Merging upstream version 6.10.3.

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 1228 insertions, 231 deletions

diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index 290bec2926..5161e7efda 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -24,6 +24,7 @@
 #include <net/xdp.h>
 #include <net/net_failover.h>
 #include <net/netdev_rx_queue.h>
+#include <net/netdev_queues.h>
 
 static int napi_weight = NAPI_POLL_WEIGHT;
 module_param(napi_weight, int, 0444);
@@ -78,6 +79,7 @@ static const unsigned long guest_offloads[] = {
 struct virtnet_stat_desc {
 	char desc[ETH_GSTRING_LEN];
 	size_t offset;
+	size_t qstat_offset;
 };
 
 struct virtnet_sq_free_stats {
@@ -93,6 +95,8 @@ struct virtnet_sq_stats {
 	u64_stats_t xdp_tx_drops;
 	u64_stats_t kicks;
 	u64_stats_t tx_timeouts;
+	u64_stats_t stop;
+	u64_stats_t wake;
 };
 
 struct virtnet_rq_stats {
@@ -107,31 +111,159 @@ struct virtnet_rq_stats {
 	u64_stats_t kicks;
 };
 
-#define VIRTNET_SQ_STAT(m)	offsetof(struct virtnet_sq_stats, m)
-#define VIRTNET_RQ_STAT(m)	offsetof(struct virtnet_rq_stats, m)
+#define VIRTNET_SQ_STAT(name, m) {name, offsetof(struct virtnet_sq_stats, m), -1}
+#define VIRTNET_RQ_STAT(name, m) {name, offsetof(struct virtnet_rq_stats, m), -1}
+
+#define VIRTNET_SQ_STAT_QSTAT(name, m)				\
+	{							\
+		name,						\
+		offsetof(struct virtnet_sq_stats, m),		\
+		offsetof(struct netdev_queue_stats_tx, m),	\
+	}
+
+#define VIRTNET_RQ_STAT_QSTAT(name, m)				\
+	{							\
+		name,						\
+		offsetof(struct virtnet_rq_stats, m),		\
+		offsetof(struct netdev_queue_stats_rx, m),	\
+	}
 
 static const struct virtnet_stat_desc virtnet_sq_stats_desc[] = {
-	{ "packets",		VIRTNET_SQ_STAT(packets) },
-	{ "bytes",		VIRTNET_SQ_STAT(bytes) },
-	{ "xdp_tx",		VIRTNET_SQ_STAT(xdp_tx) },
-	{ "xdp_tx_drops",	VIRTNET_SQ_STAT(xdp_tx_drops) },
-	{ "kicks",		VIRTNET_SQ_STAT(kicks) },
-	{ "tx_timeouts",	VIRTNET_SQ_STAT(tx_timeouts) },
+	VIRTNET_SQ_STAT("xdp_tx",       xdp_tx),
+	VIRTNET_SQ_STAT("xdp_tx_drops", xdp_tx_drops),
+	VIRTNET_SQ_STAT("kicks",        kicks),
+	VIRTNET_SQ_STAT("tx_timeouts",  tx_timeouts),
 };
 
 static const struct virtnet_stat_desc virtnet_rq_stats_desc[] = {
-	{ "packets",		VIRTNET_RQ_STAT(packets) },
-	{ "bytes",		VIRTNET_RQ_STAT(bytes) },
-	{ "drops",		VIRTNET_RQ_STAT(drops) },
-	{ "xdp_packets",	VIRTNET_RQ_STAT(xdp_packets) },
-	{ "xdp_tx",		VIRTNET_RQ_STAT(xdp_tx) },
-	{ "xdp_redirects",	VIRTNET_RQ_STAT(xdp_redirects) },
-	{ "xdp_drops",		VIRTNET_RQ_STAT(xdp_drops) },
-	{ "kicks",		VIRTNET_RQ_STAT(kicks) },
+	VIRTNET_RQ_STAT("drops",         drops),
+	VIRTNET_RQ_STAT("xdp_packets",   xdp_packets),
+	VIRTNET_RQ_STAT("xdp_tx",        xdp_tx),
+	VIRTNET_RQ_STAT("xdp_redirects", xdp_redirects),
+	VIRTNET_RQ_STAT("xdp_drops",     xdp_drops),
+	VIRTNET_RQ_STAT("kicks",         kicks),
+};
+
+static const struct virtnet_stat_desc virtnet_sq_stats_desc_qstat[] = {
+	VIRTNET_SQ_STAT_QSTAT("packets", packets),
+	VIRTNET_SQ_STAT_QSTAT("bytes",   bytes),
+	VIRTNET_SQ_STAT_QSTAT("stop",	 stop),
+	VIRTNET_SQ_STAT_QSTAT("wake",	 wake),
+};
+
+static const struct virtnet_stat_desc virtnet_rq_stats_desc_qstat[] = {
+	VIRTNET_RQ_STAT_QSTAT("packets", packets),
+	VIRTNET_RQ_STAT_QSTAT("bytes",   bytes),
+};
+
+#define VIRTNET_STATS_DESC_CQ(name) \
+	{#name, offsetof(struct virtio_net_stats_cvq, name), -1}
+
+#define VIRTNET_STATS_DESC_RX(class, name) \
+	{#name, offsetof(struct virtio_net_stats_rx_ ## class, rx_ ## name), -1}
+
+#define VIRTNET_STATS_DESC_TX(class, name) \
+	{#name, offsetof(struct virtio_net_stats_tx_ ## class, tx_ ## name), -1}
+
+
+static const struct virtnet_stat_desc virtnet_stats_cvq_desc[] = {
+	VIRTNET_STATS_DESC_CQ(command_num),
+	VIRTNET_STATS_DESC_CQ(ok_num),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_rx_basic_desc[] = {
+	VIRTNET_STATS_DESC_RX(basic, packets),
+	VIRTNET_STATS_DESC_RX(basic, bytes),
+
+	VIRTNET_STATS_DESC_RX(basic, notifications),
+	VIRTNET_STATS_DESC_RX(basic, interrupts),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_tx_basic_desc[] = {
+	VIRTNET_STATS_DESC_TX(basic, packets),
+	VIRTNET_STATS_DESC_TX(basic, bytes),
+
+	VIRTNET_STATS_DESC_TX(basic, notifications),
+	VIRTNET_STATS_DESC_TX(basic, interrupts),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_rx_csum_desc[] = {
+	VIRTNET_STATS_DESC_RX(csum, needs_csum),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_tx_gso_desc[] = {
+	VIRTNET_STATS_DESC_TX(gso, gso_packets_noseg),
+	VIRTNET_STATS_DESC_TX(gso, gso_bytes_noseg),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_rx_speed_desc[] = {
+	VIRTNET_STATS_DESC_RX(speed, ratelimit_bytes),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_tx_speed_desc[] = {
+	VIRTNET_STATS_DESC_TX(speed, ratelimit_bytes),
+};
+
+#define VIRTNET_STATS_DESC_RX_QSTAT(class, name, qstat_field)			\
+	{									\
+		#name,								\
+		offsetof(struct virtio_net_stats_rx_ ## class, rx_ ## name),	\
+		offsetof(struct netdev_queue_stats_rx, qstat_field),		\
+	}
+
+#define VIRTNET_STATS_DESC_TX_QSTAT(class, name, qstat_field)			\
+	{									\
+		#name,								\
+		offsetof(struct virtio_net_stats_tx_ ## class, tx_ ## name),	\
+		offsetof(struct netdev_queue_stats_tx, qstat_field),		\
+	}
+
+static const struct virtnet_stat_desc virtnet_stats_rx_basic_desc_qstat[] = {
+	VIRTNET_STATS_DESC_RX_QSTAT(basic, drops,         hw_drops),
+	VIRTNET_STATS_DESC_RX_QSTAT(basic, drop_overruns, hw_drop_overruns),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_tx_basic_desc_qstat[] = {
+	VIRTNET_STATS_DESC_TX_QSTAT(basic, drops,          hw_drops),
+	VIRTNET_STATS_DESC_TX_QSTAT(basic, drop_malformed, hw_drop_errors),
 };
 
-#define VIRTNET_SQ_STATS_LEN	ARRAY_SIZE(virtnet_sq_stats_desc)
-#define VIRTNET_RQ_STATS_LEN	ARRAY_SIZE(virtnet_rq_stats_desc)
+static const struct virtnet_stat_desc virtnet_stats_rx_csum_desc_qstat[] = {
+	VIRTNET_STATS_DESC_RX_QSTAT(csum, csum_valid, csum_unnecessary),
+	VIRTNET_STATS_DESC_RX_QSTAT(csum, csum_none,  csum_none),
+	VIRTNET_STATS_DESC_RX_QSTAT(csum, csum_bad,   csum_bad),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_tx_csum_desc_qstat[] = {
+	VIRTNET_STATS_DESC_TX_QSTAT(csum, csum_none,  csum_none),
+	VIRTNET_STATS_DESC_TX_QSTAT(csum, needs_csum, needs_csum),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_rx_gso_desc_qstat[] = {
+	VIRTNET_STATS_DESC_RX_QSTAT(gso, gso_packets,           hw_gro_packets),
+	VIRTNET_STATS_DESC_RX_QSTAT(gso, gso_bytes,             hw_gro_bytes),
+	VIRTNET_STATS_DESC_RX_QSTAT(gso, gso_packets_coalesced, hw_gro_wire_packets),
+	VIRTNET_STATS_DESC_RX_QSTAT(gso, gso_bytes_coalesced,   hw_gro_wire_bytes),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_tx_gso_desc_qstat[] = {
+	VIRTNET_STATS_DESC_TX_QSTAT(gso, gso_packets,        hw_gso_packets),
+	VIRTNET_STATS_DESC_TX_QSTAT(gso, gso_bytes,          hw_gso_bytes),
+	VIRTNET_STATS_DESC_TX_QSTAT(gso, gso_segments,       hw_gso_wire_packets),
+	VIRTNET_STATS_DESC_TX_QSTAT(gso, gso_segments_bytes, hw_gso_wire_bytes),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_rx_speed_desc_qstat[] = {
+	VIRTNET_STATS_DESC_RX_QSTAT(speed, ratelimit_packets, hw_drop_ratelimits),
+};
+
+static const struct virtnet_stat_desc virtnet_stats_tx_speed_desc_qstat[] = {
+	VIRTNET_STATS_DESC_TX_QSTAT(speed, ratelimit_packets, hw_drop_ratelimits),
+};
+
+#define VIRTNET_Q_TYPE_RX 0
+#define VIRTNET_Q_TYPE_TX 1
+#define VIRTNET_Q_TYPE_CQ 2
 
 struct virtnet_interrupt_coalesce {
 	u32 max_packets;
@@ -184,6 +316,9 @@ struct receive_queue {
 	/* Is dynamic interrupt moderation enabled? */
 	bool dim_enabled;
 
+	/* Used to protect dim_enabled and inter_coal */
+	struct mutex dim_lock;
+
 	/* Dynamic Interrupt Moderation */
 	struct dim dim;
 
@@ -213,9 +348,6 @@ struct receive_queue {
 
 	/* Record the last dma info to free after new pages is allocated. */
 	struct virtnet_rq_dma *last_dma;
-
-	/* Do dma by self */
-	bool do_dma;
 };
 
