Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next

Pull networking updates from David Miller:

 1) Add WireGuard

 2) Add HE and TWT support to ath11k driver, from John Crispin.

 3) Add ESP in TCP encapsulation support, from Sabrina Dubroca.

 4) Add variable window congestion control to TIPC, from Jon Maloy.

 5) Add BCM84881 PHY driver, from Russell King.

 6) Start adding netlink support for ethtool operations, from Michal
    Kubecek.

 7) Add XDP drop and TX action support to ena driver, from Sameeh
    Jubran.

 8) Add new ipv4 route notifications so that mlxsw driver does not have
    to handle identical routes itself. From Ido Schimmel.

 9) Add BPF dynamic program extensions, from Alexei Starovoitov.

10) Support RX and TX timestamping in igc, from Vinicius Costa Gomes.

11) Add support for macsec HW offloading, from Antoine Tenart.

12) Add initial support for MPTCP protocol, from Christoph Paasch,
    Matthieu Baerts, Florian Westphal, Peter Krystad, and many others.

13) Add Octeontx2 PF support, from Sunil Goutham, Geetha sowjanya, Linu
    Cherian, and others.

* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1469 commits)
  net: phy: add default ARCH_BCM_IPROC for MDIO_BCM_IPROC
  udp: segment looped gso packets correctly
  netem: change mailing list
  qed: FW 8.42.2.0 debug features
  qed: rt init valid initialization changed
  qed: Debug feature: ilt and mdump
  qed: FW 8.42.2.0 Add fw overlay feature
  qed: FW 8.42.2.0 HSI changes
  qed: FW 8.42.2.0 iscsi/fcoe changes
  qed: Add abstraction for different hsi values per chip
  qed: FW 8.42.2.0 Additional ll2 type
  qed: Use dmae to write to widebus registers in fw_funcs
  qed: FW 8.42.2.0 Parser offsets modified
  qed: FW 8.42.2.0 Queue Manager changes
  qed: FW 8.42.2.0 Expose new registers and change windows
  qed: FW 8.42.2.0 Internal ram offsets modifications
  MAINTAINERS: Add entry for Marvell OcteonTX2 Physical Function driver
  Documentation: net: octeontx2: Add RVU HW and drivers overview
  octeontx2-pf: ethtool RSS config support
  octeontx2-pf: Add basic ethtool support
  ...

17 files changed, 886 insertions, 649 deletions

diff --git a/drivers/net/ethernet/hisilicon/hns3/Makefile b/drivers/net/ethernet/hisilicon/hns3/Makefile
index d01bf536eb86..7aa2fac76c5e 100644
--- a/drivers/net/ethernet/hisilicon/hns3/Makefile
+++ b/drivers/net/ethernet/hisilicon/hns3/Makefile
@@ -3,6 +3,8 @@
 # Makefile for the HISILICON network device drivers.
 #
 
+ccflags-y += -I$(srctree)/$(src)
+
 obj-$(CONFIG_HNS3) += hns3pf/
 obj-$(CONFIG_HNS3) += hns3vf/
 
diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
index 3b5e2d7251e7..a3e4081b84ba 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
@@ -164,11 +164,7 @@ enum hnae3_reset_type {
 	HNAE3_IMP_RESET,
 	HNAE3_UNKNOWN_RESET,
 	HNAE3_NONE_RESET,
-};
-
-enum hnae3_flr_state {
-	HNAE3_FLR_DOWN,
-	HNAE3_FLR_DONE,
+	HNAE3_MAX_RESET,
 };
 
 enum hnae3_port_base_vlan_state {
@@ -575,8 +571,7 @@ struct hnae3_ae_algo {
 	const struct pci_device_id *pdev_id_table;
 };
 
-#define HNAE3_INT_NAME_EXT_LEN    32	 /* Max extra information length */
-#define HNAE3_INT_NAME_LEN        (IFNAMSIZ + HNAE3_INT_NAME_EXT_LEN)
+#define HNAE3_INT_NAME_LEN        32
 #define HNAE3_ITR_COUNTDOWN_START 100
 
 struct hnae3_tc_info {
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c
index 6b328a259efc..1d4ffc5f408a 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c
@@ -176,7 +176,7 @@ static int hns3_dbg_bd_info(struct hnae3_handle *h, const char *cmd_buf)
 		return -EINVAL;
 	}
 
-	ring  = &priv->ring[q_num];
+	ring = &priv->ring[q_num];
 	value = readl_relaxed(ring->tqp->io_base + HNS3_RING_TX_RING_TAIL_REG);
 	tx_index = (cnt == 1) ? value : tx_index;
 
@@ -209,10 +209,10 @@ static int hns3_dbg_bd_info(struct hnae3_handle *h, const char *cmd_buf)
 		 le16_to_cpu(tx_desc->tx.bdtp_fe_sc_vld_ra_ri));
 	dev_info(dev, "(TX)mss: %u\n", le16_to_cpu(tx_desc->tx.mss));
 
-	ring  = &priv->ring[q_num + h->kinfo.num_tqps];
+	ring = &priv->ring[q_num + h->kinfo.num_tqps];
 	value = readl_relaxed(ring->tqp->io_base + HNS3_RING_RX_RING_TAIL_REG);
 	rx_index = (cnt == 1) ? value : tx_index;
-	rx_desc	 = &ring->desc[rx_index];
+	rx_desc = &ring->desc[rx_index];
 
 	addr = le64_to_cpu(rx_desc->addr);
 	dev_info(dev, "RX Queue Num: %u, BD Index: %u\n", q_num, rx_index);
@@ -297,8 +297,8 @@ static ssize_t hns3_dbg_cmd_read(struct file *filp, char __user *buffer,
 	if (!buf)
 		return -ENOMEM;
 
-	len = snprintf(buf, HNS3_DBG_READ_LEN, "%s\n",
-		       "Please echo help to cmd to get help information");
+	len = scnprintf(buf, HNS3_DBG_READ_LEN, "%s\n",
+			"Please echo help to cmd to get help information");
 	uncopy_bytes = copy_to_user(buffer, buf, len);
 
 	kfree(buf);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index b3deb5e5ce29..acb796cc10d0 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -24,6 +24,12 @@
 
 #include "hnae3.h"
 #include "hns3_enet.h"
+/* All hns3 tracepoints are defined by the include below, which
+ * must be included exactly once across the whole kernel with
+ * CREATE_TRACE_POINTS defined
+ */
+#define CREATE_TRACE_POINTS
+#include "hns3_trace.h"
 
