
+{
+	report_used->used = true;
+}
+EXPORT_SYMBOL(netdev_offload_xstats_report_used);
+
+void netdev_offload_xstats_push_delta(struct net_device *dev,
+				      enum netdev_offload_xstats_type type,
+				      const struct rtnl_hw_stats64 *p_stats)
+{
+	struct rtnl_hw_stats64 *stats;
+
+	ASSERT_RTNL();
+
+	stats = netdev_offload_xstats_get_ptr(dev, type);
+	if (WARN_ON(!stats))
+		return;
+
+	netdev_hw_stats64_add(stats, p_stats);
+}
+EXPORT_SYMBOL(netdev_offload_xstats_push_delta);
+
 /**
  * netdev_get_xmit_slave - Get the xmit slave of master device
  * @dev: device
@@ -10417,6 +10680,8 @@ void unregister_netdevice_many(struct list_head *head)
 
 		dev_xdp_uninstall(dev);
 
+		netdev_offload_xstats_disable_all(dev);
+
 		/* Notify protocols, that we are about to destroy
 		 * this device. They should clean all the things.
 		 */
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
index 20a9e1686453..a66b6761b88b 100644
--- a/net/core/rtnetlink.c
+++ b/net/core/rtnetlink.c
@@ -5048,82 +5048,256 @@ static bool stats_attr_valid(unsigned int mask, int attrid, int idxattr)
 	       (!idxattr || idxattr == attrid);
 }
 
-#define IFLA_OFFLOAD_XSTATS_FIRST (IFLA_OFFLOAD_XSTATS_UNSPEC + 1)
-static int rtnl_get_offload_stats_attr_size(int attr_id)
+static bool
+rtnl_offload_xstats_have_ndo(const struct net_device *dev, int attr_id)
 {
-	switch (attr_id) {
-	case IFLA_OFFLOAD_XSTATS_CPU_HIT:
-		return sizeof(struct rtnl_link_stats64);
-	}
+	return dev->netdev_ops &&
+	       dev->netdev_ops->ndo_has_offload_stats &&
+	       dev->netdev_ops->ndo_get_offload_stats &&
+	       dev->netdev_ops->ndo_has_offload_stats(dev, attr_id);
+}
+
+static unsigned int
+rtnl_offload_xstats_get_size_ndo(const struct net_device *dev, int attr_id)
+{
+	return rtnl_offload_xstats_have_ndo(dev, attr_id) ?
+	       sizeof(struct rtnl_link_stats64) : 0;
+}
+
+static int
+rtnl_offload_xstats_fill_ndo(struct net_device *dev, int attr_id,
+			     struct sk_buff *skb)
+{
+	unsigned int size = rtnl_offload_xstats_get_size_ndo(dev, attr_id);
+	struct nlattr *attr = NULL;
+	void *attr_data;
+	int err;
+
+	if (!size)
+		return -ENODATA;
+
+	attr = nla_reserve_64bit(skb, attr_id, size,
+				 IFLA_OFFLOAD_XSTATS_UNSPEC);
+	if (!attr)
+		return -EMSGSIZE;
+
+	attr_data = nla_data(attr);
+	memset(attr_data, 0, size);
+
+	err = dev->netdev_ops->ndo_get_offload_stats(attr_id, dev, attr_data);
+	if (err)
+		return err;
 
 	return 0;
 }
 
-static int rtnl_get_offload_stats(struct sk_buff *skb, struct net_device *dev,
-				  int *prividx)
+static unsigned int
+rtnl_offload_xstats_get_size_stats(const struct net_device *dev,
+				   enum netdev_offload_xstats_type type)
 {
-	struct nlattr *attr = NULL;
-	int attr_id, size;
-	void *attr_data;
+	bool enabled = netdev_offload_xstats_enabled(dev, type);
+
+	return enabled ? sizeof(struct rtnl_hw_stats64) : 0;
+}
+
+struct rtnl_offload_xstats_request_used {
+	bool request;
+	bool used;
+};
+
+static int
+rtnl_offload_xstats_get_stats(struct net_device *dev,
+			      enum netdev_offload_xstats_type type,
+			      struct rtnl_offload_xstats_request_used *ru,
+			      struct rtnl_hw_stats64 *stats,
+			      struct netlink_ext_ack *extack)
+{
+	bool request;
+	bool used;
 	int err;
 
-	if (!(dev->netdev_ops && dev->netdev_ops->ndo_has_offload_stats &&
-	      dev->netdev_ops->ndo_get_offload_stats))
-		return -ENODATA;
-
-	for (attr_id = IFLA_OFFLOAD_XSTATS_FIRST;
-	     attr_id <= IFLA_OFFLOAD_XSTATS_MAX; attr_id++) {
-		if (attr_id < *prividx)
-			continue;
-
-		size = rtnl_get_offload_stats_attr_size(attr_id);
-		if (!size)
-			continue;
-
-		if (!dev->netdev_ops->ndo_has_offload_stats(dev, attr_id))
-			continue;
-
-		attr = nla_reserve_64bit(skb, attr_id, size,
-					 IFLA_OFFLOAD_XSTATS_UNSPEC);
-		if (!attr)
-			goto nla_put_failure;
-
-		attr_data = nla_data(attr);
-		memset(attr_data, 0, size);
-		err = dev->netdev_ops->ndo_get_offload_stats(attr_id, dev,
-							     attr_data);
-		if (err)
-			goto get_offload_stats_failure;
+	request = netdev_offload_xstats_enabled(dev, type);
+	if (!request) {
+		used = false;
+		goto out;
 	}
 