 /* This structure can contain rss message with maximum settings for indirection table and keysize
@@ -240,15 +372,6 @@ struct virtio_net_ctrl_rss {
 struct control_buf {
 	struct virtio_net_ctrl_hdr hdr;
 	virtio_net_ctrl_ack status;
-	struct virtio_net_ctrl_mq mq;
-	u8 promisc;
-	u8 allmulti;
-	__virtio16 vid;
-	__virtio64 offloads;
-	struct virtio_net_ctrl_rss rss;
-	struct virtio_net_ctrl_coal_tx coal_tx;
-	struct virtio_net_ctrl_coal_rx coal_rx;
-	struct virtio_net_ctrl_coal_vq coal_vq;
 };
 
 struct virtnet_info {
@@ -287,10 +410,14 @@ struct virtnet_info {
 	u16 rss_indir_table_size;
 	u32 rss_hash_types_supported;
 	u32 rss_hash_types_saved;
+	struct virtio_net_ctrl_rss rss;
 
 	/* Has control virtqueue */
 	bool has_cvq;
 
+	/* Lock to protect the control VQ */
+	struct mutex cvq_lock;
+
 	/* Host can handle any s/g split between our header and packet data */
 	bool any_header_sg;
 
@@ -340,6 +467,8 @@ struct virtnet_info {
 
 	/* failover when STANDBY feature enabled */
 	struct failover *failover;
+
+	u64 device_stats_cap;
 };
 
 struct padded_vnet_hdr {
@@ -425,6 +554,17 @@ static int rxq2vq(int rxq)
 	return rxq * 2;
 }
 
+static int vq_type(struct virtnet_info *vi, int qid)
+{
+	if (qid == vi->max_queue_pairs * 2)
+		return VIRTNET_Q_TYPE_CQ;
+
+	if (qid % 2)
+		return VIRTNET_Q_TYPE_TX;
+
+	return VIRTNET_Q_TYPE_RX;
+}
+
 static inline struct virtio_net_common_hdr *
 skb_vnet_common_hdr(struct sk_buff *skb)
 {
@@ -603,7 +743,6 @@ static struct sk_buff *page_to_skb(struct virtnet_info *vi,
 
 	shinfo_size = SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
 
-	/* copy small packet so we can reuse these pages */
 	if (!NET_IP_ALIGN && len > GOOD_COPY_LEN && tailroom >= shinfo_size) {
 		skb = virtnet_build_skb(buf, truesize, p - buf, len);
 		if (unlikely(!skb))
@@ -707,7 +846,7 @@ static void *virtnet_rq_get_buf(struct receive_queue *rq, u32 *len, void **ctx)
 	void *buf;
 
 	buf = virtqueue_get_buf_ctx(rq->vq, len, ctx);
-	if (buf && rq->do_dma)
+	if (buf)
 		virtnet_rq_unmap(rq, buf, *len);
 
 	return buf;
@@ -720,11 +859,6 @@ static void virtnet_rq_init_one_sg(struct receive_queue *rq, void *buf, u32 len)
 	u32 offset;
 	void *head;
 
-	if (!rq->do_dma) {
-		sg_init_one(rq->sg, buf, len);
-		return;
-	}
-
 	head = page_address(rq->alloc_frag.page);
 
 	offset = buf - head;
@@ -750,44 +884,42 @@ static void *virtnet_rq_alloc(struct receive_queue *rq, u32 size, gfp_t gfp)
 
 	head = page_address(alloc_frag->page);
 
-	if (rq->do_dma) {
-		dma = head;
-
-		/* new pages */
-		if (!alloc_frag->offset) {
-			if (rq->last_dma) {
-				/* Now, the new page is allocated, the last dma
-				 * will not be used. So the dma can be unmapped
-				 * if the ref is 0.
-				 */
-				virtnet_rq_unmap(rq, rq->last_dma, 0);
-				rq->last_dma = NULL;
-			}
+	dma = head;
 
-			dma->len = alloc_frag->size - sizeof(*dma);
+	/* new pages */
+	if (!alloc_frag->offset) {
+		if (rq->last_dma) {
+			/* Now, the new page is allocated, the last dma
+			 * will not be used. So the dma can be unmapped
+			 * if the ref is 0.
+			 */
+			virtnet_rq_unmap(rq, rq->last_dma, 0);
+			rq->last_dma = NULL;
+		}
 
-			addr = virtqueue_dma_map_single_attrs(rq->vq, dma + 1,
-							      dma->len, DMA_FROM_DEVICE, 0);
-			if (virtqueue_dma_mapping_error(rq->vq, addr))
-				return NULL;
+		dma->len = alloc_frag->size - sizeof(*dma);
 
-			dma->addr = addr;
-			dma->need_sync = virtqueue_dma_need_sync(rq->vq, addr);
+		addr = virtqueue_dma_map_single_attrs(rq->vq, dma + 1,
+						      dma->len, DMA_FROM_DEVICE, 0);
+		if (virtqueue_dma_mapping_error(rq->vq, addr))
+			return NULL;
 
-			/* Add a reference to dma to prevent the entire dma from
-			 * being released during error handling. This reference
-			 * will be freed after the pages are no longer used.
-			 */
-			get_page(alloc_frag->page);
-			dma->ref = 1;
-			alloc_frag->offset = sizeof(*dma);
+		dma->addr = addr;
+		dma->need_sync = virtqueue_dma_need_sync(rq->vq, addr);
 
-			rq->last_dma = dma;
-		}
+		/* Add a reference to dma to prevent the entire dma from
+		 * being released during error handling. This reference
+		 * will be freed after the pages are no longer used.
+		 */
+		get_page(alloc_frag->page);
+		dma->ref = 1;
+		alloc_frag->offset = sizeof(*dma);
 
-		++dma->ref;
+		rq->last_dma = dma;
 	}
 
+	++dma->ref;
+
 	buf = head + alloc_frag->offset;
 
 	get_page(alloc_frag->page);
@@ -804,12 +936,9 @@ static void virtnet_rq_set_premapped(struct virtnet_info *vi)
 	if (!vi->mergeable_rx_bufs && vi->big_packets)
 		return;
 
-	for (i = 0; i < vi->max_queue_pairs; i++) {
-		if (virtqueue_set_dma_premapped(vi->rq[i].vq))
-			continue;
-
-		vi->rq[i].do_dma = true;
-	}
+	for (i = 0; i < vi->max_queue_pairs; i++)
+		/* error should never happen */
+		BUG_ON(virtqueue_set_dma_premapped(vi->rq[i].vq));
 }
 
 static void virtnet_rq_unmap_free_buf(struct virtqueue *vq, void *buf)
@@ -820,7 +949,7 @@ static void virtnet_rq_unmap_free_buf(struct virtqueue *vq, void *buf)
 
 	rq = &vi->rq[i];
 
-	if (rq->do_dma)
+	if (!vi->big_packets || vi->mergeable_rx_bufs)
 		virtnet_rq_unmap(rq, buf, 0);
 
 	virtnet_rq_free_buf(vi, rq, buf);
@@ -875,6 +1004,9 @@ static void check_sq_full_and_disable(struct virtnet_info *vi,
 	 */
 	if (sq->vq->num_free < 2+MAX_SKB_FRAGS) {
 		netif_stop_subqueue(dev, qnum);
+		u64_stats_update_begin(&sq->stats.syncp);
+		u64_stats_inc(&sq->stats.stop);
+		u64_stats_update_end(&sq->stats.syncp);
 		if (use_napi) {
 			if (unlikely(!virtqueue_enable_cb_delayed(sq->vq)))
 				virtqueue_napi_schedule(&sq->napi, sq->vq);
@@ -883,6 +1015,9 @@ static void check_sq_full_and_disable(struct virtnet_info *vi,
 			free_old_xmit(sq, false);
 			if (sq->vq->num_free >= 2+MAX_SKB_FRAGS) {
 				netif_start_subqueue(dev, qnum);
+				u64_stats_update_begin(&sq->stats.syncp);
+				u64_stats_inc(&sq->stats.wake);
+				u64_stats_update_end(&sq->stats.syncp);
 				virtqueue_disable_cb(sq->vq);
 			}
 		}
@@ -1899,8 +2034,7 @@ static int add_recvbuf_small(struct virtnet_info *vi, struct receive_queue *rq,
 
 	err = virtqueue_add_inbuf_ctx(rq->vq, rq->sg, 1, buf, ctx, gfp);
 	if (err < 0) {
-		if (rq->do_dma)
-			virtnet_rq_unmap(rq, buf, 0);
+		virtnet_rq_unmap(rq, buf, 0);
 		put_page(virt_to_head_page(buf));
 	}
 
@@ -2014,8 +2148,7 @@ static int add_recvbuf_mergeable(struct virtnet_info *vi,
 	ctx = mergeable_len_to_ctx(len + room, headroom);
 	err = virtqueue_add_inbuf_ctx(rq->vq, rq->sg, 1, buf, ctx, gfp);
 	if (err < 0) {
-		if (rq->do_dma)
-			virtnet_rq_unmap(rq, buf, 0);
+		virtnet_rq_unmap(rq, buf, 0);
 		put_page(virt_to_head_page(buf));
 	}
 
@@ -2146,7 +2279,7 @@ static int virtnet_receive(struct receive_queue *rq, int budget,
 		}
 	} else {
 		while (packets < budget &&
-		       (buf = virtnet_rq_get_buf(rq, &len, NULL)) != NULL) {
+		       (buf = virtqueue_get_buf(rq->vq, &len)) != NULL) {
 			receive_buf(vi, rq, buf, len, NULL, xdp_xmit, &stats);
 			packets++;
 		}
@@ -2163,7 +2296,7 @@ static int virtnet_receive(struct receive_queue *rq, int budget,
 
 	u64_stats_set(&stats.packets, packets);
 	u64_stats_update_begin(&rq->stats.syncp);
-	for (i = 0; i < VIRTNET_RQ_STATS_LEN; i++) {
+	for (i = 0; i < ARRAY_SIZE(virtnet_rq_stats_desc); i++) {
 		size_t offset = virtnet_rq_stats_desc[i].offset;
 		u64_stats_t *item, *src;
 
@@ -2171,12 +2304,16 @@ static int virtnet_receive(struct receive_queue *rq, int budget,
 		src = (u64_stats_t *)((u8 *)&stats + offset);
 		u64_stats_add(item, u64_stats_read(src));
 	}
+
+	u64_stats_add(&rq->stats.packets, u64_stats_read(&stats.packets));
+	u64_stats_add(&rq->stats.bytes, u64_stats_read(&stats.bytes));
+
 	u64_stats_update_end(&rq->stats.syncp);
 
 	return packets;
 }
 
-static void virtnet_poll_cleantx(struct receive_queue *rq)
+static void virtnet_poll_cleantx(struct receive_queue *rq, int budget)
 {
 	struct virtnet_info *vi = rq->vq->vdev->priv;
 	unsigned int index = vq2rxq(rq->vq);
@@ -2194,11 +2331,17 @@ static void virtnet_poll_cleantx(struct receive_queue *rq)
 
 		do {
 			virtqueue_disable_cb(sq->vq);
-			free_old_xmit(sq, true);
+			free_old_xmit(sq, !!budget);
 		} while (unlikely(!virtqueue_enable_cb_delayed(sq->vq)));
 
-		if (sq->vq->num_free >= 2 + MAX_SKB_FRAGS)
+		if (sq->vq->num_free >= 2 + MAX_SKB_FRAGS) {
+			if (netif_tx_queue_stopped(txq)) {
+				u64_stats_update_begin(&sq->stats.syncp);
+				u64_stats_inc(&sq->stats.wake);
+				u64_stats_update_end(&sq->stats.syncp);
+			}
 			netif_tx_wake_queue(txq);
+		}
 
 		__netif_tx_unlock(txq);
 	}
@@ -2232,7 +2375,7 @@ static int virtnet_poll(struct napi_struct *napi, int budget)
 	unsigned int xdp_xmit = 0;
 	bool napi_complete;
 