 #define hns3_set_field(origin, shift, val)	((origin) |= ((val) << (shift)))
 #define hns3_tx_bd_count(S)	DIV_ROUND_UP(S, HNS3_MAX_BD_SIZE)
@@ -129,18 +135,21 @@ static int hns3_nic_init_irq(struct hns3_nic_priv *priv)
 			continue;
 
 		if (tqp_vectors->tx_group.ring && tqp_vectors->rx_group.ring) {
-			snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN - 1,
-				 "%s-%s-%d", priv->netdev->name, "TxRx",
-				 txrx_int_idx++);
+			snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN,
+				 "%s-%s-%s-%d", hns3_driver_name,
+				 pci_name(priv->ae_handle->pdev),
+				 "TxRx", txrx_int_idx++);
 			txrx_int_idx++;
 		} else if (tqp_vectors->rx_group.ring) {
-			snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN - 1,
-				 "%s-%s-%d", priv->netdev->name, "Rx",
-				 rx_int_idx++);
+			snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN,
+				 "%s-%s-%s-%d", hns3_driver_name,
+				 pci_name(priv->ae_handle->pdev),
+				 "Rx", rx_int_idx++);
 		} else if (tqp_vectors->tx_group.ring) {
-			snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN - 1,
-				 "%s-%s-%d", priv->netdev->name, "Tx",
-				 tx_int_idx++);
+			snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN,
+				 "%s-%s-%s-%d", hns3_driver_name,
+				 pci_name(priv->ae_handle->pdev),
+				 "Tx", tx_int_idx++);
 		} else {
 			/* Skip this unused q_vector */
 			continue;
@@ -157,6 +166,8 @@ static int hns3_nic_init_irq(struct hns3_nic_priv *priv)
 			return ret;
 		}
 
+		disable_irq(tqp_vectors->vector_irq);
+
 		irq_set_affinity_hint(tqp_vectors->vector_irq,
 				      &tqp_vectors->affinity_mask);
 
@@ -175,6 +186,7 @@ static void hns3_mask_vector_irq(struct hns3_enet_tqp_vector *tqp_vector,
 static void hns3_vector_enable(struct hns3_enet_tqp_vector *tqp_vector)
 {
 	napi_enable(&tqp_vector->napi);
+	enable_irq(tqp_vector->vector_irq);
 
 	/* enable vector */
 	hns3_mask_vector_irq(tqp_vector, 1);
@@ -374,18 +386,6 @@ static int hns3_nic_net_up(struct net_device *netdev)
 	if (ret)
 		return ret;
 
-	/* the device can work without cpu rmap, only aRFS needs it */
-	ret = hns3_set_rx_cpu_rmap(netdev);
-	if (ret)
-		netdev_warn(netdev, "set rx cpu rmap fail, ret=%d!\n", ret);
-
-	/* get irq resource for all vectors */
-	ret = hns3_nic_init_irq(priv);
-	if (ret) {
-		netdev_err(netdev, "init irq failed! ret=%d\n", ret);
-		goto free_rmap;
-	}
-
 	clear_bit(HNS3_NIC_STATE_DOWN, &priv->state);
 
 	/* enable the vectors */
@@ -398,22 +398,15 @@ static int hns3_nic_net_up(struct net_device *netdev)
 
 	/* start the ae_dev */
 	ret = h->ae_algo->ops->start ? h->ae_algo->ops->start(h) : 0;
-	if (ret)
-		goto out_start_err;
-
-	return 0;
-
-out_start_err:
-	set_bit(HNS3_NIC_STATE_DOWN, &priv->state);
-	while (j--)
-		hns3_tqp_disable(h->kinfo.tqp[j]);
+	if (ret) {
+		set_bit(HNS3_NIC_STATE_DOWN, &priv->state);
+		while (j--)
+			hns3_tqp_disable(h->kinfo.tqp[j]);
 
-	for (j = i - 1; j >= 0; j--)
-		hns3_vector_disable(&priv->tqp_vector[j]);
+		for (j = i - 1; j >= 0; j--)
+			hns3_vector_disable(&priv->tqp_vector[j]);
+	}
 
-	hns3_nic_uninit_irq(priv);
-free_rmap:
-	hns3_free_rx_cpu_rmap(netdev);
 	return ret;
 }
 
@@ -510,11 +503,6 @@ static void hns3_nic_net_down(struct net_device *netdev)
 	if (ops->stop)
 		ops->stop(priv->ae_handle);
 
-	hns3_free_rx_cpu_rmap(netdev);
-
-	/* free irq resources */
-	hns3_nic_uninit_irq(priv);
-
 	/* delay ring buffer clearing to hns3_reset_notify_uninit_enet
 	 * during reset process, because driver may not be able
 	 * to disable the ring through firmware when downing the netdev.
@@ -736,6 +724,8 @@ static int hns3_set_tso(struct sk_buff *skb, u32 *paylen,
 	/* get MSS for TSO */
 	*mss = skb_shinfo(skb)->gso_size;
 
+	trace_hns3_tso(skb);
+
 	return 0;
 }
 
@@ -1140,6 +1130,7 @@ static int hns3_fill_desc(struct hns3_enet_ring *ring, void *priv,
 		desc->tx.bdtp_fe_sc_vld_ra_ri =
 			cpu_to_le16(BIT(HNS3_TXD_VLD_B));
 
+		trace_hns3_tx_desc(ring, ring->next_to_use);
 		ring_ptr_move_fw(ring, next_to_use);
 		return HNS3_LIKELY_BD_NUM;
 	}
@@ -1163,6 +1154,7 @@ static int hns3_fill_desc(struct hns3_enet_ring *ring, void *priv,
 		desc->tx.bdtp_fe_sc_vld_ra_ri =
 				cpu_to_le16(BIT(HNS3_TXD_VLD_B));
 
+		trace_hns3_tx_desc(ring, ring->next_to_use);
 		/* move ring pointer to next */
 		ring_ptr_move_fw(ring, next_to_use);
 
@@ -1288,6 +1280,14 @@ static bool hns3_skb_need_linearized(struct sk_buff *skb, unsigned int *bd_size,
 	return false;
 }
 