-	if (!attr)
+	err = netdev_offload_xstats_get(dev, type, stats, &used, extack);
+	if (err)
+		return err;
+
+out:
+	if (ru) {
+		ru->request = request;
+		ru->used = used;
+	}
+	return 0;
+}
+
+static int
+rtnl_offload_xstats_fill_hw_s_info_one(struct sk_buff *skb, int attr_id,
+				       struct rtnl_offload_xstats_request_used *ru)
+{
+	struct nlattr *nest;
+
+	nest = nla_nest_start(skb, attr_id);
+	if (!nest)
+		return -EMSGSIZE;
+
+	if (nla_put_u8(skb, IFLA_OFFLOAD_XSTATS_HW_S_INFO_REQUEST, ru->request))
+		goto nla_put_failure;
+
+	if (nla_put_u8(skb, IFLA_OFFLOAD_XSTATS_HW_S_INFO_USED, ru->used))
+		goto nla_put_failure;
+
+	nla_nest_end(skb, nest);
+	return 0;
+
+nla_put_failure:
+	nla_nest_cancel(skb, nest);
+	return -EMSGSIZE;
+}
+
+static int
+rtnl_offload_xstats_fill_hw_s_info(struct sk_buff *skb, struct net_device *dev,
+				   struct netlink_ext_ack *extack)
+{
+	enum netdev_offload_xstats_type t_l3 = NETDEV_OFFLOAD_XSTATS_TYPE_L3;
+	struct rtnl_offload_xstats_request_used ru_l3;
+	struct nlattr *nest;
+	int err;
+
+	err = rtnl_offload_xstats_get_stats(dev, t_l3, &ru_l3, NULL, extack);
+	if (err)
+		return err;
+
+	nest = nla_nest_start(skb, IFLA_OFFLOAD_XSTATS_HW_S_INFO);
+	if (!nest)
+		return -EMSGSIZE;
+
+	if (rtnl_offload_xstats_fill_hw_s_info_one(skb,
+						   IFLA_OFFLOAD_XSTATS_L3_STATS,
+						   &ru_l3))
+		goto nla_put_failure;
+
+	nla_nest_end(skb, nest);
+	return 0;
+
+nla_put_failure:
+	nla_nest_cancel(skb, nest);
+	return -EMSGSIZE;
+}
+
+static int rtnl_offload_xstats_fill(struct sk_buff *skb, struct net_device *dev,
+				    int *prividx, u32 off_filter_mask,
+				    struct netlink_ext_ack *extack)
+{
+	enum netdev_offload_xstats_type t_l3 = NETDEV_OFFLOAD_XSTATS_TYPE_L3;
+	int attr_id_hw_s_info = IFLA_OFFLOAD_XSTATS_HW_S_INFO;
+	int attr_id_l3_stats = IFLA_OFFLOAD_XSTATS_L3_STATS;
+	int attr_id_cpu_hit = IFLA_OFFLOAD_XSTATS_CPU_HIT;
+	bool have_data = false;
+	int err;
+
+	if (*prividx <= attr_id_cpu_hit &&
+	    (off_filter_mask &
+	     IFLA_STATS_FILTER_BIT(attr_id_cpu_hit))) {
+		err = rtnl_offload_xstats_fill_ndo(dev, attr_id_cpu_hit, skb);
+		if (!err) {
+			have_data = true;
+		} else if (err != -ENODATA) {
+			*prividx = attr_id_cpu_hit;
+			return err;
+		}
+	}
+
+	if (*prividx <= attr_id_hw_s_info &&
+	    (off_filter_mask & IFLA_STATS_FILTER_BIT(attr_id_hw_s_info))) {
+		*prividx = attr_id_hw_s_info;
+
+		err = rtnl_offload_xstats_fill_hw_s_info(skb, dev, extack);
+		if (err)
+			return err;
+
+		have_data = true;
+		*prividx = 0;
+	}
+
+	if (*prividx <= attr_id_l3_stats &&
+	    (off_filter_mask & IFLA_STATS_FILTER_BIT(attr_id_l3_stats))) {
+		unsigned int size_l3;
+		struct nlattr *attr;
+
+		*prividx = attr_id_l3_stats;
+
+		size_l3 = rtnl_offload_xstats_get_size_stats(dev, t_l3);
+		attr = nla_reserve_64bit(skb, attr_id_l3_stats, size_l3,
+					 IFLA_OFFLOAD_XSTATS_UNSPEC);
+		if (!attr)
+			return -EMSGSIZE;
+
+		err = rtnl_offload_xstats_get_stats(dev, t_l3, NULL,
+						    nla_data(attr), extack);
+		if (err)
+			return err;
+
+		have_data = true;
+		*prividx = 0;
+	}
+
+	if (!have_data)
 		return -ENODATA;
 
 	*prividx = 0;
 	return 0;
-
-nla_put_failure:
-	err = -EMSGSIZE;
-get_offload_stats_failure:
-	*prividx = attr_id;
-	return err;
 }
 