-	virtnet_poll_cleantx(rq);
+	virtnet_poll_cleantx(rq, budget);
 
 	received = virtnet_receive(rq, budget, &xdp_xmit);
 	rq->packets_in_napi += received;
@@ -2243,6 +2386,10 @@ static int virtnet_poll(struct napi_struct *napi, int budget)
 	/* Out of packets? */
 	if (received < budget) {
 		napi_complete = virtqueue_napi_complete(napi, rq->vq, received);
+		/* Intentionally not taking dim_lock here. This may result in a
+		 * spurious net_dim call. But if that happens virtnet_rx_dim_work
+		 * will not act on the scheduled work.
+		 */
 		if (napi_complete && rq->dim_enabled)
 			virtnet_rx_dim_update(vi, rq);
 	}
@@ -2342,10 +2489,16 @@ static int virtnet_poll_tx(struct napi_struct *napi, int budget)
 	txq = netdev_get_tx_queue(vi->dev, index);
 	__netif_tx_lock(txq, raw_smp_processor_id());
 	virtqueue_disable_cb(sq->vq);
-	free_old_xmit(sq, true);
+	free_old_xmit(sq, !!budget);
 
-	if (sq->vq->num_free >= 2 + MAX_SKB_FRAGS)
+	if (sq->vq->num_free >= 2 + MAX_SKB_FRAGS) {
+		if (netif_tx_queue_stopped(txq)) {
+			u64_stats_update_begin(&sq->stats.syncp);
+			u64_stats_inc(&sq->stats.wake);
+			u64_stats_update_end(&sq->stats.syncp);
+		}
 		netif_tx_wake_queue(txq);
+	}
 
 	opaque = virtqueue_enable_cb_prepare(sq->vq);
 
@@ -2545,16 +2698,19 @@ static int virtnet_tx_resize(struct virtnet_info *vi,
  * supported by the hypervisor, as indicated by feature bits, should
  * never fail unless improperly formatted.
  */
-static bool virtnet_send_command(struct virtnet_info *vi, u8 class, u8 cmd,
-				 struct scatterlist *out)
+static bool virtnet_send_command_reply(struct virtnet_info *vi, u8 class, u8 cmd,
+				       struct scatterlist *out,
+				       struct scatterlist *in)
 {
-	struct scatterlist *sgs[4], hdr, stat;
-	unsigned out_num = 0, tmp;
+	struct scatterlist *sgs[5], hdr, stat;
+	u32 out_num = 0, tmp, in_num = 0;
+	bool ok;
 	int ret;
 
 	/* Caller should know better */
 	BUG_ON(!virtio_has_feature(vi->vdev, VIRTIO_NET_F_CTRL_VQ));
 
+	mutex_lock(&vi->cvq_lock);
 	vi->ctrl->status = ~0;
 	vi->ctrl->hdr.class = class;
 	vi->ctrl->hdr.cmd = cmd;
@@ -2567,18 +2723,22 @@ static bool virtnet_send_command(struct virtnet_info *vi, u8 class, u8 cmd,
 
 	/* Add return status. */
 	sg_init_one(&stat, &vi->ctrl->status, sizeof(vi->ctrl->status));
-	sgs[out_num] = &stat;
+	sgs[out_num + in_num++] = &stat;
+
+	if (in)
+		sgs[out_num + in_num++] = in;
 
-	BUG_ON(out_num + 1 > ARRAY_SIZE(sgs));
-	ret = virtqueue_add_sgs(vi->cvq, sgs, out_num, 1, vi, GFP_ATOMIC);
+	BUG_ON(out_num + in_num > ARRAY_SIZE(sgs));
+	ret = virtqueue_add_sgs(vi->cvq, sgs, out_num, in_num, vi, GFP_ATOMIC);
 	if (ret < 0) {
 		dev_warn(&vi->vdev->dev,
 			 "Failed to add sgs for command vq: %d\n.", ret);
+		mutex_unlock(&vi->cvq_lock);
 		return false;
 	}
 
 	if (unlikely(!virtqueue_kick(vi->cvq)))
-		return vi->ctrl->status == VIRTIO_NET_OK;
+		goto unlock;
 
 	/* Spin for a response, the kick causes an ioport write, trapping
 	 * into the hypervisor, so the request should be handled immediately.
@@ -2589,7 +2749,16 @@ static bool virtnet_send_command(struct virtnet_info *vi, u8 class, u8 cmd,
 		cpu_relax();
 	}
 
-	return vi->ctrl->status == VIRTIO_NET_OK;
+unlock:
+	ok = vi->ctrl->status == VIRTIO_NET_OK;
+	mutex_unlock(&vi->cvq_lock);
+	return ok;
+}
+
+static bool virtnet_send_command(struct virtnet_info *vi, u8 class, u8 cmd,
+				 struct scatterlist *out)
+{
+	return virtnet_send_command_reply(vi, class, cmd, out, NULL);
 }
 
 static int virtnet_set_mac_address(struct net_device *dev, void *p)
@@ -2681,23 +2850,26 @@ static void virtnet_stats(struct net_device *dev,
 
 static void virtnet_ack_link_announce(struct virtnet_info *vi)
 {
-	rtnl_lock();
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_ANNOUNCE,
 				  VIRTIO_NET_CTRL_ANNOUNCE_ACK, NULL))
 		dev_warn(&vi->dev->dev, "Failed to ack link announce.\n");
-	rtnl_unlock();
 }
 
-static int _virtnet_set_queues(struct virtnet_info *vi, u16 queue_pairs)
+static int virtnet_set_queues(struct virtnet_info *vi, u16 queue_pairs)
 {
+	struct virtio_net_ctrl_mq *mq __free(kfree) = NULL;
 	struct scatterlist sg;
 	struct net_device *dev = vi->dev;
 
 	if (!vi->has_cvq || !virtio_has_feature(vi->vdev, VIRTIO_NET_F_MQ))
 		return 0;
 
-	vi->ctrl->mq.virtqueue_pairs = cpu_to_virtio16(vi->vdev, queue_pairs);
-	sg_init_one(&sg, &vi->ctrl->mq, sizeof(vi->ctrl->mq));
+	mq = kzalloc(sizeof(*mq), GFP_KERNEL);
+	if (!mq)
+		return -ENOMEM;
+
+	mq->virtqueue_pairs = cpu_to_virtio16(vi->vdev, queue_pairs);
+	sg_init_one(&sg, mq, sizeof(*mq));
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_MQ,
 				  VIRTIO_NET_CTRL_MQ_VQ_PAIRS_SET, &sg)) {
@@ -2714,16 +2886,6 @@ static int _virtnet_set_queues(struct virtnet_info *vi, u16 queue_pairs)
 	return 0;
 }
 
-static int virtnet_set_queues(struct virtnet_info *vi, u16 queue_pairs)
-{
-	int err;
-
-	rtnl_lock();
-	err = _virtnet_set_queues(vi, queue_pairs);
-	rtnl_unlock();
-	return err;
-}
-
 static int virtnet_close(struct net_device *dev)
 {
 	struct virtnet_info *vi = netdev_priv(dev);
@@ -2746,6 +2908,7 @@ static void virtnet_rx_mode_work(struct work_struct *work)
 {
 	struct virtnet_info *vi =
 		container_of(work, struct virtnet_info, rx_mode_work);
+	u8 *promisc_allmulti  __free(kfree) = NULL;
 	struct net_device *dev = vi->dev;
 	struct scatterlist sg[2];
 	struct virtio_net_ctrl_mac *mac_data;
@@ -2759,24 +2922,29 @@ static void virtnet_rx_mode_work(struct work_struct *work)
 	if (!virtio_has_feature(vi->vdev, VIRTIO_NET_F_CTRL_RX))
 		return;
 
-	rtnl_lock();
+	promisc_allmulti = kzalloc(sizeof(*promisc_allmulti), GFP_KERNEL);
+	if (!promisc_allmulti) {
+		dev_warn(&dev->dev, "Failed to set RX mode, no memory.\n");
+		return;
+	}
 
-	vi->ctrl->promisc = ((dev->flags & IFF_PROMISC) != 0);
-	vi->ctrl->allmulti = ((dev->flags & IFF_ALLMULTI) != 0);
+	rtnl_lock();
 
-	sg_init_one(sg, &vi->ctrl->promisc, sizeof(vi->ctrl->promisc));
+	*promisc_allmulti = !!(dev->flags & IFF_PROMISC);
+	sg_init_one(sg, promisc_allmulti, sizeof(*promisc_allmulti));
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_RX,
 				  VIRTIO_NET_CTRL_RX_PROMISC, sg))
 		dev_warn(&dev->dev, "Failed to %sable promisc mode.\n",
-			 vi->ctrl->promisc ? "en" : "dis");
+			 *promisc_allmulti ? "en" : "dis");
 
-	sg_init_one(sg, &vi->ctrl->allmulti, sizeof(vi->ctrl->allmulti));
+	*promisc_allmulti = !!(dev->flags & IFF_ALLMULTI);
+	sg_init_one(sg, promisc_allmulti, sizeof(*promisc_allmulti));
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_RX,
 				  VIRTIO_NET_CTRL_RX_ALLMULTI, sg))
 		dev_warn(&dev->dev, "Failed to %sable allmulti mode.\n",
-			 vi->ctrl->allmulti ? "en" : "dis");
+			 *promisc_allmulti ? "en" : "dis");
 
 	netif_addr_lock_bh(dev);
 
@@ -2837,10 +3005,15 @@ static int virtnet_vlan_rx_add_vid(struct net_device *dev,
 				   __be16 proto, u16 vid)
 {
 	struct virtnet_info *vi = netdev_priv(dev);
+	__virtio16 *_vid __free(kfree) = NULL;
 	struct scatterlist sg;
 
-	vi->ctrl->vid = cpu_to_virtio16(vi->vdev, vid);
-	sg_init_one(&sg, &vi->ctrl->vid, sizeof(vi->ctrl->vid));
+	_vid = kzalloc(sizeof(*_vid), GFP_KERNEL);
+	if (!_vid)
+		return -ENOMEM;
+
+	*_vid = cpu_to_virtio16(vi->vdev, vid);
+	sg_init_one(&sg, _vid, sizeof(*_vid));
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_VLAN,
 				  VIRTIO_NET_CTRL_VLAN_ADD, &sg))
@@ -2852,10 +3025,15 @@ static int virtnet_vlan_rx_kill_vid(struct net_device *dev,
 				    __be16 proto, u16 vid)
 {
 	struct virtnet_info *vi = netdev_priv(dev);
+	__virtio16 *_vid __free(kfree) = NULL;
 	struct scatterlist sg;
 
-	vi->ctrl->vid = cpu_to_virtio16(vi->vdev, vid);
-	sg_init_one(&sg, &vi->ctrl->vid, sizeof(vi->ctrl->vid));
+	_vid = kzalloc(sizeof(*_vid), GFP_KERNEL);
+	if (!_vid)
+		return -ENOMEM;
+
+	*_vid = cpu_to_virtio16(vi->vdev, vid);
+	sg_init_one(&sg, _vid, sizeof(*_vid));
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_VLAN,
 				  VIRTIO_NET_CTRL_VLAN_DEL, &sg))
@@ -2968,12 +3146,17 @@ static void virtnet_cpu_notif_remove(struct virtnet_info *vi)
 static int virtnet_send_ctrl_coal_vq_cmd(struct virtnet_info *vi,
 					 u16 vqn, u32 max_usecs, u32 max_packets)
 {
+	struct virtio_net_ctrl_coal_vq *coal_vq __free(kfree) = NULL;
 	struct scatterlist sgs;
 