+void hns3_shinfo_pack(struct skb_shared_info *shinfo, __u32 *size)
+{
+	int i = 0;
+
+	for (i = 0; i < MAX_SKB_FRAGS; i++)
+		size[i] = skb_frag_size(&shinfo->frags[i]);
+}
+
 static int hns3_nic_maybe_stop_tx(struct hns3_enet_ring *ring,
 				  struct net_device *netdev,
 				  struct sk_buff *skb)
@@ -1299,8 +1299,10 @@ static int hns3_nic_maybe_stop_tx(struct hns3_enet_ring *ring,
 	bd_num = hns3_tx_bd_num(skb, bd_size);
 	if (unlikely(bd_num > HNS3_MAX_NON_TSO_BD_NUM)) {
 		if (bd_num <= HNS3_MAX_TSO_BD_NUM && skb_is_gso(skb) &&
-		    !hns3_skb_need_linearized(skb, bd_size, bd_num))
+		    !hns3_skb_need_linearized(skb, bd_size, bd_num)) {
+			trace_hns3_over_8bd(skb);
 			goto out;
+		}
 
 		if (__skb_linearize(skb))
 			return -ENOMEM;
@@ -1308,8 +1310,10 @@ static int hns3_nic_maybe_stop_tx(struct hns3_enet_ring *ring,
 		bd_num = hns3_tx_bd_count(skb->len);
 		if ((skb_is_gso(skb) && bd_num > HNS3_MAX_TSO_BD_NUM) ||
 		    (!skb_is_gso(skb) &&
-		     bd_num > HNS3_MAX_NON_TSO_BD_NUM))
+		     bd_num > HNS3_MAX_NON_TSO_BD_NUM)) {
+			trace_hns3_over_8bd(skb);
 			return -ENOMEM;
+		}
 
 		u64_stats_update_begin(&ring->syncp);
 		ring->stats.tx_copy++;
@@ -1454,6 +1458,7 @@ out:
 					(ring->desc_num - 1);
 	ring->desc[pre_ntu].tx.bdtp_fe_sc_vld_ra_ri |=
 				cpu_to_le16(BIT(HNS3_TXD_FE_B));
+	trace_hns3_tx_desc(ring, pre_ntu);
 
 	/* Complete translate all packets */
 	dev_queue = netdev_get_tx_queue(netdev, ring->queue_index);
@@ -1562,6 +1567,37 @@ static int hns3_nic_set_features(struct net_device *netdev,
 	return 0;
 }
 
+static netdev_features_t hns3_features_check(struct sk_buff *skb,
+					     struct net_device *dev,
+					     netdev_features_t features)
+{
+#define HNS3_MAX_HDR_LEN	480U
+#define HNS3_MAX_L4_HDR_LEN	60U
+
+	size_t len;
+
+	if (skb->ip_summed != CHECKSUM_PARTIAL)
+		return features;
+
+	if (skb->encapsulation)
+		len = skb_inner_transport_header(skb) - skb->data;
+	else
+		len = skb_transport_header(skb) - skb->data;
+
+	/* Assume L4 is 60 byte as TCP is the only protocol with a
+	 * a flexible value, and it's max len is 60 bytes.
+	 */
+	len += HNS3_MAX_L4_HDR_LEN;
+
+	/* Hardware only supports checksum on the skb with a max header
+	 * len of 480 bytes.
+	 */
+	if (len > HNS3_MAX_HDR_LEN)
+		features &= ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK);
+
+	return features;
+}
+
 static void hns3_nic_get_stats64(struct net_device *netdev,
 				 struct rtnl_link_stats64 *stats)
 {
@@ -1875,7 +1911,7 @@ static bool hns3_get_tx_timeo_queue_info(struct net_device *ndev)
 	return true;
 }
 
-static void hns3_nic_net_timeout(struct net_device *ndev)
+static void hns3_nic_net_timeout(struct net_device *ndev, unsigned int txqueue)
 {
 	struct hns3_nic_priv *priv = netdev_priv(ndev);
 	struct hnae3_handle *h = priv->ae_handle;
@@ -1976,6 +2012,7 @@ static const struct net_device_ops hns3_nic_netdev_ops = {
 	.ndo_do_ioctl		= hns3_nic_do_ioctl,
 	.ndo_change_mtu		= hns3_nic_change_mtu,
 	.ndo_set_features	= hns3_nic_set_features,
+	.ndo_features_check	= hns3_features_check,
 	.ndo_get_stats64	= hns3_nic_get_stats64,
 	.ndo_setup_tc		= hns3_nic_setup_tc,
 	.ndo_set_rx_mode	= hns3_nic_set_rx_mode,
@@ -2057,10 +2094,8 @@ static int hns3_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	int ret;
 
 	ae_dev = devm_kzalloc(&pdev->dev, sizeof(*ae_dev), GFP_KERNEL);
-	if (!ae_dev) {
-		ret = -ENOMEM;
-		return ret;
-	}
+	if (!ae_dev)
+		return -ENOMEM;
 
 	ae_dev->pdev = pdev;
 	ae_dev->flag = ent->driver_data;
@@ -2503,8 +2538,8 @@ void hns3_clean_tx_ring(struct hns3_enet_ring *ring)
 	rmb(); /* Make sure head is ready before touch any data */
 
 	if (unlikely(!is_valid_clean_head(ring, head))) {
-		netdev_err(netdev, "wrong head (%d, %d-%d)\n", head,
-			   ring->next_to_use, ring->next_to_clean);
+		hns3_rl_err(netdev, "wrong head (%d, %d-%d)\n", head,
+			    ring->next_to_use, ring->next_to_clean);
 
 		u64_stats_update_begin(&ring->syncp);
 		ring->stats.io_err_cnt++;
@@ -2590,6 +2625,12 @@ static void hns3_nic_alloc_rx_buffers(struct hns3_enet_ring *ring,
 	writel_relaxed(i, ring->tqp->io_base + HNS3_RING_RX_RING_HEAD_REG);
 }
 
+static bool hns3_page_is_reusable(struct page *page)
+{
+	return page_to_nid(page) == numa_mem_id() &&
+		!page_is_pfmemalloc(page);
+}
+
 static void hns3_nic_reuse_page(struct sk_buff *skb, int i,
 				struct hns3_enet_ring *ring, int pull_len,
 				struct hns3_desc_cb *desc_cb)
@@ -2604,7 +2645,7 @@ static void hns3_nic_reuse_page(struct sk_buff *skb, int i,
 	/* Avoid re-using remote pages, or the stack is still using the page
 	 * when page_offset rollback to zero, flag default unreuse
 	 */
-	if (unlikely(page_to_nid(desc_cb->priv) != numa_mem_id()) ||
+	if (unlikely(!hns3_page_is_reusable(desc_cb->priv)) ||
 	    (!desc_cb->page_offset && page_count(desc_cb->priv) > 1))
 		return;
 
@@ -2674,6 +2715,9 @@ static int hns3_gro_complete(struct sk_buff *skb, u32 l234info)
 	skb->csum_start = (unsigned char *)th - skb->head;
 	skb->csum_offset = offsetof(struct tcphdr, check);
 	skb->ip_summed = CHECKSUM_PARTIAL;
+
+	trace_hns3_gro(skb);
+
 	return 0;
 }
 