-static int rtnl_get_offload_stats_size(const struct net_device *dev)
+static unsigned int
+rtnl_offload_xstats_get_size_hw_s_info_one(const struct net_device *dev,
+					   enum netdev_offload_xstats_type type)
 {
+	bool enabled = netdev_offload_xstats_enabled(dev, type);
+
+	return nla_total_size(0) +
+		/* IFLA_OFFLOAD_XSTATS_HW_S_INFO_REQUEST */
+		nla_total_size(sizeof(u8)) +
+		/* IFLA_OFFLOAD_XSTATS_HW_S_INFO_USED */
+		(enabled ? nla_total_size(sizeof(u8)) : 0) +
+		0;
+}
+
+static unsigned int
+rtnl_offload_xstats_get_size_hw_s_info(const struct net_device *dev)
+{
+	enum netdev_offload_xstats_type t_l3 = NETDEV_OFFLOAD_XSTATS_TYPE_L3;
+
+	return nla_total_size(0) +
+		/* IFLA_OFFLOAD_XSTATS_L3_STATS */
+		rtnl_offload_xstats_get_size_hw_s_info_one(dev, t_l3) +
+		0;
+}
+
+static int rtnl_offload_xstats_get_size(const struct net_device *dev,
+					u32 off_filter_mask)
+{
+	enum netdev_offload_xstats_type t_l3 = NETDEV_OFFLOAD_XSTATS_TYPE_L3;
+	int attr_id_cpu_hit = IFLA_OFFLOAD_XSTATS_CPU_HIT;
 	int nla_size = 0;
-	int attr_id;
 	int size;
 
-	if (!(dev->netdev_ops && dev->netdev_ops->ndo_has_offload_stats &&
-	      dev->netdev_ops->ndo_get_offload_stats))
-		return 0;
+	if (off_filter_mask &
+	    IFLA_STATS_FILTER_BIT(attr_id_cpu_hit)) {
+		size = rtnl_offload_xstats_get_size_ndo(dev, attr_id_cpu_hit);
+		nla_size += nla_total_size_64bit(size);
+	}
 
-	for (attr_id = IFLA_OFFLOAD_XSTATS_FIRST;
-	     attr_id <= IFLA_OFFLOAD_XSTATS_MAX; attr_id++) {
-		if (!dev->netdev_ops->ndo_has_offload_stats(dev, attr_id))
-			continue;
-		size = rtnl_get_offload_stats_attr_size(attr_id);
+	if (off_filter_mask &
+	    IFLA_STATS_FILTER_BIT(IFLA_OFFLOAD_XSTATS_HW_S_INFO))
+		nla_size += rtnl_offload_xstats_get_size_hw_s_info(dev);
+
+	if (off_filter_mask &
+	    IFLA_STATS_FILTER_BIT(IFLA_OFFLOAD_XSTATS_L3_STATS)) {
+		size = rtnl_offload_xstats_get_size_stats(dev, t_l3);
 		nla_size += nla_total_size_64bit(size);
 	}
 
@@ -5133,11 +5307,21 @@ static int rtnl_get_offload_stats_size(const struct net_device *dev)
 	return nla_size;
 }
 
+struct rtnl_stats_dump_filters {
+	/* mask[0] filters outer attributes. Then individual nests have their
+	 * filtering mask at the index of the nested attribute.
+	 */
+	u32 mask[IFLA_STATS_MAX + 1];
+};
+
 static int rtnl_fill_statsinfo(struct sk_buff *skb, struct net_device *dev,
 			       int type, u32 pid, u32 seq, u32 change,
-			       unsigned int flags, unsigned int filter_mask,
-			       int *idxattr, int *prividx)
+			       unsigned int flags,
+			       const struct rtnl_stats_dump_filters *filters,
+			       int *idxattr, int *prividx,
+			       struct netlink_ext_ack *extack)
 {
+	unsigned int filter_mask = filters->mask[0];
 	struct if_stats_msg *ifsm;
 	struct nlmsghdr *nlh;
 	struct nlattr *attr;
@@ -5163,8 +5347,10 @@ static int rtnl_fill_statsinfo(struct sk_buff *skb, struct net_device *dev,
 		attr = nla_reserve_64bit(skb, IFLA_STATS_LINK_64,
 					 sizeof(struct rtnl_link_stats64),
 					 IFLA_STATS_UNSPEC);
-		if (!attr)
+		if (!attr) {
+			err = -EMSGSIZE;
 			goto nla_put_failure;
+		}
 
 		sp = nla_data(attr);
 		dev_get_stats(dev, sp);
@@ -5177,8 +5363,10 @@ static int rtnl_fill_statsinfo(struct sk_buff *skb, struct net_device *dev,
 			*idxattr = IFLA_STATS_LINK_XSTATS;
 			attr = nla_nest_start_noflag(skb,
 						     IFLA_STATS_LINK_XSTATS);
-			if (!attr)
+			if (!attr) {
+				err = -EMSGSIZE;
 				goto nla_put_failure;
+			}
 
 			err = ops->fill_linkxstats(skb, dev, prividx, *idxattr);
 			nla_nest_end(skb, attr);
@@ -5200,8 +5388,10 @@ static int rtnl_fill_statsinfo(struct sk_buff *skb, struct net_device *dev,
 			*idxattr = IFLA_STATS_LINK_XSTATS_SLAVE;
 			attr = nla_nest_start_noflag(skb,
 						     IFLA_STATS_LINK_XSTATS_SLAVE);
-			if (!attr)
+			if (!attr) {
+				err = -EMSGSIZE;
 				goto nla_put_failure;
+			}
 
 			err = ops->fill_linkxstats(skb, dev, prividx, *idxattr);
 			nla_nest_end(skb, attr);
@@ -5213,13 +5403,19 @@ static int rtnl_fill_statsinfo(struct sk_buff *skb, struct net_device *dev,
 
 	if (stats_attr_valid(filter_mask, IFLA_STATS_LINK_OFFLOAD_XSTATS,
 			     *idxattr)) {
+		u32 off_filter_mask;
+
+		off_filter_mask = filters->mask[IFLA_STATS_LINK_OFFLOAD_XSTATS];
 		*idxattr = IFLA_STATS_LINK_OFFLOAD_XSTATS;
 		attr = nla_nest_start_noflag(skb,
 					     IFLA_STATS_LINK_OFFLOAD_XSTATS);
-		if (!attr)
+		if (!attr) {
+			err = -EMSGSIZE;
 			goto nla_put_failure;
+		}
 