-	vi->ctrl->coal_vq.vqn = cpu_to_le16(vqn);
-	vi->ctrl->coal_vq.coal.max_usecs = cpu_to_le32(max_usecs);
-	vi->ctrl->coal_vq.coal.max_packets = cpu_to_le32(max_packets);
-	sg_init_one(&sgs, &vi->ctrl->coal_vq, sizeof(vi->ctrl->coal_vq));
+	coal_vq = kzalloc(sizeof(*coal_vq), GFP_KERNEL);
+	if (!coal_vq)
+		return -ENOMEM;
+
+	coal_vq->vqn = cpu_to_le16(vqn);
+	coal_vq->coal.max_usecs = cpu_to_le32(max_usecs);
+	coal_vq->coal.max_packets = cpu_to_le32(max_packets);
+	sg_init_one(&sgs, coal_vq, sizeof(*coal_vq));
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_NOTF_COAL,
 				  VIRTIO_NET_CTRL_NOTF_COAL_VQ_SET,
@@ -3084,9 +3267,11 @@ static int virtnet_set_ringparam(struct net_device *dev,
 				return err;
 
 			/* The reason is same as the transmit virtqueue reset */
+			mutex_lock(&vi->rq[i].dim_lock);
 			err = virtnet_send_rx_ctrl_coal_vq_cmd(vi, i,
 							       vi->intr_coal_rx.max_usecs,
 							       vi->intr_coal_rx.max_packets);
+			mutex_unlock(&vi->rq[i].dim_lock);
 			if (err)
 				return err;
 		}
@@ -3105,25 +3290,29 @@ static bool virtnet_commit_rss_command(struct virtnet_info *vi)
 	sg_init_table(sgs, 4);
 
 	sg_buf_size = offsetof(struct virtio_net_ctrl_rss, indirection_table);
-	sg_set_buf(&sgs[0], &vi->ctrl->rss, sg_buf_size);
+	sg_set_buf(&sgs[0], &vi->rss, sg_buf_size);
 
-	sg_buf_size = sizeof(uint16_t) * (vi->ctrl->rss.indirection_table_mask + 1);
-	sg_set_buf(&sgs[1], vi->ctrl->rss.indirection_table, sg_buf_size);
+	sg_buf_size = sizeof(uint16_t) * (vi->rss.indirection_table_mask + 1);
+	sg_set_buf(&sgs[1], vi->rss.indirection_table, sg_buf_size);
 
 	sg_buf_size = offsetof(struct virtio_net_ctrl_rss, key)
 			- offsetof(struct virtio_net_ctrl_rss, max_tx_vq);
-	sg_set_buf(&sgs[2], &vi->ctrl->rss.max_tx_vq, sg_buf_size);
+	sg_set_buf(&sgs[2], &vi->rss.max_tx_vq, sg_buf_size);
 
 	sg_buf_size = vi->rss_key_size;
-	sg_set_buf(&sgs[3], vi->ctrl->rss.key, sg_buf_size);
+	sg_set_buf(&sgs[3], vi->rss.key, sg_buf_size);
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_MQ,
 				  vi->has_rss ? VIRTIO_NET_CTRL_MQ_RSS_CONFIG
-				  : VIRTIO_NET_CTRL_MQ_HASH_CONFIG, sgs)) {
-		dev_warn(&dev->dev, "VIRTIONET issue with committing RSS sgs\n");
-		return false;
-	}
+				  : VIRTIO_NET_CTRL_MQ_HASH_CONFIG, sgs))
+		goto err;
+
 	return true;
+
+err:
+	dev_warn(&dev->dev, "VIRTIONET issue with committing RSS sgs\n");
+	return false;
+
 }
 
 static void virtnet_init_default_rss(struct virtnet_info *vi)
@@ -3131,21 +3320,21 @@ static void virtnet_init_default_rss(struct virtnet_info *vi)
 	u32 indir_val = 0;
 	int i = 0;
 
-	vi->ctrl->rss.hash_types = vi->rss_hash_types_supported;
+	vi->rss.hash_types = vi->rss_hash_types_supported;
 	vi->rss_hash_types_saved = vi->rss_hash_types_supported;
-	vi->ctrl->rss.indirection_table_mask = vi->rss_indir_table_size
+	vi->rss.indirection_table_mask = vi->rss_indir_table_size
 						? vi->rss_indir_table_size - 1 : 0;
-	vi->ctrl->rss.unclassified_queue = 0;
+	vi->rss.unclassified_queue = 0;
 
 	for (; i < vi->rss_indir_table_size; ++i) {
 		indir_val = ethtool_rxfh_indir_default(i, vi->curr_queue_pairs);
-		vi->ctrl->rss.indirection_table[i] = indir_val;
+		vi->rss.indirection_table[i] = indir_val;
 	}
 
-	vi->ctrl->rss.max_tx_vq = vi->has_rss ? vi->curr_queue_pairs : 0;
-	vi->ctrl->rss.hash_key_length = vi->rss_key_size;
+	vi->rss.max_tx_vq = vi->has_rss ? vi->curr_queue_pairs : 0;
+	vi->rss.hash_key_length = vi->rss_key_size;
 
-	netdev_rss_key_fill(vi->ctrl->rss.key, vi->rss_key_size);
+	netdev_rss_key_fill(vi->rss.key, vi->rss_key_size);
 }
 
 static void virtnet_get_hashflow(const struct virtnet_info *vi, struct ethtool_rxnfc *info)
@@ -3256,7 +3445,7 @@ static bool virtnet_set_hashflow(struct virtnet_info *vi, struct ethtool_rxnfc *
 
 	if (new_hashtypes != vi->rss_hash_types_saved) {
 		vi->rss_hash_types_saved = new_hashtypes;
-		vi->ctrl->rss.hash_types = vi->rss_hash_types_saved;
+		vi->rss.hash_types = vi->rss_hash_types_saved;
 		if (vi->dev->features & NETIF_F_RXHASH)
 			return virtnet_commit_rss_command(vi);
 	}
@@ -3301,7 +3490,7 @@ static int virtnet_set_channels(struct net_device *dev,
 		return -EINVAL;
 
 	cpus_read_lock();
-	err = _virtnet_set_queues(vi, queue_pairs);
+	err = virtnet_set_queues(vi, queue_pairs);
 	if (err) {
 		cpus_read_unlock();
 		goto err;
@@ -3315,25 +3504,654 @@ static int virtnet_set_channels(struct net_device *dev,
 	return err;
 }
 