@@ -2794,7 +2838,6 @@ static bool hns3_parse_vlan_tag(struct hns3_enet_ring *ring,
 static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length,
 			  unsigned char *va)
 {
-#define HNS3_NEED_ADD_FRAG	1
 	struct hns3_desc_cb *desc_cb = &ring->desc_cb[ring->next_to_clean];
 	struct net_device *netdev = ring_to_netdev(ring);
 	struct sk_buff *skb;
@@ -2811,6 +2854,7 @@ static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length,
 		return -ENOMEM;
 	}
 
+	trace_hns3_rx_desc(ring);
 	prefetchw(skb->data);
 
 	ring->pending_buf = 1;
@@ -2820,7 +2864,7 @@ static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length,
 		memcpy(__skb_put(skb, length), va, ALIGN(length, sizeof(long)));
 
 		/* We can reuse buffer as-is, just make sure it is local */
-		if (likely(page_to_nid(desc_cb->priv) == numa_mem_id()))
+		if (likely(hns3_page_is_reusable(desc_cb->priv)))
 			desc_cb->reuse_flag = 1;
 		else /* This page cannot be reused so discard it */
 			put_page(desc_cb->priv);
@@ -2838,33 +2882,19 @@ static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length,
 			    desc_cb);
 	ring_ptr_move_fw(ring, next_to_clean);
 
-	return HNS3_NEED_ADD_FRAG;
+	return 0;
 }
 
-static int hns3_add_frag(struct hns3_enet_ring *ring, struct hns3_desc *desc,
-			 bool pending)
+static int hns3_add_frag(struct hns3_enet_ring *ring)
 {
 	struct sk_buff *skb = ring->skb;
 	struct sk_buff *head_skb = skb;
 	struct sk_buff *new_skb;
 	struct hns3_desc_cb *desc_cb;
-	struct hns3_desc *pre_desc;
+	struct hns3_desc *desc;
 	u32 bd_base_info;
-	int pre_bd;
 
-	/* if there is pending bd, the SW param next_to_clean has moved
-	 * to next and the next is NULL
-	 */
-	if (pending) {
-		pre_bd = (ring->next_to_clean - 1 + ring->desc_num) %
-			 ring->desc_num;
-		pre_desc = &ring->desc[pre_bd];
-		bd_base_info = le32_to_cpu(pre_desc->rx.bd_base_info);
-	} else {
-		bd_base_info = le32_to_cpu(desc->rx.bd_base_info);
-	}
-
-	while (!(bd_base_info & BIT(HNS3_RXD_FE_B))) {
+	do {
 		desc = &ring->desc[ring->next_to_clean];
 		desc_cb = &ring->desc_cb[ring->next_to_clean];
 		bd_base_info = le32_to_cpu(desc->rx.bd_base_info);
@@ -2899,9 +2929,10 @@ static int hns3_add_frag(struct hns3_enet_ring *ring, struct hns3_desc *desc,
 		}
 
 		hns3_nic_reuse_page(skb, ring->frag_num++, ring, 0, desc_cb);
+		trace_hns3_rx_desc(ring);
 		ring_ptr_move_fw(ring, next_to_clean);
 		ring->pending_buf++;
-	}
+	} while (!(bd_base_info & BIT(HNS3_RXD_FE_B)));
 
 	return 0;
 }
@@ -3069,28 +3100,23 @@ static int hns3_handle_rx_bd(struct hns3_enet_ring *ring)
 
 		if (ret < 0) /* alloc buffer fail */
 			return ret;
-		if (ret > 0) { /* need add frag */
-			ret = hns3_add_frag(ring, desc, false);
+		if (!(bd_base_info & BIT(HNS3_RXD_FE_B))) { /* need add frag */
+			ret = hns3_add_frag(ring);
 			if (ret)
 				return ret;
-
-			/* As the head data may be changed when GRO enable, copy
-			 * the head data in after other data rx completed
-			 */
-			memcpy(skb->data, ring->va,
-			       ALIGN(ring->pull_len, sizeof(long)));
 		}
 	} else {
-		ret = hns3_add_frag(ring, desc, true);
+		ret = hns3_add_frag(ring);
 		if (ret)
 			return ret;
+	}
 
-		/* As the head data may be changed when GRO enable, copy
-		 * the head data in after other data rx completed
-		 */
+	/* As the head data may be changed when GRO enable, copy
+	 * the head data in after other data rx completed
+	 */
+	if (skb->len > HNS3_RX_HEAD_SIZE)
 		memcpy(skb->data, ring->va,
 		       ALIGN(ring->pull_len, sizeof(long)));
-	}
 
 	ret = hns3_handle_bdinfo(ring, skb);
 	if (unlikely(ret)) {
@@ -3596,26 +3622,25 @@ static void hns3_nic_uninit_vector_data(struct hns3_nic_priv *priv)
 		if (!tqp_vector->rx_group.ring && !tqp_vector->tx_group.ring)
 			continue;
 
-		hns3_get_vector_ring_chain(tqp_vector, &vector_ring_chain);
+		/* Since the mapping can be overwritten, when fail to get the
+		 * chain between vector and ring, we should go on to deal with
+		 * the remaining options.
+		 */
+		if (hns3_get_vector_ring_chain(tqp_vector, &vector_ring_chain))
+			dev_warn(priv->dev, "failed to get ring chain\n");
 
 		h->ae_algo->ops->unmap_ring_from_vector(h,
 			tqp_vector->vector_irq, &vector_ring_chain);
 
 		hns3_free_vector_ring_chain(tqp_vector, &vector_ring_chain);
 
-		if (tqp_vector->irq_init_flag == HNS3_VECTOR_INITED) {
-			irq_set_affinity_hint(tqp_vector->vector_irq, NULL);
-			free_irq(tqp_vector->vector_irq, tqp_vector);
-			tqp_vector->irq_init_flag = HNS3_VECTOR_NOT_INITED;
-		}
-
 		hns3_clear_ring_group(&tqp_vector->rx_group);
 		hns3_clear_ring_group(&tqp_vector->tx_group);
 		netif_napi_del(&priv->tqp_vector[i].napi);
 	}
 }
 
-static int hns3_nic_dealloc_vector_data(struct hns3_nic_priv *priv)
+static void hns3_nic_dealloc_vector_data(struct hns3_nic_priv *priv)
 {
 	struct hnae3_handle *h = priv->ae_handle;
 	struct pci_dev *pdev = h->pdev;
@@ -3627,11 +3652,10 @@ static int hns3_nic_dealloc_vector_data(struct hns3_nic_priv *priv)
 		tqp_vector = &priv->tqp_vector[i];
 		ret = h->ae_algo->ops->put_vector(h, tqp_vector->vector_irq);
 		if (ret)
-			return ret;
+			return;
 	}
 
 	devm_kfree(&pdev->dev, priv->tqp_vector);
-	return 0;
 }
 
 static void hns3_ring_get_cfg(struct hnae3_queue *q, struct hns3_nic_priv *priv,
@@ -4030,6 +4054,18 @@ static int hns3_client_init(struct hnae3_handle *handle)
 		goto out_reg_netdev_fail;
 	}
 