-		err = rtnl_get_offload_stats(skb, dev, prividx);
+		err = rtnl_offload_xstats_fill(skb, dev, prividx,
+					       off_filter_mask, extack);
 		if (err == -ENODATA)
 			nla_nest_cancel(skb, attr);
 		else
@@ -5235,8 +5431,10 @@ static int rtnl_fill_statsinfo(struct sk_buff *skb, struct net_device *dev,
 
 		*idxattr = IFLA_STATS_AF_SPEC;
 		attr = nla_nest_start_noflag(skb, IFLA_STATS_AF_SPEC);
-		if (!attr)
+		if (!attr) {
+			err = -EMSGSIZE;
 			goto nla_put_failure;
+		}
 
 		rcu_read_lock();
 		list_for_each_entry_rcu(af_ops, &rtnl_af_ops, list) {
@@ -5280,13 +5478,14 @@ nla_put_failure:
 	else
 		nlmsg_end(skb, nlh);
 
-	return -EMSGSIZE;
+	return err;
 }
 
 static size_t if_nlmsg_stats_size(const struct net_device *dev,
-				  u32 filter_mask)
+				  const struct rtnl_stats_dump_filters *filters)
 {
 	size_t size = NLMSG_ALIGN(sizeof(struct if_stats_msg));
+	unsigned int filter_mask = filters->mask[0];
 
 	if (stats_attr_valid(filter_mask, IFLA_STATS_LINK_64, 0))
 		size += nla_total_size_64bit(sizeof(struct rtnl_link_stats64));
@@ -5322,8 +5521,12 @@ static size_t if_nlmsg_stats_size(const struct net_device *dev,
 		}
 	}
 
-	if (stats_attr_valid(filter_mask, IFLA_STATS_LINK_OFFLOAD_XSTATS, 0))
-		size += rtnl_get_offload_stats_size(dev);
+	if (stats_attr_valid(filter_mask, IFLA_STATS_LINK_OFFLOAD_XSTATS, 0)) {
+		u32 off_filter_mask;
+
+		off_filter_mask = filters->mask[IFLA_STATS_LINK_OFFLOAD_XSTATS];
+		size += rtnl_offload_xstats_get_size(dev, off_filter_mask);
+	}
 
 	if (stats_attr_valid(filter_mask, IFLA_STATS_AF_SPEC, 0)) {
 		struct rtnl_af_ops *af_ops;
@@ -5347,6 +5550,79 @@ static size_t if_nlmsg_stats_size(const struct net_device *dev,
 	return size;
 }
 
+#define RTNL_STATS_OFFLOAD_XSTATS_VALID ((1 << __IFLA_OFFLOAD_XSTATS_MAX) - 1)
+
+static const struct nla_policy
+rtnl_stats_get_policy_filters[IFLA_STATS_MAX + 1] = {
+	[IFLA_STATS_LINK_OFFLOAD_XSTATS] =
+		    NLA_POLICY_MASK(NLA_U32, RTNL_STATS_OFFLOAD_XSTATS_VALID),
+};
+
+static const struct nla_policy
+rtnl_stats_get_policy[IFLA_STATS_GETSET_MAX + 1] = {
+	[IFLA_STATS_GET_FILTERS] =
+		    NLA_POLICY_NESTED(rtnl_stats_get_policy_filters),
+};
+
+static const struct nla_policy
+ifla_stats_set_policy[IFLA_STATS_GETSET_MAX + 1] = {
+	[IFLA_STATS_SET_OFFLOAD_XSTATS_L3_STATS] = NLA_POLICY_MAX(NLA_U8, 1),
+};
+
+static int rtnl_stats_get_parse_filters(struct nlattr *ifla_filters,
+					struct rtnl_stats_dump_filters *filters,
+					struct netlink_ext_ack *extack)
+{
+	struct nlattr *tb[IFLA_STATS_MAX + 1];
+	int err;
+	int at;
+
+	err = nla_parse_nested(tb, IFLA_STATS_MAX, ifla_filters,
+			       rtnl_stats_get_policy_filters, extack);
+	if (err < 0)
+		return err;
+
+	for (at = 1; at <= IFLA_STATS_MAX; at++) {
+		if (tb[at]) {
+			if (!(filters->mask[0] & IFLA_STATS_FILTER_BIT(at))) {
+				NL_SET_ERR_MSG(extack, "Filtered attribute not enabled in filter_mask");
+				return -EINVAL;
+			}
+			filters->mask[at] = nla_get_u32(tb[at]);
+		}
+	}
+
+	return 0;
+}
+
+static int rtnl_stats_get_parse(const struct nlmsghdr *nlh,
+				u32 filter_mask,
+				struct rtnl_stats_dump_filters *filters,
+				struct netlink_ext_ack *extack)
+{
+	struct nlattr *tb[IFLA_STATS_GETSET_MAX + 1];
+	int err;
+	int i;
+
+	filters->mask[0] = filter_mask;
+	for (i = 1; i < ARRAY_SIZE(filters->mask); i++)
+		filters->mask[i] = -1U;
+
+	err = nlmsg_parse(nlh, sizeof(struct if_stats_msg), tb,
+			  IFLA_STATS_GETSET_MAX, rtnl_stats_get_policy, extack);
+	if (err < 0)
+		return err;
+
+	if (tb[IFLA_STATS_GET_FILTERS]) {
+		err = rtnl_stats_get_parse_filters(tb[IFLA_STATS_GET_FILTERS],
+						   filters, extack);
+		if (err)
+			return err;
+	}
+
+	return 0;
+}
+
 static int rtnl_valid_stats_req(const struct nlmsghdr *nlh, bool strict_check,
 				bool is_dump, struct netlink_ext_ack *extack)
 {
@@ -5369,10 +5645,6 @@ static int rtnl_valid_stats_req(const struct nlmsghdr *nlh, bool strict_check,
 		NL_SET_ERR_MSG(extack, "Invalid values in header for stats dump request");
 		return -EINVAL;
 	}
-	if (nlmsg_attrlen(nlh, sizeof(*ifsm))) {
-		NL_SET_ERR_MSG(extack, "Invalid attributes after stats header");
-		return -EINVAL;
-	}
 	if (ifsm->filter_mask >= IFLA_STATS_FILTER_BIT(IFLA_STATS_MAX + 1)) {
 		NL_SET_ERR_MSG(extack, "Invalid stats requested through filter mask");
 		return -EINVAL;
@@ -5384,12 +5656,12 @@ static int rtnl_valid_stats_req(const struct nlmsghdr *nlh, bool strict_check,
 static int rtnl_stats_get(struct sk_buff *skb, struct nlmsghdr *nlh,
 			  struct netlink_ext_ack *extack)
 {
+	struct rtnl_stats_dump_filters filters;
 	struct net *net = sock_net(skb->sk);
 	struct net_device *dev = NULL;
 	int idxattr = 0, prividx = 0;
 	struct if_stats_msg *ifsm;
 	struct sk_buff *nskb;
-	u32 filter_mask;
 	int err;
 