+static void virtnet_stats_sprintf(u8 **p, const char *fmt, const char *noq_fmt,
+				  int num, int qid, const struct virtnet_stat_desc *desc)
+{
+	int i;
+
+	if (qid < 0) {
+		for (i = 0; i < num; ++i)
+			ethtool_sprintf(p, noq_fmt, desc[i].desc);
+	} else {
+		for (i = 0; i < num; ++i)
+			ethtool_sprintf(p, fmt, qid, desc[i].desc);
+	}
+}
+
+/* qid == -1: for rx/tx queue total field */
+static void virtnet_get_stats_string(struct virtnet_info *vi, int type, int qid, u8 **data)
+{
+	const struct virtnet_stat_desc *desc;
+	const char *fmt, *noq_fmt;
+	u8 *p = *data;
+	u32 num;
+
+	if (type == VIRTNET_Q_TYPE_CQ && qid >= 0) {
+		noq_fmt = "cq_hw_%s";
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_CVQ) {
+			desc = &virtnet_stats_cvq_desc[0];
+			num = ARRAY_SIZE(virtnet_stats_cvq_desc);
+
+			virtnet_stats_sprintf(&p, NULL, noq_fmt, num, -1, desc);
+		}
+	}
+
+	if (type == VIRTNET_Q_TYPE_RX) {
+		fmt = "rx%u_%s";
+		noq_fmt = "rx_%s";
+
+		desc = &virtnet_rq_stats_desc[0];
+		num = ARRAY_SIZE(virtnet_rq_stats_desc);
+
+		virtnet_stats_sprintf(&p, fmt, noq_fmt, num, qid, desc);
+
+		fmt = "rx%u_hw_%s";
+		noq_fmt = "rx_hw_%s";
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_BASIC) {
+			desc = &virtnet_stats_rx_basic_desc[0];
+			num = ARRAY_SIZE(virtnet_stats_rx_basic_desc);
+
+			virtnet_stats_sprintf(&p, fmt, noq_fmt, num, qid, desc);
+		}
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_CSUM) {
+			desc = &virtnet_stats_rx_csum_desc[0];
+			num = ARRAY_SIZE(virtnet_stats_rx_csum_desc);
+
+			virtnet_stats_sprintf(&p, fmt, noq_fmt, num, qid, desc);
+		}
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_SPEED) {
+			desc = &virtnet_stats_rx_speed_desc[0];
+			num = ARRAY_SIZE(virtnet_stats_rx_speed_desc);
+
+			virtnet_stats_sprintf(&p, fmt, noq_fmt, num, qid, desc);
+		}
+	}
+
+	if (type == VIRTNET_Q_TYPE_TX) {
+		fmt = "tx%u_%s";
+		noq_fmt = "tx_%s";
+
+		desc = &virtnet_sq_stats_desc[0];
+		num = ARRAY_SIZE(virtnet_sq_stats_desc);
+
+		virtnet_stats_sprintf(&p, fmt, noq_fmt, num, qid, desc);
+
+		fmt = "tx%u_hw_%s";
+		noq_fmt = "tx_hw_%s";
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_BASIC) {
+			desc = &virtnet_stats_tx_basic_desc[0];
+			num = ARRAY_SIZE(virtnet_stats_tx_basic_desc);
+
+			virtnet_stats_sprintf(&p, fmt, noq_fmt, num, qid, desc);
+		}
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_GSO) {
+			desc = &virtnet_stats_tx_gso_desc[0];
+			num = ARRAY_SIZE(virtnet_stats_tx_gso_desc);
+
+			virtnet_stats_sprintf(&p, fmt, noq_fmt, num, qid, desc);
+		}
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_SPEED) {
+			desc = &virtnet_stats_tx_speed_desc[0];
+			num = ARRAY_SIZE(virtnet_stats_tx_speed_desc);
+
+			virtnet_stats_sprintf(&p, fmt, noq_fmt, num, qid, desc);
+		}
+	}
+
+	*data = p;
+}
+
+struct virtnet_stats_ctx {
+	/* The stats are write to qstats or ethtool -S */
+	bool to_qstat;
+
+	/* Used to calculate the offset inside the output buffer. */
+	u32 desc_num[3];
+
+	/* The actual supported stat types. */
+	u32 bitmap[3];
+
+	/* Used to calculate the reply buffer size. */
+	u32 size[3];
+
+	/* Record the output buffer. */
+	u64 *data;
+};
+
+static void virtnet_stats_ctx_init(struct virtnet_info *vi,
+				   struct virtnet_stats_ctx *ctx,
+				   u64 *data, bool to_qstat)
+{
+	u32 queue_type;
+
+	ctx->data = data;
+	ctx->to_qstat = to_qstat;
+
+	if (to_qstat) {
+		ctx->desc_num[VIRTNET_Q_TYPE_RX] = ARRAY_SIZE(virtnet_rq_stats_desc_qstat);
+		ctx->desc_num[VIRTNET_Q_TYPE_TX] = ARRAY_SIZE(virtnet_sq_stats_desc_qstat);
+
+		queue_type = VIRTNET_Q_TYPE_RX;
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_BASIC) {
+			ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_RX_BASIC;
+			ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_rx_basic_desc_qstat);
+			ctx->size[queue_type]     += sizeof(struct virtio_net_stats_rx_basic);
+		}
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_CSUM) {
+			ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_RX_CSUM;
+			ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_rx_csum_desc_qstat);
+			ctx->size[queue_type]     += sizeof(struct virtio_net_stats_rx_csum);
+		}
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_GSO) {
+			ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_RX_GSO;
+			ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_rx_gso_desc_qstat);
+			ctx->size[queue_type]     += sizeof(struct virtio_net_stats_rx_gso);
+		}
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_SPEED) {
+			ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_RX_SPEED;
+			ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_rx_speed_desc_qstat);
+			ctx->size[queue_type]     += sizeof(struct virtio_net_stats_rx_speed);
+		}
+
+		queue_type = VIRTNET_Q_TYPE_TX;
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_BASIC) {
+			ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_TX_BASIC;
+			ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_tx_basic_desc_qstat);
+			ctx->size[queue_type]     += sizeof(struct virtio_net_stats_tx_basic);
+		}
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_CSUM) {
+			ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_TX_CSUM;
+			ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_tx_csum_desc_qstat);
+			ctx->size[queue_type]     += sizeof(struct virtio_net_stats_tx_csum);
+		}
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_GSO) {
+			ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_TX_GSO;
+			ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_tx_gso_desc_qstat);
+			ctx->size[queue_type]     += sizeof(struct virtio_net_stats_tx_gso);
+		}
+
+		if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_SPEED) {
+			ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_TX_SPEED;
+			ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_tx_speed_desc_qstat);
+			ctx->size[queue_type]     += sizeof(struct virtio_net_stats_tx_speed);
+		}
+
+		return;
+	}
+
+	ctx->desc_num[VIRTNET_Q_TYPE_RX] = ARRAY_SIZE(virtnet_rq_stats_desc);
+	ctx->desc_num[VIRTNET_Q_TYPE_TX] = ARRAY_SIZE(virtnet_sq_stats_desc);
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_CVQ) {
+		queue_type = VIRTNET_Q_TYPE_CQ;
+
+		ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_CVQ;
+		ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_cvq_desc);
+		ctx->size[queue_type]     += sizeof(struct virtio_net_stats_cvq);
+	}
+
+	queue_type = VIRTNET_Q_TYPE_RX;
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_BASIC) {
+		ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_RX_BASIC;
+		ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_rx_basic_desc);
+		ctx->size[queue_type]     += sizeof(struct virtio_net_stats_rx_basic);
+	}
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_CSUM) {
+		ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_RX_CSUM;
+		ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_rx_csum_desc);
+		ctx->size[queue_type]     += sizeof(struct virtio_net_stats_rx_csum);
+	}
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_SPEED) {
+		ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_RX_SPEED;
+		ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_rx_speed_desc);
+		ctx->size[queue_type]     += sizeof(struct virtio_net_stats_rx_speed);
+	}
+
+	queue_type = VIRTNET_Q_TYPE_TX;
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_BASIC) {
+		ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_TX_BASIC;
+		ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_tx_basic_desc);
+		ctx->size[queue_type]     += sizeof(struct virtio_net_stats_tx_basic);
+	}
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_GSO) {
+		ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_TX_GSO;
+		ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_tx_gso_desc);
+		ctx->size[queue_type]     += sizeof(struct virtio_net_stats_tx_gso);
+	}
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_SPEED) {
+		ctx->bitmap[queue_type]   |= VIRTIO_NET_STATS_TYPE_TX_SPEED;
+		ctx->desc_num[queue_type] += ARRAY_SIZE(virtnet_stats_tx_speed_desc);
+		ctx->size[queue_type]     += sizeof(struct virtio_net_stats_tx_speed);
+	}
+}
+
+/* stats_sum_queue - Calculate the sum of the same fields in sq or rq.
+ * @sum: the position to store the sum values
+ * @num: field num
+ * @q_value: the first queue fields
+ * @q_num: number of the queues
+ */
+static void stats_sum_queue(u64 *sum, u32 num, u64 *q_value, u32 q_num)
+{
+	u32 step = num;
+	int i, j;
+	u64 *p;
+
+	for (i = 0; i < num; ++i) {
+		p = sum + i;
+		*p = 0;
+
+		for (j = 0; j < q_num; ++j)
+			*p += *(q_value + i + j * step);
+	}
+}
+
+static void virtnet_fill_total_fields(struct virtnet_info *vi,
+				      struct virtnet_stats_ctx *ctx)
+{
+	u64 *data, *first_rx_q, *first_tx_q;
+	u32 num_cq, num_rx, num_tx;
+
+	num_cq = ctx->desc_num[VIRTNET_Q_TYPE_CQ];
+	num_rx = ctx->desc_num[VIRTNET_Q_TYPE_RX];
+	num_tx = ctx->desc_num[VIRTNET_Q_TYPE_TX];
+
+	first_rx_q = ctx->data + num_rx + num_tx + num_cq;
+	first_tx_q = first_rx_q + vi->curr_queue_pairs * num_rx;
+
+	data = ctx->data;
+
+	stats_sum_queue(data, num_rx, first_rx_q, vi->curr_queue_pairs);
+
+	data = ctx->data + num_rx;
+
+	stats_sum_queue(data, num_tx, first_tx_q, vi->curr_queue_pairs);
+}
+
+static void virtnet_fill_stats_qstat(struct virtnet_info *vi, u32 qid,
+				     struct virtnet_stats_ctx *ctx,
+				     const u8 *base, bool drv_stats, u8 reply_type)
+{
+	const struct virtnet_stat_desc *desc;
+	const u64_stats_t *v_stat;
+	u64 offset, bitmap;
+	const __le64 *v;
+	u32 queue_type;
+	int i, num;
+
+	queue_type = vq_type(vi, qid);
+	bitmap = ctx->bitmap[queue_type];
+
+	if (drv_stats) {
+		if (queue_type == VIRTNET_Q_TYPE_RX) {
+			desc = &virtnet_rq_stats_desc_qstat[0];
+			num = ARRAY_SIZE(virtnet_rq_stats_desc_qstat);
+		} else {
+			desc = &virtnet_sq_stats_desc_qstat[0];
+			num = ARRAY_SIZE(virtnet_sq_stats_desc_qstat);
+		}
+
+		for (i = 0; i < num; ++i) {
+			offset = desc[i].qstat_offset / sizeof(*ctx->data);
+			v_stat = (const u64_stats_t *)(base + desc[i].offset);
+			ctx->data[offset] = u64_stats_read(v_stat);
+		}
+		return;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_RX_BASIC) {
+		desc = &virtnet_stats_rx_basic_desc_qstat[0];
+		num = ARRAY_SIZE(virtnet_stats_rx_basic_desc_qstat);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_RX_BASIC)
+			goto found;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_RX_CSUM) {
+		desc = &virtnet_stats_rx_csum_desc_qstat[0];
+		num = ARRAY_SIZE(virtnet_stats_rx_csum_desc_qstat);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_RX_CSUM)
+			goto found;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_RX_GSO) {
+		desc = &virtnet_stats_rx_gso_desc_qstat[0];
+		num = ARRAY_SIZE(virtnet_stats_rx_gso_desc_qstat);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_RX_GSO)
+			goto found;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_RX_SPEED) {
+		desc = &virtnet_stats_rx_speed_desc_qstat[0];
+		num = ARRAY_SIZE(virtnet_stats_rx_speed_desc_qstat);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_RX_SPEED)
+			goto found;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_TX_BASIC) {
+		desc = &virtnet_stats_tx_basic_desc_qstat[0];
+		num = ARRAY_SIZE(virtnet_stats_tx_basic_desc_qstat);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_TX_BASIC)
+			goto found;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_TX_CSUM) {
+		desc = &virtnet_stats_tx_csum_desc_qstat[0];
+		num = ARRAY_SIZE(virtnet_stats_tx_csum_desc_qstat);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_TX_CSUM)
+			goto found;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_TX_GSO) {
+		desc = &virtnet_stats_tx_gso_desc_qstat[0];
+		num = ARRAY_SIZE(virtnet_stats_tx_gso_desc_qstat);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_TX_GSO)
+			goto found;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_TX_SPEED) {
+		desc = &virtnet_stats_tx_speed_desc_qstat[0];
+		num = ARRAY_SIZE(virtnet_stats_tx_speed_desc_qstat);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_TX_SPEED)
+			goto found;
+	}
+
+	return;
+
+found:
+	for (i = 0; i < num; ++i) {
+		offset = desc[i].qstat_offset / sizeof(*ctx->data);
+		v = (const __le64 *)(base + desc[i].offset);
+		ctx->data[offset] = le64_to_cpu(*v);
+	}
+}
+
+/* virtnet_fill_stats - copy the stats to qstats or ethtool -S
+ * The stats source is the device or the driver.
+ *
+ * @vi: virtio net info
+ * @qid: the vq id
+ * @ctx: stats ctx (initiated by virtnet_stats_ctx_init())
+ * @base: pointer to the device reply or the driver stats structure.
+ * @drv_stats: designate the base type (device reply, driver stats)
+ * @type: the type of the device reply (if drv_stats is true, this must be zero)
+ */
+static void virtnet_fill_stats(struct virtnet_info *vi, u32 qid,
+			       struct virtnet_stats_ctx *ctx,
+			       const u8 *base, bool drv_stats, u8 reply_type)
+{
+	u32 queue_type, num_rx, num_tx, num_cq;
+	const struct virtnet_stat_desc *desc;
+	const u64_stats_t *v_stat;
+	u64 offset, bitmap;
+	const __le64 *v;
+	int i, num;
+
+	if (ctx->to_qstat)
+		return virtnet_fill_stats_qstat(vi, qid, ctx, base, drv_stats, reply_type);
+
+	num_cq = ctx->desc_num[VIRTNET_Q_TYPE_CQ];
+	num_rx = ctx->desc_num[VIRTNET_Q_TYPE_RX];
+	num_tx = ctx->desc_num[VIRTNET_Q_TYPE_TX];
+
+	queue_type = vq_type(vi, qid);
+	bitmap = ctx->bitmap[queue_type];
+
+	/* skip the total fields of pairs */
+	offset = num_rx + num_tx;
+
+	if (queue_type == VIRTNET_Q_TYPE_TX) {
+		offset += num_cq + num_rx * vi->curr_queue_pairs + num_tx * (qid / 2);
+
+		num = ARRAY_SIZE(virtnet_sq_stats_desc);
+		if (drv_stats) {
+			desc = &virtnet_sq_stats_desc[0];
+			goto drv_stats;
+		}
+
+		offset += num;
+
+	} else if (queue_type == VIRTNET_Q_TYPE_RX) {
+		offset += num_cq + num_rx * (qid / 2);
+
+		num = ARRAY_SIZE(virtnet_rq_stats_desc);
+		if (drv_stats) {
+			desc = &virtnet_rq_stats_desc[0];
+			goto drv_stats;
+		}
+
+		offset += num;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_CVQ) {
+		desc = &virtnet_stats_cvq_desc[0];
+		num = ARRAY_SIZE(virtnet_stats_cvq_desc);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_CVQ)
+			goto found;
+
+		offset += num;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_RX_BASIC) {
+		desc = &virtnet_stats_rx_basic_desc[0];
+		num = ARRAY_SIZE(virtnet_stats_rx_basic_desc);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_RX_BASIC)
+			goto found;
+
+		offset += num;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_RX_CSUM) {
+		desc = &virtnet_stats_rx_csum_desc[0];
+		num = ARRAY_SIZE(virtnet_stats_rx_csum_desc);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_RX_CSUM)
+			goto found;
+
+		offset += num;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_RX_SPEED) {
+		desc = &virtnet_stats_rx_speed_desc[0];
+		num = ARRAY_SIZE(virtnet_stats_rx_speed_desc);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_RX_SPEED)
+			goto found;
+
+		offset += num;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_TX_BASIC) {
+		desc = &virtnet_stats_tx_basic_desc[0];
+		num = ARRAY_SIZE(virtnet_stats_tx_basic_desc);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_TX_BASIC)
+			goto found;
+
+		offset += num;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_TX_GSO) {
+		desc = &virtnet_stats_tx_gso_desc[0];
+		num = ARRAY_SIZE(virtnet_stats_tx_gso_desc);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_TX_GSO)
+			goto found;
+
+		offset += num;
+	}
+
+	if (bitmap & VIRTIO_NET_STATS_TYPE_TX_SPEED) {
+		desc = &virtnet_stats_tx_speed_desc[0];
+		num = ARRAY_SIZE(virtnet_stats_tx_speed_desc);
+		if (reply_type == VIRTIO_NET_STATS_TYPE_REPLY_TX_SPEED)
+			goto found;
+
+		offset += num;
+	}
+
+	return;
+
+found:
+	for (i = 0; i < num; ++i) {
+		v = (const __le64 *)(base + desc[i].offset);
+		ctx->data[offset + i] = le64_to_cpu(*v);
+	}
+
+	return;
+
+drv_stats:
+	for (i = 0; i < num; ++i) {
+		v_stat = (const u64_stats_t *)(base + desc[i].offset);
+		ctx->data[offset + i] = u64_stats_read(v_stat);
+	}
+}
+
+static int __virtnet_get_hw_stats(struct virtnet_info *vi,
+				  struct virtnet_stats_ctx *ctx,
+				  struct virtio_net_ctrl_queue_stats *req,
+				  int req_size, void *reply, int res_size)
+{
+	struct virtio_net_stats_reply_hdr *hdr;
+	struct scatterlist sgs_in, sgs_out;
+	void *p;
+	u32 qid;
+	int ok;
+
+	sg_init_one(&sgs_out, req, req_size);
+	sg_init_one(&sgs_in, reply, res_size);
+
+	ok = virtnet_send_command_reply(vi, VIRTIO_NET_CTRL_STATS,
+					VIRTIO_NET_CTRL_STATS_GET,
+					&sgs_out, &sgs_in);
+
+	if (!ok)
+		return ok;
+
+	for (p = reply; p - reply < res_size; p += le16_to_cpu(hdr->size)) {
+		hdr = p;
+		qid = le16_to_cpu(hdr->vq_index);
+		virtnet_fill_stats(vi, qid, ctx, p, false, hdr->type);
+	}
+
+	return 0;
+}
+
+static void virtnet_make_stat_req(struct virtnet_info *vi,
+				  struct virtnet_stats_ctx *ctx,
+				  struct virtio_net_ctrl_queue_stats *req,
+				  int qid, int *idx)
+{
+	int qtype = vq_type(vi, qid);
+	u64 bitmap = ctx->bitmap[qtype];
+
+	if (!bitmap)
+		return;
+
+	req->stats[*idx].vq_index = cpu_to_le16(qid);
+	req->stats[*idx].types_bitmap[0] = cpu_to_le64(bitmap);
+	*idx += 1;
+}
+
+/* qid: -1: get stats of all vq.
+ *     > 0: get the stats for the special vq. This must not be cvq.
+ */
+static int virtnet_get_hw_stats(struct virtnet_info *vi,
+				struct virtnet_stats_ctx *ctx, int qid)
+{
+	int qnum, i, j, res_size, qtype, last_vq, first_vq;
+	struct virtio_net_ctrl_queue_stats *req;
+	bool enable_cvq;
+	void *reply;
+	int ok;
+
+	if (!virtio_has_feature(vi->vdev, VIRTIO_NET_F_DEVICE_STATS))
+		return 0;
+
+	if (qid == -1) {
+		last_vq = vi->curr_queue_pairs * 2 - 1;
+		first_vq = 0;
+		enable_cvq = true;
+	} else {
+		last_vq = qid;
+		first_vq = qid;
+		enable_cvq = false;
+	}
+
+	qnum = 0;
+	res_size = 0;
+	for (i = first_vq; i <= last_vq ; ++i) {
+		qtype = vq_type(vi, i);
+		if (ctx->bitmap[qtype]) {
+			++qnum;
+			res_size += ctx->size[qtype];
+		}
+	}
+
+	if (enable_cvq && ctx->bitmap[VIRTNET_Q_TYPE_CQ]) {
+		res_size += ctx->size[VIRTNET_Q_TYPE_CQ];
+		qnum += 1;
+	}
+
+	req = kcalloc(qnum, sizeof(*req), GFP_KERNEL);
+	if (!req)
+		return -ENOMEM;
+
+	reply = kmalloc(res_size, GFP_KERNEL);
+	if (!reply) {
+		kfree(req);
+		return -ENOMEM;
+	}
+
+	j = 0;
+	for (i = first_vq; i <= last_vq ; ++i)
+		virtnet_make_stat_req(vi, ctx, req, i, &j);
+
+	if (enable_cvq)
+		virtnet_make_stat_req(vi, ctx, req, vi->max_queue_pairs * 2, &j);
+
+	ok = __virtnet_get_hw_stats(vi, ctx, req, sizeof(*req) * j, reply, res_size);
+
+	kfree(req);
+	kfree(reply);
+
+	return ok;
+}
+
 static void virtnet_get_strings(struct net_device *dev, u32 stringset, u8 *data)
 {
 	struct virtnet_info *vi = netdev_priv(dev);
-	unsigned int i, j;
+	unsigned int i;
 	u8 *p = data;
 