+	/* the device can work without cpu rmap, only aRFS needs it */
+	ret = hns3_set_rx_cpu_rmap(netdev);
+	if (ret)
+		dev_warn(priv->dev, "set rx cpu rmap fail, ret=%d\n", ret);
+
+	ret = hns3_nic_init_irq(priv);
+	if (ret) {
+		dev_err(priv->dev, "init irq failed! ret=%d\n", ret);
+		hns3_free_rx_cpu_rmap(netdev);
+		goto out_init_irq_fail;
+	}
+
 	ret = hns3_client_start(handle);
 	if (ret) {
 		dev_err(priv->dev, "hns3_client_start fail! ret=%d\n", ret);
@@ -4051,6 +4087,9 @@ static int hns3_client_init(struct hnae3_handle *handle)
 	return ret;
 
 out_client_start:
+	hns3_free_rx_cpu_rmap(netdev);
+	hns3_nic_uninit_irq(priv);
+out_init_irq_fail:
 	unregister_netdev(netdev);
 out_reg_netdev_fail:
 	hns3_uninit_phy(netdev);
@@ -4088,15 +4127,17 @@ static void hns3_client_uninit(struct hnae3_handle *handle, bool reset)
 		goto out_netdev_free;
 	}
 
+	hns3_free_rx_cpu_rmap(netdev);
+
+	hns3_nic_uninit_irq(priv);
+
 	hns3_del_all_fd_rules(netdev, true);
 
 	hns3_clear_all_ring(handle, true);
 
 	hns3_nic_uninit_vector_data(priv);
 
-	ret = hns3_nic_dealloc_vector_data(priv);
-	if (ret)
-		netdev_err(netdev, "dealloc vector error\n");
+	hns3_nic_dealloc_vector_data(priv);
 
 	ret = hns3_uninit_all_ring(priv);
 	if (ret)
@@ -4423,17 +4464,32 @@ static int hns3_reset_notify_init_enet(struct hnae3_handle *handle)
 	if (ret)
 		goto err_uninit_vector;
 
+	/* the device can work without cpu rmap, only aRFS needs it */
+	ret = hns3_set_rx_cpu_rmap(netdev);
+	if (ret)
+		dev_warn(priv->dev, "set rx cpu rmap fail, ret=%d\n", ret);
+
+	ret = hns3_nic_init_irq(priv);
+	if (ret) {
+		dev_err(priv->dev, "init irq failed! ret=%d\n", ret);
+		hns3_free_rx_cpu_rmap(netdev);
+		goto err_init_irq_fail;
+	}
+
 	ret = hns3_client_start(handle);
 	if (ret) {
 		dev_err(priv->dev, "hns3_client_start fail! ret=%d\n", ret);
-		goto err_uninit_ring;
+		goto err_client_start_fail;
 	}
 
 	set_bit(HNS3_NIC_STATE_INITED, &priv->state);
 
 	return ret;
 
-err_uninit_ring:
+err_client_start_fail:
+	hns3_free_rx_cpu_rmap(netdev);
+	hns3_nic_uninit_irq(priv);
+err_init_irq_fail:
 	hns3_uninit_all_ring(priv);
 err_uninit_vector:
 	hns3_nic_uninit_vector_data(priv);
@@ -4483,6 +4539,8 @@ static int hns3_reset_notify_uninit_enet(struct hnae3_handle *handle)
 		return 0;
 	}
 
+	hns3_free_rx_cpu_rmap(netdev);
+	hns3_nic_uninit_irq(priv);
 	hns3_clear_all_ring(handle, true);
 	hns3_reset_tx_queue(priv->ae_handle);
 
@@ -4490,9 +4548,7 @@ static int hns3_reset_notify_uninit_enet(struct hnae3_handle *handle)
 
 	hns3_store_coal(priv);
 
-	ret = hns3_nic_dealloc_vector_data(priv);
-	if (ret)
-		netdev_err(netdev, "dealloc vector error\n");
+	hns3_nic_dealloc_vector_data(priv);
 
 	ret = hns3_uninit_all_ring(priv);
 	if (ret)
@@ -4658,7 +4714,7 @@ static int __init hns3_init_module(void)
 	pr_info("%s: %s\n", hns3_driver_name, hns3_copyright);
 
 	client.type = HNAE3_CLIENT_KNIC;
-	snprintf(client.name, HNAE3_CLIENT_NAME_LENGTH - 1, "%s",
+	snprintf(client.name, HNAE3_CLIENT_NAME_LENGTH, "%s",
 		 hns3_driver_name);
 
 	client.ops = &client_ops;
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
index 9d47abd5c37c..abefd7a179f7 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
@@ -673,4 +673,5 @@ void hns3_dbg_init(struct hnae3_handle *handle);
 void hns3_dbg_uninit(struct hnae3_handle *handle);
 void hns3_dbg_register_debugfs(const char *debugfs_dir_name);
 void hns3_dbg_unregister_debugfs(void);
+void hns3_shinfo_pack(struct skb_shared_info *shinfo, __u32 *size);
 #endif
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
index 6e0212b79438..c03856e63320 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
@@ -423,9 +423,8 @@ static void *hns3_update_strings(u8 *data, const struct hns3_stats *stats,
 			data[ETH_GSTRING_LEN - 1] = '\0';
 
 			/* first, prepend the prefix string */
-			n1 = snprintf(data, MAX_PREFIX_SIZE, "%s%d_",
-				      prefix, i);
-			n1 = min_t(uint, n1, MAX_PREFIX_SIZE - 1);
+			n1 = scnprintf(data, MAX_PREFIX_SIZE, "%s%d_",
+				       prefix, i);
 			size_left = (ETH_GSTRING_LEN - 1) - n1;
 