 	err = rtnl_valid_stats_req(nlh, netlink_strict_get_check(skb),
@@ -5406,19 +5678,22 @@ static int rtnl_stats_get(struct sk_buff *skb, struct nlmsghdr *nlh,
 	if (!dev)
 		return -ENODEV;
 
-	filter_mask = ifsm->filter_mask;
-	if (!filter_mask) {
+	if (!ifsm->filter_mask) {
 		NL_SET_ERR_MSG(extack, "Filter mask must be set for stats get");
 		return -EINVAL;
 	}
 
-	nskb = nlmsg_new(if_nlmsg_stats_size(dev, filter_mask), GFP_KERNEL);
+	err = rtnl_stats_get_parse(nlh, ifsm->filter_mask, &filters, extack);
+	if (err)
+		return err;
+
+	nskb = nlmsg_new(if_nlmsg_stats_size(dev, &filters), GFP_KERNEL);
 	if (!nskb)
 		return -ENOBUFS;
 
 	err = rtnl_fill_statsinfo(nskb, dev, RTM_NEWSTATS,
 				  NETLINK_CB(skb).portid, nlh->nlmsg_seq, 0,
-				  0, filter_mask, &idxattr, &prividx);
+				  0, &filters, &idxattr, &prividx, extack);
 	if (err < 0) {
 		/* -EMSGSIZE implies BUG in if_nlmsg_stats_size */
 		WARN_ON(err == -EMSGSIZE);
@@ -5434,12 +5709,12 @@ static int rtnl_stats_dump(struct sk_buff *skb, struct netlink_callback *cb)
 {
 	struct netlink_ext_ack *extack = cb->extack;
 	int h, s_h, err, s_idx, s_idxattr, s_prividx;
+	struct rtnl_stats_dump_filters filters;
 	struct net *net = sock_net(skb->sk);
 	unsigned int flags = NLM_F_MULTI;
 	struct if_stats_msg *ifsm;
 	struct hlist_head *head;
 	struct net_device *dev;
-	u32 filter_mask = 0;
 	int idx = 0;
 
 	s_h = cb->args[0];
@@ -5454,12 +5729,16 @@ static int rtnl_stats_dump(struct sk_buff *skb, struct netlink_callback *cb)
 		return err;
 
 	ifsm = nlmsg_data(cb->nlh);
-	filter_mask = ifsm->filter_mask;
-	if (!filter_mask) {
+	if (!ifsm->filter_mask) {
 		NL_SET_ERR_MSG(extack, "Filter mask must be set for stats dump");
 		return -EINVAL;
 	}
 
+	err = rtnl_stats_get_parse(cb->nlh, ifsm->filter_mask, &filters,
+				   extack);
+	if (err)
+		return err;
+
 	for (h = s_h; h < NETDEV_HASHENTRIES; h++, s_idx = 0) {
 		idx = 0;
 		head = &net->dev_index_head[h];
@@ -5469,8 +5748,9 @@ static int rtnl_stats_dump(struct sk_buff *skb, struct netlink_callback *cb)
 			err = rtnl_fill_statsinfo(skb, dev, RTM_NEWSTATS,
 						  NETLINK_CB(cb->skb).portid,
 						  cb->nlh->nlmsg_seq, 0,
-						  flags, filter_mask,
-						  &s_idxattr, &s_prividx);
+						  flags, &filters,
+						  &s_idxattr, &s_prividx,
+						  extack);
 			/* If we ran out of room on the first message,
 			 * we're in trouble
 			 */
@@ -5494,6 +5774,107 @@ out:
 	return skb->len;
 }
 