 	switch (stringset) {
 	case ETH_SS_STATS:
-		for (i = 0; i < vi->curr_queue_pairs; i++) {
-			for (j = 0; j < VIRTNET_RQ_STATS_LEN; j++)
-				ethtool_sprintf(&p, "rx_queue_%u_%s", i,
-						virtnet_rq_stats_desc[j].desc);
-		}
+		/* Generate the total field names. */
+		virtnet_get_stats_string(vi, VIRTNET_Q_TYPE_RX, -1, &p);
+		virtnet_get_stats_string(vi, VIRTNET_Q_TYPE_TX, -1, &p);
 
-		for (i = 0; i < vi->curr_queue_pairs; i++) {
-			for (j = 0; j < VIRTNET_SQ_STATS_LEN; j++)
-				ethtool_sprintf(&p, "tx_queue_%u_%s", i,
-						virtnet_sq_stats_desc[j].desc);
-		}
+		virtnet_get_stats_string(vi, VIRTNET_Q_TYPE_CQ, 0, &p);
+
+		for (i = 0; i < vi->curr_queue_pairs; ++i)
+			virtnet_get_stats_string(vi, VIRTNET_Q_TYPE_RX, i, &p);
+
+		for (i = 0; i < vi->curr_queue_pairs; ++i)
+			virtnet_get_stats_string(vi, VIRTNET_Q_TYPE_TX, i, &p);
 		break;
 	}
 }
@@ -3341,11 +4159,17 @@ static void virtnet_get_strings(struct net_device *dev, u32 stringset, u8 *data)
 static int virtnet_get_sset_count(struct net_device *dev, int sset)
 {
 	struct virtnet_info *vi = netdev_priv(dev);
+	struct virtnet_stats_ctx ctx = {0};
+	u32 pair_count;
 
 	switch (sset) {
 	case ETH_SS_STATS:
-		return vi->curr_queue_pairs * (VIRTNET_RQ_STATS_LEN +
-					       VIRTNET_SQ_STATS_LEN);
+		virtnet_stats_ctx_init(vi, &ctx, NULL, false);
+
+		pair_count = ctx.desc_num[VIRTNET_Q_TYPE_RX] + ctx.desc_num[VIRTNET_Q_TYPE_TX];
+
+		return pair_count + ctx.desc_num[VIRTNET_Q_TYPE_CQ] +
+			vi->curr_queue_pairs * pair_count;
 	default:
 		return -EOPNOTSUPP;
 	}
@@ -3355,40 +4179,32 @@ static void virtnet_get_ethtool_stats(struct net_device *dev,
 				      struct ethtool_stats *stats, u64 *data)
 {
 	struct virtnet_info *vi = netdev_priv(dev);
-	unsigned int idx = 0, start, i, j;
+	struct virtnet_stats_ctx ctx = {0};
+	unsigned int start, i;
 	const u8 *stats_base;
-	const u64_stats_t *p;
-	size_t offset;
+
+	virtnet_stats_ctx_init(vi, &ctx, data, false);
+	if (virtnet_get_hw_stats(vi, &ctx, -1))
+		dev_warn(&vi->dev->dev, "Failed to get hw stats.\n");
 
 	for (i = 0; i < vi->curr_queue_pairs; i++) {
 		struct receive_queue *rq = &vi->rq[i];
+		struct send_queue *sq = &vi->sq[i];
 
 		stats_base = (const u8 *)&rq->stats;
 		do {
 			start = u64_stats_fetch_begin(&rq->stats.syncp);
-			for (j = 0; j < VIRTNET_RQ_STATS_LEN; j++) {
-				offset = virtnet_rq_stats_desc[j].offset;
-				p = (const u64_stats_t *)(stats_base + offset);
-				data[idx + j] = u64_stats_read(p);
-			}
+			virtnet_fill_stats(vi, i * 2, &ctx, stats_base, true, 0);
 		} while (u64_stats_fetch_retry(&rq->stats.syncp, start));
-		idx += VIRTNET_RQ_STATS_LEN;
-	}
-
-	for (i = 0; i < vi->curr_queue_pairs; i++) {
-		struct send_queue *sq = &vi->sq[i];
 
 		stats_base = (const u8 *)&sq->stats;
 		do {
 			start = u64_stats_fetch_begin(&sq->stats.syncp);
-			for (j = 0; j < VIRTNET_SQ_STATS_LEN; j++) {
-				offset = virtnet_sq_stats_desc[j].offset;
-				p = (const u64_stats_t *)(stats_base + offset);
-				data[idx + j] = u64_stats_read(p);
-			}
+			virtnet_fill_stats(vi, i * 2 + 1, &ctx, stats_base, true, 0);
 		} while (u64_stats_fetch_retry(&sq->stats.syncp, start));
-		idx += VIRTNET_SQ_STATS_LEN;
 	}
+
+	virtnet_fill_total_fields(vi, &ctx);
 }
 
 static void virtnet_get_channels(struct net_device *dev,
@@ -3428,12 +4244,17 @@ static int virtnet_get_link_ksettings(struct net_device *dev,
 static int virtnet_send_tx_notf_coal_cmds(struct virtnet_info *vi,
 					  struct ethtool_coalesce *ec)
 {
+	struct virtio_net_ctrl_coal_tx *coal_tx __free(kfree) = NULL;
 	struct scatterlist sgs_tx;
 	int i;
 
-	vi->ctrl->coal_tx.tx_usecs = cpu_to_le32(ec->tx_coalesce_usecs);
-	vi->ctrl->coal_tx.tx_max_packets = cpu_to_le32(ec->tx_max_coalesced_frames);
-	sg_init_one(&sgs_tx, &vi->ctrl->coal_tx, sizeof(vi->ctrl->coal_tx));
+	coal_tx = kzalloc(sizeof(*coal_tx), GFP_KERNEL);
+	if (!coal_tx)
+		return -ENOMEM;
+
+	coal_tx->tx_usecs = cpu_to_le32(ec->tx_coalesce_usecs);
+	coal_tx->tx_max_packets = cpu_to_le32(ec->tx_max_coalesced_frames);
+	sg_init_one(&sgs_tx, coal_tx, sizeof(*coal_tx));
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_NOTF_COAL,
 				  VIRTIO_NET_CTRL_NOTF_COAL_TX_SET,
@@ -3453,6 +4274,7 @@ static int virtnet_send_tx_notf_coal_cmds(struct virtnet_info *vi,
 static int virtnet_send_rx_notf_coal_cmds(struct virtnet_info *vi,
 					  struct ethtool_coalesce *ec)
 {
+	struct virtio_net_ctrl_coal_rx *coal_rx __free(kfree) = NULL;
 	bool rx_ctrl_dim_on = !!ec->use_adaptive_rx_coalesce;
 	struct scatterlist sgs_rx;
 	int i;
@@ -3466,24 +4288,34 @@ static int virtnet_send_rx_notf_coal_cmds(struct virtnet_info *vi,
 
 	if (rx_ctrl_dim_on && !vi->rx_dim_enabled) {
 		vi->rx_dim_enabled = true;
-		for (i = 0; i < vi->max_queue_pairs; i++)
+		for (i = 0; i < vi->max_queue_pairs; i++) {
+			mutex_lock(&vi->rq[i].dim_lock);
 			vi->rq[i].dim_enabled = true;
+			mutex_unlock(&vi->rq[i].dim_lock);
+		}
 		return 0;
 	}
 