 			/* now, concatenate the stats string to it */
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_trace.h b/drivers/net/ethernet/hisilicon/hns3/hns3_trace.h
new file mode 100644
index 000000000000..7bddcca148a5
--- /dev/null
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_trace.h
@@ -0,0 +1,139 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+/* Copyright (c) 2018-2019 Hisilicon Limited. */
+
+/* This must be outside ifdef _HNS3_TRACE_H */
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM hns3
+
+#if !defined(_HNS3_TRACE_H_) || defined(TRACE_HEADER_MULTI_READ)
+#define _HNS3_TRACE_H_
+
+#include <linux/tracepoint.h>
+
+#define DESC_NR		(sizeof(struct hns3_desc) / sizeof(u32))
+
+DECLARE_EVENT_CLASS(hns3_skb_template,
+	TP_PROTO(struct sk_buff *skb),
+	TP_ARGS(skb),
+
+	TP_STRUCT__entry(
+		__field(unsigned int, headlen)
+		__field(unsigned int, len)
+		__field(__u8, nr_frags)
+		__field(__u8, ip_summed)
+		__field(unsigned int, hdr_len)
+		__field(unsigned short, gso_size)
+		__field(unsigned short, gso_segs)
+		__field(unsigned int, gso_type)
+		__field(bool, fraglist)
+		__array(__u32, size, MAX_SKB_FRAGS)
+	),
+
+	TP_fast_assign(
+		__entry->headlen = skb_headlen(skb);
+		__entry->len = skb->len;
+		__entry->nr_frags = skb_shinfo(skb)->nr_frags;
+		__entry->gso_size = skb_shinfo(skb)->gso_size;
+		__entry->gso_segs = skb_shinfo(skb)->gso_segs;
+		__entry->gso_type = skb_shinfo(skb)->gso_type;
+		__entry->hdr_len = skb->encapsulation ?
+		skb_inner_transport_offset(skb) + inner_tcp_hdrlen(skb) :
+		skb_transport_offset(skb) + tcp_hdrlen(skb);
+		__entry->ip_summed = skb->ip_summed;
+		__entry->fraglist = skb_has_frag_list(skb);
+		hns3_shinfo_pack(skb_shinfo(skb), __entry->size);
+	),
+
+	TP_printk(
+		"len: %u, %u, %u, cs: %u, gso: %u, %u, %x, frag(%d %u): %s",
+		__entry->headlen, __entry->len, __entry->hdr_len,
+		__entry->ip_summed, __entry->gso_size, __entry->gso_segs,
+		__entry->gso_type, __entry->fraglist, __entry->nr_frags,
+		__print_array(__entry->size, MAX_SKB_FRAGS, sizeof(__u32))
+	)
+);
+
+DEFINE_EVENT(hns3_skb_template, hns3_over_8bd,
+	TP_PROTO(struct sk_buff *skb),
+	TP_ARGS(skb));
+
+DEFINE_EVENT(hns3_skb_template, hns3_gro,
+	TP_PROTO(struct sk_buff *skb),
+	TP_ARGS(skb));
+
+DEFINE_EVENT(hns3_skb_template, hns3_tso,
+	TP_PROTO(struct sk_buff *skb),
+	TP_ARGS(skb));
+
+TRACE_EVENT(hns3_tx_desc,
+	TP_PROTO(struct hns3_enet_ring *ring, int cur_ntu),
+	TP_ARGS(ring, cur_ntu),
+
+	TP_STRUCT__entry(
+		__field(int, index)
+		__field(int, ntu)
+		__field(int, ntc)
+		__field(dma_addr_t, desc_dma)
+		__array(u32, desc, DESC_NR)
+		__string(devname, ring->tqp->handle->kinfo.netdev->name)
+	),
+
+	TP_fast_assign(
+		__entry->index = ring->tqp->tqp_index;
+		__entry->ntu = ring->next_to_use;
+		__entry->ntc = ring->next_to_clean;
+		__entry->desc_dma = ring->desc_dma_addr,
+		memcpy(__entry->desc, &ring->desc[cur_ntu],
+		       sizeof(struct hns3_desc));
+		__assign_str(devname, ring->tqp->handle->kinfo.netdev->name);
+	),
+
+	TP_printk(
+		"%s-%d-%d/%d desc(%pad): %s",
+		__get_str(devname), __entry->index, __entry->ntu,
+		__entry->ntc, &__entry->desc_dma,
+		__print_array(__entry->desc, DESC_NR, sizeof(u32))
+	)
+);
+
+TRACE_EVENT(hns3_rx_desc,
+	TP_PROTO(struct hns3_enet_ring *ring),
+	TP_ARGS(ring),
+
+	TP_STRUCT__entry(
+		__field(int, index)
+		__field(int, ntu)
+		__field(int, ntc)
+		__field(dma_addr_t, desc_dma)
+		__field(dma_addr_t, buf_dma)
+		__array(u32, desc, DESC_NR)
+		__string(devname, ring->tqp->handle->kinfo.netdev->name)
+	),
+
+	TP_fast_assign(
+		__entry->index = ring->tqp->tqp_index;
+		__entry->ntu = ring->next_to_use;
+		__entry->ntc = ring->next_to_clean;
+		__entry->desc_dma = ring->desc_dma_addr;
+		__entry->buf_dma = ring->desc_cb[ring->next_to_clean].dma;
+		memcpy(__entry->desc, &ring->desc[ring->next_to_clean],
+		       sizeof(struct hns3_desc));
+		__assign_str(devname, ring->tqp->handle->kinfo.netdev->name);
+	),
+
+	TP_printk(
+		"%s-%d-%d/%d desc(%pad) buf(%pad): %s",
+		__get_str(devname), __entry->index, __entry->ntu,
+		__entry->ntc, &__entry->desc_dma, &__entry->buf_dma,
+		__print_array(__entry->desc, DESC_NR, sizeof(u32))
+	)
+);
+
+#endif /* _HNS3_TRACE_H_ */
+
+/* This must be outside ifdef _HNS3_TRACE_H */
+#undef TRACE_INCLUDE_PATH
+#define TRACE_INCLUDE_PATH .
+#undef TRACE_INCLUDE_FILE
+#define TRACE_INCLUDE_FILE hns3_trace
+#include <trace/define_trace.h>
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c
index 940ead3970d1..7f509eff562e 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c
@@ -479,19 +479,6 @@ static void hclge_cmd_uninit_regs(struct hclge_hw *hw)
 	hclge_write_dev(hw, HCLGE_NIC_CRQ_TAIL_REG, 0);
 }
 
-static void hclge_destroy_queue(struct hclge_cmq_ring *ring)
-{
-	spin_lock(&ring->lock);
-	hclge_free_cmd_desc(ring);
-	spin_unlock(&ring->lock);
-}
-
-static void hclge_destroy_cmd_queue(struct hclge_hw *hw)
-{
-	hclge_destroy_queue(&hw->cmq.csq);
-	hclge_destroy_queue(&hw->cmq.crq);
-}
-
 void hclge_cmd_uninit(struct hclge_dev *hdev)
 {
 	spin_lock_bh(&hdev->hw.cmq.csq.lock);
@@ -501,5 +488,6 @@ void hclge_cmd_uninit(struct hclge_dev *hdev)
 	spin_unlock(&hdev->hw.cmq.crq.lock);
 	spin_unlock_bh(&hdev->hw.cmq.csq.lock);
 