+void rtnl_offload_xstats_notify(struct net_device *dev)
+{
+	struct rtnl_stats_dump_filters response_filters = {};
+	struct net *net = dev_net(dev);
+	int idxattr = 0, prividx = 0;
+	struct sk_buff *skb;
+	int err = -ENOBUFS;
+
+	ASSERT_RTNL();
+
+	response_filters.mask[0] |=
+		IFLA_STATS_FILTER_BIT(IFLA_STATS_LINK_OFFLOAD_XSTATS);
+	response_filters.mask[IFLA_STATS_LINK_OFFLOAD_XSTATS] |=
+		IFLA_STATS_FILTER_BIT(IFLA_OFFLOAD_XSTATS_HW_S_INFO);
+
+	skb = nlmsg_new(if_nlmsg_stats_size(dev, &response_filters),
+			GFP_KERNEL);
+	if (!skb)
+		goto errout;
+
+	err = rtnl_fill_statsinfo(skb, dev, RTM_NEWSTATS, 0, 0, 0, 0,
+				  &response_filters, &idxattr, &prividx, NULL);
+	if (err < 0) {
+		kfree_skb(skb);
+		goto errout;
+	}
+
+	rtnl_notify(skb, net, 0, RTNLGRP_STATS, NULL, GFP_KERNEL);
+	return;
+
+errout:
+	rtnl_set_sk_err(net, RTNLGRP_STATS, err);
+}
+EXPORT_SYMBOL(rtnl_offload_xstats_notify);
+
+static int rtnl_stats_set(struct sk_buff *skb, struct nlmsghdr *nlh,
+			  struct netlink_ext_ack *extack)
+{
+	enum netdev_offload_xstats_type t_l3 = NETDEV_OFFLOAD_XSTATS_TYPE_L3;
+	struct rtnl_stats_dump_filters response_filters = {};
+	struct nlattr *tb[IFLA_STATS_GETSET_MAX + 1];
+	struct net *net = sock_net(skb->sk);
+	struct net_device *dev = NULL;
+	struct if_stats_msg *ifsm;
+	bool notify = false;
+	int err;
+
+	err = rtnl_valid_stats_req(nlh, netlink_strict_get_check(skb),
+				   false, extack);
+	if (err)
+		return err;
+
+	ifsm = nlmsg_data(nlh);
+	if (ifsm->family != AF_UNSPEC) {
+		NL_SET_ERR_MSG(extack, "Address family should be AF_UNSPEC");
+		return -EINVAL;
+	}
+
+	if (ifsm->ifindex > 0)
+		dev = __dev_get_by_index(net, ifsm->ifindex);
+	else
+		return -EINVAL;
+
+	if (!dev)
+		return -ENODEV;
+
+	if (ifsm->filter_mask) {
+		NL_SET_ERR_MSG(extack, "Filter mask must be 0 for stats set");
+		return -EINVAL;
+	}
+
+	err = nlmsg_parse(nlh, sizeof(*ifsm), tb, IFLA_STATS_GETSET_MAX,
+			  ifla_stats_set_policy, extack);
+	if (err < 0)
+		return err;
+
+	if (tb[IFLA_STATS_SET_OFFLOAD_XSTATS_L3_STATS]) {
+		u8 req = nla_get_u8(tb[IFLA_STATS_SET_OFFLOAD_XSTATS_L3_STATS]);
+
+		if (req)
+			err = netdev_offload_xstats_enable(dev, t_l3, extack);
+		else
+			err = netdev_offload_xstats_disable(dev, t_l3);
+
+		if (!err)
+			notify = true;
+		else if (err != -EALREADY)
+			return err;
+
+		response_filters.mask[0] |=
+			IFLA_STATS_FILTER_BIT(IFLA_STATS_LINK_OFFLOAD_XSTATS);
+		response_filters.mask[IFLA_STATS_LINK_OFFLOAD_XSTATS] |=
+			IFLA_STATS_FILTER_BIT(IFLA_OFFLOAD_XSTATS_HW_S_INFO);
+	}
+
+	if (notify)
+		rtnl_offload_xstats_notify(dev);
+
+	return 0;
+}
+
 /* Process one rtnetlink message. */
 
 static int rtnetlink_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh,
@@ -5719,4 +6100,5 @@ void __init rtnetlink_init(void)
 