+	coal_rx = kzalloc(sizeof(*coal_rx), GFP_KERNEL);
+	if (!coal_rx)
+		return -ENOMEM;
+
 	if (!rx_ctrl_dim_on && vi->rx_dim_enabled) {
 		vi->rx_dim_enabled = false;
-		for (i = 0; i < vi->max_queue_pairs; i++)
+		for (i = 0; i < vi->max_queue_pairs; i++) {
+			mutex_lock(&vi->rq[i].dim_lock);
 			vi->rq[i].dim_enabled = false;
+			mutex_unlock(&vi->rq[i].dim_lock);
+		}
 	}
 
 	/* Since the per-queue coalescing params can be set,
 	 * we need apply the global new params even if they
 	 * are not updated.
 	 */
-	vi->ctrl->coal_rx.rx_usecs = cpu_to_le32(ec->rx_coalesce_usecs);
-	vi->ctrl->coal_rx.rx_max_packets = cpu_to_le32(ec->rx_max_coalesced_frames);
-	sg_init_one(&sgs_rx, &vi->ctrl->coal_rx, sizeof(vi->ctrl->coal_rx));
+	coal_rx->rx_usecs = cpu_to_le32(ec->rx_coalesce_usecs);
+	coal_rx->rx_max_packets = cpu_to_le32(ec->rx_max_coalesced_frames);
+	sg_init_one(&sgs_rx, coal_rx, sizeof(*coal_rx));
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_NOTF_COAL,
 				  VIRTIO_NET_CTRL_NOTF_COAL_RX_SET,
@@ -3493,8 +4325,10 @@ static int virtnet_send_rx_notf_coal_cmds(struct virtnet_info *vi,
 	vi->intr_coal_rx.max_usecs = ec->rx_coalesce_usecs;
 	vi->intr_coal_rx.max_packets = ec->rx_max_coalesced_frames;
 	for (i = 0; i < vi->max_queue_pairs; i++) {
+		mutex_lock(&vi->rq[i].dim_lock);
 		vi->rq[i].intr_coal.max_usecs = ec->rx_coalesce_usecs;
 		vi->rq[i].intr_coal.max_packets = ec->rx_max_coalesced_frames;
+		mutex_unlock(&vi->rq[i].dim_lock);
 	}
 
 	return 0;
@@ -3521,19 +4355,24 @@ static int virtnet_send_rx_notf_coal_vq_cmds(struct virtnet_info *vi,
 					     u16 queue)
 {
 	bool rx_ctrl_dim_on = !!ec->use_adaptive_rx_coalesce;
-	bool cur_rx_dim = vi->rq[queue].dim_enabled;
 	u32 max_usecs, max_packets;
+	bool cur_rx_dim;
 	int err;
 
+	mutex_lock(&vi->rq[queue].dim_lock);
+	cur_rx_dim = vi->rq[queue].dim_enabled;
 	max_usecs = vi->rq[queue].intr_coal.max_usecs;
 	max_packets = vi->rq[queue].intr_coal.max_packets;
 
 	if (rx_ctrl_dim_on && (ec->rx_coalesce_usecs != max_usecs ||
-			       ec->rx_max_coalesced_frames != max_packets))
+			       ec->rx_max_coalesced_frames != max_packets)) {
+		mutex_unlock(&vi->rq[queue].dim_lock);
 		return -EINVAL;
+	}
 
 	if (rx_ctrl_dim_on && !cur_rx_dim) {
 		vi->rq[queue].dim_enabled = true;
+		mutex_unlock(&vi->rq[queue].dim_lock);
 		return 0;
 	}
 
@@ -3546,10 +4385,8 @@ static int virtnet_send_rx_notf_coal_vq_cmds(struct virtnet_info *vi,
 	err = virtnet_send_rx_ctrl_coal_vq_cmd(vi, queue,
 					       ec->rx_coalesce_usecs,
 					       ec->rx_max_coalesced_frames);
-	if (err)
-		return err;
-
-	return 0;
+	mutex_unlock(&vi->rq[queue].dim_lock);
+	return err;
 }
 
 static int virtnet_send_notf_coal_vq_cmds(struct virtnet_info *vi,
@@ -3579,39 +4416,27 @@ static void virtnet_rx_dim_work(struct work_struct *work)
 	struct virtnet_info *vi = rq->vq->vdev->priv;
 	struct net_device *dev = vi->dev;
 	struct dim_cq_moder update_moder;
-	int i, qnum, err;
+	int qnum, err;
 
-	if (!rtnl_trylock())
-		return;
+	qnum = rq - vi->rq;
 
-	/* Each rxq's work is queued by "net_dim()->schedule_work()"
-	 * in response to NAPI traffic changes. Note that dim->profile_ix
-	 * for each rxq is updated prior to the queuing action.
-	 * So we only need to traverse and update profiles for all rxqs
-	 * in the work which is holding rtnl_lock.
-	 */
-	for (i = 0; i < vi->curr_queue_pairs; i++) {
-		rq = &vi->rq[i];
-		dim = &rq->dim;
-		qnum = rq - vi->rq;
-
-		if (!rq->dim_enabled)
-			continue;
-
-		update_moder = net_dim_get_rx_moderation(dim->mode, dim->profile_ix);
-		if (update_moder.usec != rq->intr_coal.max_usecs ||
-		    update_moder.pkts != rq->intr_coal.max_packets) {
-			err = virtnet_send_rx_ctrl_coal_vq_cmd(vi, qnum,
-							       update_moder.usec,
-							       update_moder.pkts);
-			if (err)
-				pr_debug("%s: Failed to send dim parameters on rxq%d\n",
-					 dev->name, qnum);
-		}
-	}
+	mutex_lock(&rq->dim_lock);
+	if (!rq->dim_enabled)
+		goto out;
 
+	update_moder = net_dim_get_rx_moderation(dim->mode, dim->profile_ix);
+	if (update_moder.usec != rq->intr_coal.max_usecs ||
+	    update_moder.pkts != rq->intr_coal.max_packets) {
+		err = virtnet_send_rx_ctrl_coal_vq_cmd(vi, qnum,
+						       update_moder.usec,
+						       update_moder.pkts);
+		if (err)
+			pr_debug("%s: Failed to send dim parameters on rxq%d\n",
+				 dev->name, qnum);
+	}
+out:
 	dim->state = DIM_START_MEASURE;
-	rtnl_unlock();
+	mutex_unlock(&rq->dim_lock);
 }
 
 static int virtnet_coal_params_supported(struct ethtool_coalesce *ec)
@@ -3749,11 +4574,13 @@ static int virtnet_get_per_queue_coalesce(struct net_device *dev,
 		return -EINVAL;
 
 	if (virtio_has_feature(vi->vdev, VIRTIO_NET_F_VQ_NOTF_COAL)) {
+		mutex_lock(&vi->rq[queue].dim_lock);
 		ec->rx_coalesce_usecs = vi->rq[queue].intr_coal.max_usecs;
 		ec->tx_coalesce_usecs = vi->sq[queue].intr_coal.max_usecs;
 		ec->tx_max_coalesced_frames = vi->sq[queue].intr_coal.max_packets;
 		ec->rx_max_coalesced_frames = vi->rq[queue].intr_coal.max_packets;
 		ec->use_adaptive_rx_coalesce = vi->rq[queue].dim_enabled;
+		mutex_unlock(&vi->rq[queue].dim_lock);
 	} else {
 		ec->rx_max_coalesced_frames = 1;
 
@@ -3809,11 +4636,11 @@ static int virtnet_get_rxfh(struct net_device *dev,
 
 	if (rxfh->indir) {
 		for (i = 0; i < vi->rss_indir_table_size; ++i)
-			rxfh->indir[i] = vi->ctrl->rss.indirection_table[i];
+			rxfh->indir[i] = vi->rss.indirection_table[i];
 	}
 
 	if (rxfh->key)
-		memcpy(rxfh->key, vi->ctrl->rss.key, vi->rss_key_size);
+		memcpy(rxfh->key, vi->rss.key, vi->rss_key_size);
 
 	rxfh->hfunc = ETH_RSS_HASH_TOP;
 
@@ -3837,7 +4664,7 @@ static int virtnet_set_rxfh(struct net_device *dev,
 			return -EOPNOTSUPP;
 
 		for (i = 0; i < vi->rss_indir_table_size; ++i)
-			vi->ctrl->rss.indirection_table[i] = rxfh->indir[i];
+			vi->rss.indirection_table[i] = rxfh->indir[i];
 		update = true;
 	}
 
@@ -3849,7 +4676,7 @@ static int virtnet_set_rxfh(struct net_device *dev,
 		if (!vi->has_rss && !vi->has_rss_hash_report)
 			return -EOPNOTSUPP;
 
-		memcpy(vi->ctrl->rss.key, rxfh->key, vi->rss_key_size);
+		memcpy(vi->rss.key, rxfh->key, vi->rss_key_size);
 		update = true;
 	}
 
@@ -3923,6 +4750,97 @@ static const struct ethtool_ops virtnet_ethtool_ops = {
 	.set_rxnfc = virtnet_set_rxnfc,
 };
 
+static void virtnet_get_queue_stats_rx(struct net_device *dev, int i,
+				       struct netdev_queue_stats_rx *stats)
+{
+	struct virtnet_info *vi = netdev_priv(dev);
+	struct receive_queue *rq = &vi->rq[i];
+	struct virtnet_stats_ctx ctx = {0};
+
+	virtnet_stats_ctx_init(vi, &ctx, (void *)stats, true);
+
+	virtnet_get_hw_stats(vi, &ctx, i * 2);
+	virtnet_fill_stats(vi, i * 2, &ctx, (void *)&rq->stats, true, 0);
+}
+
+static void virtnet_get_queue_stats_tx(struct net_device *dev, int i,
+				       struct netdev_queue_stats_tx *stats)
+{
+	struct virtnet_info *vi = netdev_priv(dev);
+	struct send_queue *sq = &vi->sq[i];
+	struct virtnet_stats_ctx ctx = {0};
+
+	virtnet_stats_ctx_init(vi, &ctx, (void *)stats, true);
+
+	virtnet_get_hw_stats(vi, &ctx, i * 2 + 1);
+	virtnet_fill_stats(vi, i * 2 + 1, &ctx, (void *)&sq->stats, true, 0);
+}
+
+static void virtnet_get_base_stats(struct net_device *dev,
+				   struct netdev_queue_stats_rx *rx,
+				   struct netdev_queue_stats_tx *tx)
+{
+	struct virtnet_info *vi = netdev_priv(dev);
+
+	/* The queue stats of the virtio-net will not be reset. So here we
+	 * return 0.
+	 */
+	rx->bytes = 0;
+	rx->packets = 0;
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_BASIC) {
+		rx->hw_drops = 0;
+		rx->hw_drop_overruns = 0;
+	}
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_CSUM) {
+		rx->csum_unnecessary = 0;
+		rx->csum_none = 0;
+		rx->csum_bad = 0;
+	}
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_GSO) {
+		rx->hw_gro_packets = 0;
+		rx->hw_gro_bytes = 0;
+		rx->hw_gro_wire_packets = 0;
+		rx->hw_gro_wire_bytes = 0;
+	}
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_RX_SPEED)
+		rx->hw_drop_ratelimits = 0;
+
+	tx->bytes = 0;
+	tx->packets = 0;
+	tx->stop = 0;
+	tx->wake = 0;
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_BASIC) {
+		tx->hw_drops = 0;
+		tx->hw_drop_errors = 0;
+	}
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_CSUM) {
+		tx->csum_none = 0;
+		tx->needs_csum = 0;
+	}
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_GSO) {
+		tx->hw_gso_packets = 0;
+		tx->hw_gso_bytes = 0;
+		tx->hw_gso_wire_packets = 0;
+		tx->hw_gso_wire_bytes = 0;
+	}
+
+	if (vi->device_stats_cap & VIRTIO_NET_STATS_TYPE_TX_SPEED)
+		tx->hw_drop_ratelimits = 0;
+}
+
+static const struct netdev_stat_ops virtnet_stat_ops = {
+	.get_queue_stats_rx	= virtnet_get_queue_stats_rx,
+	.get_queue_stats_tx	= virtnet_get_queue_stats_tx,
+	.get_base_stats		= virtnet_get_base_stats,
+};
+
 static void virtnet_freeze_down(struct virtio_device *vdev)
 {
 	struct virtnet_info *vi = vdev->priv;
@@ -3969,10 +4887,16 @@ static int virtnet_restore_up(struct virtio_device *vdev)
 