-	hclge_destroy_cmd_queue(&hdev->hw);
+	hclge_free_cmd_desc(&hdev->hw.cmq.csq);
+	hclge_free_cmd_desc(&hdev->hw.cmq.crq);
 }
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
index d97da67f07a1..96498d9b4754 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
@@ -6,6 +6,7 @@
 #include <linux/types.h>
 #include <linux/io.h>
 #include <linux/etherdevice.h>
+#include "hnae3.h"
 
 #define HCLGE_CMDQ_TX_TIMEOUT		30000
 #define HCLGE_DESC_DATA_LEN		6
@@ -63,6 +64,7 @@ enum hclge_cmd_status {
 struct hclge_misc_vector {
 	u8 __iomem *addr;
 	int vector_irq;
+	char name[HNAE3_INT_NAME_LEN];
 };
 
 struct hclge_cmq {
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
index 112df34b3869..67fad80035d3 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
@@ -73,8 +73,6 @@ static struct hclge_dbg_reg_type_info hclge_dbg_reg_info[] = {
 
 static int hclge_dbg_get_dfx_bd_num(struct hclge_dev *hdev, int offset)
 {
-#define HCLGE_GET_DFX_REG_TYPE_CNT	4
-
 	struct hclge_desc desc[HCLGE_GET_DFX_REG_TYPE_CNT];
 	int entries_per_desc;
 	int index;
@@ -886,8 +884,8 @@ static void hclge_dbg_dump_mng_table(struct hclge_dev *hdev)
 	}
 }
 
-static void hclge_dbg_fd_tcam_read(struct hclge_dev *hdev, u8 stage,
-				   bool sel_x, u32 loc)
+static int hclge_dbg_fd_tcam_read(struct hclge_dev *hdev, u8 stage,
+				  bool sel_x, u32 loc)
 {
 	struct hclge_fd_tcam_config_1_cmd *req1;
 	struct hclge_fd_tcam_config_2_cmd *req2;
@@ -912,7 +910,7 @@ static void hclge_dbg_fd_tcam_read(struct hclge_dev *hdev, u8 stage,
 
 	ret = hclge_cmd_send(&hdev->hw, desc, 3);
 	if (ret)
-		return;
+		return ret;
 
 	dev_info(&hdev->pdev->dev, " read result tcam key %s(%u):\n",
 		 sel_x ? "x" : "y", loc);
@@ -931,16 +929,76 @@ static void hclge_dbg_fd_tcam_read(struct hclge_dev *hdev, u8 stage,
 	req = (u32 *)req3->tcam_data;
 	for (i = 0; i < 5; i++)
 		dev_info(&hdev->pdev->dev, "%08x\n", *req++);
+
+	return ret;
+}
+
+static int hclge_dbg_get_rules_location(struct hclge_dev *hdev, u16 *rule_locs)
+{
+	struct hclge_fd_rule *rule;
+	struct hlist_node *node;
+	int cnt = 0;
+
+	spin_lock_bh(&hdev->fd_rule_lock);
+	hlist_for_each_entry_safe(rule, node, &hdev->fd_rule_list, rule_node) {
+		rule_locs[cnt] = rule->location;
+		cnt++;
+	}
+	spin_unlock_bh(&hdev->fd_rule_lock);
+
+	if (cnt != hdev->hclge_fd_rule_num)
+		return -EINVAL;
+
+	return cnt;
 }
 
 static void hclge_dbg_fd_tcam(struct hclge_dev *hdev)
 {
-	u32 i;
+	int i, ret, rule_cnt;
+	u16 *rule_locs;
+
+	if (!hnae3_dev_fd_supported(hdev)) {
+		dev_err(&hdev->pdev->dev,
+			"Only FD-supported dev supports dump fd tcam\n");
+		return;
+	}
+
+	if (!hdev->hclge_fd_rule_num ||
+	    !hdev->fd_cfg.rule_num[HCLGE_FD_STAGE_1])
+		return;
 
-	for (i = 0; i < hdev->fd_cfg.rule_num[0]; i++) {
-		hclge_dbg_fd_tcam_read(hdev, 0, true, i);
-		hclge_dbg_fd_tcam_read(hdev, 0, false, i);
+	rule_locs = kcalloc(hdev->fd_cfg.rule_num[HCLGE_FD_STAGE_1],
+			    sizeof(u16), GFP_KERNEL);
+	if (!rule_locs)
+		return;
+
+	rule_cnt = hclge_dbg_get_rules_location(hdev, rule_locs);
+	if (rule_cnt <= 0) {
+		dev_err(&hdev->pdev->dev,
+			"failed to get rule number, ret = %d\n", rule_cnt);
+		kfree(rule_locs);
+		return;
 	}
+
+	for (i = 0; i < rule_cnt; i++) {
+		ret = hclge_dbg_fd_tcam_read(hdev, 0, true, rule_locs[i]);
+		if (ret) {
+			dev_err(&hdev->pdev->dev,
+				"failed to get fd tcam key x, ret = %d\n", ret);
+			kfree(rule_locs);
+			return;
+		}
+
+		ret = hclge_dbg_fd_tcam_read(hdev, 0, false, rule_locs[i]);
+		if (ret) {
+			dev_err(&hdev->pdev->dev,
+				"failed to get fd tcam key y, ret = %d\n", ret);
+			kfree(rule_locs);
+			return;
+		}
+	}
+
+	kfree(rule_locs);
 }
 
 void hclge_dbg_dump_rst_info(struct hclge_dev *hdev)
@@ -976,6 +1034,14 @@ void hclge_dbg_dump_rst_info(struct hclge_dev *hdev)
 	dev_info(&hdev->pdev->dev, "hdev state: 0x%lx\n", hdev->state);
 }
 