 	rtnl_register(PF_UNSPEC, RTM_GETSTATS, rtnl_stats_get, rtnl_stats_dump,
 		      0);
+	rtnl_register(PF_UNSPEC, RTM_SETSTATS, rtnl_stats_set, NULL, 0);
 }
diff --git a/security/selinux/nlmsgtab.c b/security/selinux/nlmsgtab.c
index 6ad3ee02e023..d8ceee9e0d6f 100644
--- a/security/selinux/nlmsgtab.c
+++ b/security/selinux/nlmsgtab.c
@@ -76,6 +76,7 @@ static const struct nlmsg_perm nlmsg_route_perms[] =
 	{ RTM_GETNSID,		NETLINK_ROUTE_SOCKET__NLMSG_READ  },
 	{ RTM_NEWSTATS,		NETLINK_ROUTE_SOCKET__NLMSG_READ },
 	{ RTM_GETSTATS,		NETLINK_ROUTE_SOCKET__NLMSG_READ  },
+	{ RTM_SETSTATS,		NETLINK_ROUTE_SOCKET__NLMSG_WRITE },
 	{ RTM_NEWCACHEREPORT,	NETLINK_ROUTE_SOCKET__NLMSG_READ },
 	{ RTM_NEWCHAIN,		NETLINK_ROUTE_SOCKET__NLMSG_WRITE },
 	{ RTM_DELCHAIN,		NETLINK_ROUTE_SOCKET__NLMSG_WRITE },
diff --git a/tools/testing/selftests/net/forwarding/hw_stats_l3.sh b/tools/testing/selftests/net/forwarding/hw_stats_l3.sh
new file mode 100755
index 000000000000..1c11c4256d06
--- /dev/null
+++ b/tools/testing/selftests/net/forwarding/hw_stats_l3.sh
@@ -0,0 +1,332 @@
+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+
+# +--------------------+                     +----------------------+
+# | H1                 |                     |                   H2 |
+# |                    |                     |                      |
+# |          $h1.200 + |                     | + $h2.200            |
+# |     192.0.2.1/28 | |                     | | 192.0.2.18/28      |
+# | 2001:db8:1::1/64 | |                     | | 2001:db8:2::1/64   |
+# |                  | |                     | |                    |
+# |              $h1 + |                     | + $h2                |
+# |                  | |                     | |                    |
+# +------------------|-+                     +-|--------------------+
+#                    |                         |
+# +------------------|-------------------------|--------------------+
+# | SW               |                         |                    |
+# |                  |                         |                    |
+# |             $rp1 +                         + $rp2               |
+# |                  |                         |                    |
+# |         $rp1.200 +                         + $rp2.200           |
+# |     192.0.2.2/28                             192.0.2.17/28      |
+# | 2001:db8:1::2/64                             2001:db8:2::2/64   |
+# |                                                                 |
+# +-----------------------------------------------------------------+
+
+ALL_TESTS="
+	ping_ipv4
+	ping_ipv6
+	test_stats_rx_ipv4
+	test_stats_tx_ipv4
+	test_stats_rx_ipv6
+	test_stats_tx_ipv6
+	respin_enablement
+	test_stats_rx_ipv4
+	test_stats_tx_ipv4
+	test_stats_rx_ipv6
+	test_stats_tx_ipv6
+	reapply_config
+	ping_ipv4
+	ping_ipv6
+	test_stats_rx_ipv4
+	test_stats_tx_ipv4
+	test_stats_rx_ipv6
+	test_stats_tx_ipv6
+	test_stats_report_rx
+	test_stats_report_tx
+	test_destroy_enabled
+	test_double_enable
+"
+NUM_NETIFS=4
+source lib.sh
+
+h1_create()
+{
+	simple_if_init $h1
+	vlan_create $h1 200 v$h1 192.0.2.1/28 2001:db8:1::1/64
+	ip route add 192.0.2.16/28 vrf v$h1 nexthop via 192.0.2.2
+	ip -6 route add 2001:db8:2::/64 vrf v$h1 nexthop via 2001:db8:1::2
+}
+
+h1_destroy()
+{
+	ip -6 route del 2001:db8:2::/64 vrf v$h1 nexthop via 2001:db8:1::2
+	ip route del 192.0.2.16/28 vrf v$h1 nexthop via 192.0.2.2
+	vlan_destroy $h1 200
+	simple_if_fini $h1
+}
+
+h2_create()
+{
+	simple_if_init $h2
+	vlan_create $h2 200 v$h2 192.0.2.18/28 2001:db8:2::1/64
+	ip route add 192.0.2.0/28 vrf v$h2 nexthop via 192.0.2.17
+	ip -6 route add 2001:db8:1::/64 vrf v$h2 nexthop via 2001:db8:2::2
+}
+
+h2_destroy()
+{
+	ip -6 route del 2001:db8:1::/64 vrf v$h2 nexthop via 2001:db8:2::2
+	ip route del 192.0.2.0/28 vrf v$h2 nexthop via 192.0.2.17
+	vlan_destroy $h2 200
+	simple_if_fini $h2
+}
+
+router_rp1_200_create()
+{
+	ip link add name $rp1.200 up \
+		link $rp1 addrgenmode eui64 type vlan id 200
+	ip address add dev $rp1.200 192.0.2.2/28
+	ip address add dev $rp1.200 2001:db8:1::2/64
+	ip stats set dev $rp1.200 l3_stats on
+}
+
+router_rp1_200_destroy()
+{
+	ip stats set dev $rp1.200 l3_stats off
+	ip address del dev $rp1.200 2001:db8:1::2/64
+	ip address del dev $rp1.200 192.0.2.2/28