 static int virtnet_set_guest_offloads(struct virtnet_info *vi, u64 offloads)
 {
+	__virtio64 *_offloads __free(kfree) = NULL;
 	struct scatterlist sg;
-	vi->ctrl->offloads = cpu_to_virtio64(vi->vdev, offloads);
 
-	sg_init_one(&sg, &vi->ctrl->offloads, sizeof(vi->ctrl->offloads));
+	_offloads = kzalloc(sizeof(*_offloads), GFP_KERNEL);
+	if (!_offloads)
+		return -ENOMEM;
+
+	*_offloads = cpu_to_virtio64(vi->vdev, offloads);
+
+	sg_init_one(&sg, _offloads, sizeof(*_offloads));
 
 	if (!virtnet_send_command(vi, VIRTIO_NET_CTRL_GUEST_OFFLOADS,
 				  VIRTIO_NET_CTRL_GUEST_OFFLOADS_SET, &sg)) {
@@ -4072,7 +4996,7 @@ static int virtnet_xdp_set(struct net_device *dev, struct bpf_prog *prog,
 		synchronize_net();
 	}
 
-	err = _virtnet_set_queues(vi, curr_qp + xdp_qp);
+	err = virtnet_set_queues(vi, curr_qp + xdp_qp);
 	if (err)
 		goto err;
 	netif_set_real_num_rx_queues(dev, curr_qp + xdp_qp);
@@ -4174,9 +5098,9 @@ static int virtnet_set_features(struct net_device *dev,
 
 	if ((dev->features ^ features) & NETIF_F_RXHASH) {
 		if (features & NETIF_F_RXHASH)
-			vi->ctrl->rss.hash_types = vi->rss_hash_types_saved;
+			vi->rss.hash_types = vi->rss_hash_types_saved;
 		else
-			vi->ctrl->rss.hash_types = VIRTIO_NET_HASH_REPORT_NONE;
+			vi->rss.hash_types = VIRTIO_NET_HASH_REPORT_NONE;
 
 		if (!virtnet_commit_rss_command(vi))
 			return -EINVAL;
@@ -4305,7 +5229,7 @@ static void free_receive_page_frags(struct virtnet_info *vi)
 	int i;
 	for (i = 0; i < vi->max_queue_pairs; i++)
 		if (vi->rq[i].alloc_frag.page) {
-			if (vi->rq[i].do_dma && vi->rq[i].last_dma)
+			if (vi->rq[i].last_dma)
 				virtnet_rq_unmap(&vi->rq[i], vi->rq[i].last_dma, 0);
 			put_page(vi->rq[i].alloc_frag.page);
 		}
@@ -4488,6 +5412,7 @@ static int virtnet_alloc_queues(struct virtnet_info *vi)
 
 		u64_stats_init(&vi->rq[i].stats.syncp);
 		u64_stats_init(&vi->sq[i].stats.syncp);
+		mutex_init(&vi->rq[i].dim_lock);
 	}
 
 	return 0;
@@ -4655,6 +5580,48 @@ static void virtnet_set_big_packets(struct virtnet_info *vi, const int mtu)
 	}
 }
 
+#define VIRTIO_NET_HASH_REPORT_MAX_TABLE      10
+static enum xdp_rss_hash_type
+virtnet_xdp_rss_type[VIRTIO_NET_HASH_REPORT_MAX_TABLE] = {
+	[VIRTIO_NET_HASH_REPORT_NONE] = XDP_RSS_TYPE_NONE,
+	[VIRTIO_NET_HASH_REPORT_IPv4] = XDP_RSS_TYPE_L3_IPV4,
+	[VIRTIO_NET_HASH_REPORT_TCPv4] = XDP_RSS_TYPE_L4_IPV4_TCP,
+	[VIRTIO_NET_HASH_REPORT_UDPv4] = XDP_RSS_TYPE_L4_IPV4_UDP,
+	[VIRTIO_NET_HASH_REPORT_IPv6] = XDP_RSS_TYPE_L3_IPV6,
+	[VIRTIO_NET_HASH_REPORT_TCPv6] = XDP_RSS_TYPE_L4_IPV6_TCP,
+	[VIRTIO_NET_HASH_REPORT_UDPv6] = XDP_RSS_TYPE_L4_IPV6_UDP,
+	[VIRTIO_NET_HASH_REPORT_IPv6_EX] = XDP_RSS_TYPE_L3_IPV6_EX,
+	[VIRTIO_NET_HASH_REPORT_TCPv6_EX] = XDP_RSS_TYPE_L4_IPV6_TCP_EX,
+	[VIRTIO_NET_HASH_REPORT_UDPv6_EX] = XDP_RSS_TYPE_L4_IPV6_UDP_EX
+};
+
+static int virtnet_xdp_rx_hash(const struct xdp_md *_ctx, u32 *hash,
+			       enum xdp_rss_hash_type *rss_type)
+{
+	const struct xdp_buff *xdp = (void *)_ctx;
+	struct virtio_net_hdr_v1_hash *hdr_hash;
+	struct virtnet_info *vi;
+	u16 hash_report;
+
+	if (!(xdp->rxq->dev->features & NETIF_F_RXHASH))
+		return -ENODATA;
+
+	vi = netdev_priv(xdp->rxq->dev);
+	hdr_hash = (struct virtio_net_hdr_v1_hash *)(xdp->data - vi->hdr_len);
+	hash_report = __le16_to_cpu(hdr_hash->hash_report);
+
+	if (hash_report >= VIRTIO_NET_HASH_REPORT_MAX_TABLE)
+		hash_report = VIRTIO_NET_HASH_REPORT_NONE;
+
+	*rss_type = virtnet_xdp_rss_type[hash_report];
+	*hash = __le32_to_cpu(hdr_hash->hash_value);
+	return 0;
+}
+
+static const struct xdp_metadata_ops virtnet_xdp_metadata_ops = {
+	.xmo_rx_hash			= virtnet_xdp_rx_hash,
+};
+
 static int virtnet_probe(struct virtio_device *vdev)
 {
 	int i, err = -ENOMEM;
@@ -4684,6 +5651,7 @@ static int virtnet_probe(struct virtio_device *vdev)
 	dev->priv_flags |= IFF_UNICAST_FLT | IFF_LIVE_ADDR_CHANGE |
 			   IFF_TX_SKB_NO_LINEAR;
 	dev->netdev_ops = &virtnet_netdev;
+	dev->stat_ops = &virtnet_stat_ops;
 	dev->features = NETIF_F_HIGHDMA;
 
 	dev->ethtool_ops = &virtnet_ethtool_ops;
@@ -4791,6 +5759,7 @@ static int virtnet_probe(struct virtio_device *vdev)
 				  VIRTIO_NET_RSS_HASH_TYPE_UDP_EX);
 
 		dev->hw_features |= NETIF_F_RXHASH;
+		dev->xdp_metadata_ops = &virtnet_xdp_metadata_ops;
 	}
 
 	if (vi->has_rss_hash_report)
@@ -4808,6 +5777,8 @@ static int virtnet_probe(struct virtio_device *vdev)
 	if (virtio_has_feature(vdev, VIRTIO_NET_F_CTRL_VQ))
 		vi->has_cvq = true;
 
+	mutex_init(&vi->cvq_lock);
+
 	if (virtio_has_feature(vdev, VIRTIO_NET_F_MTU)) {
 		mtu = virtio_cread16(vdev,
 				     offsetof(struct virtio_net_config,
@@ -4899,7 +5870,7 @@ static int virtnet_probe(struct virtio_device *vdev)
 
 	virtio_device_ready(vdev);
 
-	_virtnet_set_queues(vi, vi->curr_queue_pairs);
+	virtnet_set_queues(vi, vi->curr_queue_pairs);
 
 	/* a random MAC address has been assigned, notify the device.
 	 * We don't fail probe if VIRTIO_NET_F_CTRL_MAC_ADDR is not there
@@ -4919,6 +5890,33 @@ static int virtnet_probe(struct virtio_device *vdev)
 		}
 	}
 
+	if (virtio_has_feature(vi->vdev, VIRTIO_NET_F_DEVICE_STATS)) {
+		struct virtio_net_stats_capabilities *stats_cap  __free(kfree) = NULL;
+		struct scatterlist sg;
+		__le64 v;
+
+		stats_cap = kzalloc(sizeof(*stats_cap), GFP_KERNEL);
+		if (!stats_cap) {
+			rtnl_unlock();
+			err = -ENOMEM;
+			goto free_unregister_netdev;
+		}
+
+		sg_init_one(&sg, stats_cap, sizeof(*stats_cap));
+
+		if (!virtnet_send_command_reply(vi, VIRTIO_NET_CTRL_STATS,
+						VIRTIO_NET_CTRL_STATS_QUERY,
+						NULL, &sg)) {
+			pr_debug("virtio_net: fail to get stats capability\n");
+			rtnl_unlock();
+			err = -EINVAL;
+			goto free_unregister_netdev;
+		}
+
+		v = stats_cap->supported_stats_types[0];
+		vi->device_stats_cap = le64_to_cpu(v);
+	}
+
 	rtnl_unlock();
 
 	err = virtnet_cpu_notif_add(vi);
@@ -5047,7 +6045,7 @@ static struct virtio_device_id id_table[] = {
 	VIRTIO_NET_F_SPEED_DUPLEX, VIRTIO_NET_F_STANDBY, \
 	VIRTIO_NET_F_RSS, VIRTIO_NET_F_HASH_REPORT, VIRTIO_NET_F_NOTF_COAL, \
 	VIRTIO_NET_F_VQ_NOTF_COAL, \
-	VIRTIO_NET_F_GUEST_HDRLEN
+	VIRTIO_NET_F_GUEST_HDRLEN, VIRTIO_NET_F_DEVICE_STATS
 
 static unsigned int features[] = {
 	VIRTNET_FEATURES,
@@ -5065,7 +6063,6 @@ static struct virtio_driver virtio_net_driver = {
 	.feature_table_legacy = features_legacy,
 	.feature_table_size_legacy = ARRAY_SIZE(features_legacy),
 	.driver.name =	KBUILD_MODNAME,
-	.driver.owner =	THIS_MODULE,
 	.id_table =	id_table,
 	.validate =	virtnet_validate,
 	.probe =	virtnet_probe,