+	ip link del dev $rp1.200
+}
+
+router_create()
+{
+	ip link set dev $rp1 up
+	router_rp1_200_create
+
+	ip link set dev $rp2 up
+	vlan_create $rp2 200 "" 192.0.2.17/28 2001:db8:2::2/64
+}
+
+router_destroy()
+{
+	vlan_destroy $rp2 200
+	ip link set dev $rp2 down
+
+	router_rp1_200_destroy
+	ip link set dev $rp1 down
+}
+
+setup_prepare()
+{
+	h1=${NETIFS[p1]}
+	rp1=${NETIFS[p2]}
+
+	rp2=${NETIFS[p3]}
+	h2=${NETIFS[p4]}
+
+	rp1mac=$(mac_get $rp1)
+	rp2mac=$(mac_get $rp2)
+
+	vrf_prepare
+
+	h1_create
+	h2_create
+
+	router_create
+
+	forwarding_enable
+}
+
+cleanup()
+{
+	pre_cleanup
+
+	forwarding_restore
+
+	router_destroy
+
+	h2_destroy
+	h1_destroy
+
+	vrf_cleanup
+}
+
+ping_ipv4()
+{
+	ping_test $h1.200 192.0.2.18 " IPv4"
+}
+
+ping_ipv6()
+{
+	ping_test $h1.200 2001:db8:2::1 " IPv6"
+}
+
+get_l3_stat()
+{
+	local selector=$1; shift
+
+	ip -j stats show dev $rp1.200 group offload subgroup l3_stats |
+		  jq '.[0].stats64.'$selector
+}
+
+send_packets_rx_ipv4()
+{
+	# Send 21 packets instead of 20, because the first one might trap and go
+	# through the SW datapath, which might not bump the HW counter.
+	$MZ $h1.200 -c 21 -d 20msec -p 100 \
+	    -a own -b $rp1mac -A 192.0.2.1 -B 192.0.2.18 \
+	    -q -t udp sp=54321,dp=12345
+}
+
+send_packets_rx_ipv6()
+{
+	$MZ $h1.200 -6 -c 21 -d 20msec -p 100 \
+	    -a own -b $rp1mac -A 2001:db8:1::1 -B 2001:db8:2::1 \
+	    -q -t udp sp=54321,dp=12345
+}
+
+send_packets_tx_ipv4()
+{
+	$MZ $h2.200 -c 21 -d 20msec -p 100 \
+	    -a own -b $rp2mac -A 192.0.2.18 -B 192.0.2.1 \
+	    -q -t udp sp=54321,dp=12345
+}
+
+send_packets_tx_ipv6()
+{
+	$MZ $h2.200 -6 -c 21 -d 20msec -p 100 \
+	    -a own -b $rp2mac -A 2001:db8:2::1 -B 2001:db8:1::1 \
+	    -q -t udp sp=54321,dp=12345
+}
+
+___test_stats()
+{
+	local dir=$1; shift
+	local prot=$1; shift
+
+	local a
+	local b
+
+	a=$(get_l3_stat ${dir}.packets)
+	send_packets_${dir}_${prot}
+	"$@"
+	b=$(busywait "$TC_HIT_TIMEOUT" until_counter_is ">= $a + 20" \
+		       get_l3_stat ${dir}.packets)
+	check_err $? "Traffic not reflected in the counter: $a -> $b"
+}
+
+__test_stats()
+{
+	local dir=$1; shift
+	local prot=$1; shift
+
+	RET=0
+	___test_stats "$dir" "$prot"
+	log_test "Test $dir packets: $prot"
+}
+
+test_stats_rx_ipv4()
+{
+	__test_stats rx ipv4
+}
+
+test_stats_tx_ipv4()
+{
+	__test_stats tx ipv4
+}
+
+test_stats_rx_ipv6()
+{
+	__test_stats rx ipv6
+}
+
+test_stats_tx_ipv6()
+{
+	__test_stats tx ipv6
+}
+
+# Make sure everything works well even after stats have been disabled and
+# reenabled on the same device without touching the L3 configuration.
+respin_enablement()
+{
+	log_info "Turning stats off and on again"
+	ip stats set dev $rp1.200 l3_stats off
+	ip stats set dev $rp1.200 l3_stats on
+}
+
+# For the initial run, l3_stats is enabled on a completely set up netdevice. Now
+# do it the other way around: enabling the L3 stats on an L2 netdevice, and only
+# then apply the L3 configuration.
+reapply_config()
+{
+	log_info "Reapplying configuration"
+
+	router_rp1_200_destroy
+
+	ip link add name $rp1.200 link $rp1 addrgenmode none type vlan id 200
+	ip stats set dev $rp1.200 l3_stats on
+	ip link set dev $rp1.200 up addrgenmode eui64
+	ip address add dev $rp1.200 192.0.2.2/28
+	ip address add dev $rp1.200 2001:db8:1::2/64
+}
+
+__test_stats_report()
+{
+	local dir=$1; shift
+	local prot=$1; shift
+
+	local a
+	local b
+
+	RET=0
+
+	a=$(get_l3_stat ${dir}.packets)
+	send_packets_${dir}_${prot}
+	ip address flush dev $rp1.200
+	b=$(busywait "$TC_HIT_TIMEOUT" until_counter_is ">= $a + 20" \
+		       get_l3_stat ${dir}.packets)
+	check_err $? "Traffic not reflected in the counter: $a -> $b"
+	log_test "Test ${dir} packets: stats pushed on loss of L3"
+
+	ip stats set dev $rp1.200 l3_stats off
+	ip link del dev $rp1.200
+	router_rp1_200_create
+}
+
+test_stats_report_rx()
+{
+	__test_stats_report rx ipv4
+}
+
+test_stats_report_tx()
+{
+	__test_stats_report tx ipv4
+}
+
+test_destroy_enabled()
+{
+	RET=0
+
+	ip link del dev $rp1.200
+	router_rp1_200_create
+
+	log_test "Destroy l3_stats-enabled netdev"
+}
+
+test_double_enable()
+{
+	RET=0
+	___test_stats rx ipv4 \
+		ip stats set dev $rp1.200 l3_stats on
+	log_test "Test stat retention across a spurious enablement"
+}
+
+trap cleanup EXIT
+
+setup_prepare
+setup_wait
+
+tests_run
+
+exit $EXIT_STATUS