+static void hclge_dbg_dump_serv_info(struct hclge_dev *hdev)
+{
+	dev_info(&hdev->pdev->dev, "last_serv_processed: %lu\n",
+		 hdev->last_serv_processed);
+	dev_info(&hdev->pdev->dev, "last_serv_cnt: %lu\n",
+		 hdev->serv_processed_cnt);
+}
+
 static void hclge_dbg_get_m7_stats_info(struct hclge_dev *hdev)
 {
 	struct hclge_desc *desc_src, *desc_tmp;
@@ -1227,6 +1293,8 @@ int hclge_dbg_run_cmd(struct hnae3_handle *handle, const char *cmd_buf)
 		hclge_dbg_dump_reg_cmd(hdev, &cmd_buf[sizeof(DUMP_REG)]);
 	} else if (strncmp(cmd_buf, "dump reset info", 15) == 0) {
 		hclge_dbg_dump_rst_info(hdev);
+	} else if (strncmp(cmd_buf, "dump serv info", 14) == 0) {
+		hclge_dbg_dump_serv_info(hdev);
 	} else if (strncmp(cmd_buf, "dump m7 info", 12) == 0) {
 		hclge_dbg_get_m7_stats_info(hdev);
 	} else if (strncmp(cmd_buf, "dump ncl_config", 15) == 0) {
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c
index dc66b4e13377..c85b72dc44d2 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c
@@ -505,7 +505,7 @@ static const struct hclge_hw_error hclge_ssu_mem_ecc_err_int[] = {
 
 static const struct hclge_hw_error hclge_ssu_port_based_err_int[] = {
 	{ .int_msk = BIT(0), .msg = "roc_pkt_without_key_port",
-	  .reset_level = HNAE3_GLOBAL_RESET },
+	  .reset_level = HNAE3_FUNC_RESET },
 	{ .int_msk = BIT(1), .msg = "tpu_pkt_without_key_port",
 	  .reset_level = HNAE3_GLOBAL_RESET },
 	{ .int_msk = BIT(2), .msg = "igu_pkt_without_key_port",
@@ -599,7 +599,7 @@ static const struct hclge_hw_error hclge_ssu_ets_tcg_int[] = {
 
 static const struct hclge_hw_error hclge_ssu_port_based_pf_int[] = {
 	{ .int_msk = BIT(0), .msg = "roc_pkt_without_key_port",
-	  .reset_level = HNAE3_GLOBAL_RESET },
+	  .reset_level = HNAE3_FUNC_RESET },
 	{ .int_msk = BIT(9), .msg = "low_water_line_err_port",
 	  .reset_level = HNAE3_NONE_RESET },
 	{ .int_msk = BIT(10), .msg = "hi_water_line_err_port",
@@ -1898,10 +1898,8 @@ static int hclge_handle_all_hw_msix_error(struct hclge_dev *hdev,
 
 	bd_num = max_t(u32, mpf_bd_num, pf_bd_num);
 	desc = kcalloc(bd_num, sizeof(struct hclge_desc), GFP_KERNEL);
-	if (!desc) {
-		ret = -ENOMEM;
-		goto out;
-	}
+	if (!desc)
+		return -ENOMEM;
 
 	ret = hclge_handle_mpf_msix_error(hdev, desc, mpf_bd_num,
 					  reset_requests);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 13dbd249f35f..ec5f6eeb639b 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -72,6 +72,8 @@ static int hclge_set_default_loopback(struct hclge_dev *hdev);
 
 static struct hnae3_ae_algo ae_algo;
 
+static struct workqueue_struct *hclge_wq;
+
 static const struct pci_device_id ae_algo_pci_tbl[] = {
 	{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_GE), 0},
 	{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE), 0},
@@ -416,7 +418,7 @@ static int hclge_mac_update_stats_defective(struct hclge_dev *hdev)
 {
 #define HCLGE_MAC_CMD_NUM 21
 
-	u64 *data = (u64 *)(&hdev->hw_stats.mac_stats);
+	u64 *data = (u64 *)(&hdev->mac_stats);
 	struct hclge_desc desc[HCLGE_MAC_CMD_NUM];
 	__le64 *desc_data;
 	int i, k, n;
@@ -453,7 +455,7 @@ static int hclge_mac_update_stats_defective(struct hclge_dev *hdev)
 
 static int hclge_mac_update_stats_complete(struct hclge_dev *hdev, u32 desc_num)
 {
-	u64 *data = (u64 *)(&hdev->hw_stats.mac_stats);
+	u64 *data = (u64 *)(&hdev->mac_stats);
 	struct hclge_desc *desc;
 	__le64 *desc_data;
 	u16 i, k, n;
@@ -802,7 +804,7 @@ static void hclge_get_stats(struct hnae3_handle *handle, u64 *data)
 	struct hclge_dev *hdev = vport->back;
 	u64 *p;
 
-	p = hclge_comm_get_stats(&hdev->hw_stats.mac_stats, g_mac_stats_string,
+	p = hclge_comm_get_stats(&hdev->mac_stats, g_mac_stats_string,
 				 ARRAY_SIZE(g_mac_stats_string), data);
 	p = hclge_tqps_get_stats(handle, p);
 }
@@ -815,8 +817,8 @@ static void hclge_get_mac_stat(struct hnae3_handle *handle,
 
 	hclge_update_stats(handle, NULL);
 
-	mac_stats->tx_pause_cnt = hdev->hw_stats.mac_stats.mac_tx_mac_pause_num;
-	mac_stats->rx_pause_cnt = hdev->hw_stats.mac_stats.mac_rx_mac_pause_num;
+	mac_stats->tx_pause_cnt = hdev->mac_stats.mac_tx_mac_pause_num;
+	mac_stats->rx_pause_cnt = hdev->mac_stats.mac_rx_mac_pause_num;
 }
 
 static int hclge_parse_func_status(struct hclge_dev *hdev,
@@ -860,9 +862,7 @@ static int hclge_query_function_status(struct hclge_dev *hdev)
 		usleep_range(1000, 2000);
 	} while (timeout++ < HCLGE_QUERY_MAX_CNT);
 
-	ret = hclge_parse_func_status(hdev, req);
-
-	return ret;
+	return hclge_parse_func_status(hdev, req);
 }
 
 static int hclge_query_pf_resource(struct hclge_dev *hdev)
@@ -880,12 +880,12 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev)
 	}
 
 	req = (struct hclge_pf_res_cmd *)desc.data;
-	hdev->num_tqps = __le16_to_cpu(req->tqp_num);
-	hdev->pkt_buf_size = __le16_to_cpu(req->buf_size) << HCLGE_BUF_UNIT_S;
+	hdev->num_tqps = le16_to_cpu(req->tqp_num);
+	hdev->pkt_buf_size = le16_to_cpu(req->buf_size) << HCLGE_BUF_UNIT_S;
 
 	if (req->tx_buf_size)
 		hdev->tx_buf_size =
-			__le16_to_cpu(req->tx_buf_size) << HCLGE_BUF_UNIT_S;
+			le16_to_cpu(req->tx_buf_size) << HCLGE_BUF_UNIT_S;
 	else
 		hdev->tx_buf_size = HCLGE_DEFAULT_TX_BUF;
 
@@ -893,7 +893,7 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev)
 
 	if (req->dv_buf_size)
 		hdev->dv_buf_size =
-			__le16_to_cpu(req->dv_buf_size) << HCLGE_BUF_UNIT_S;
+			le16_to_cpu(req->dv_buf_size) << HCLGE_BUF_UNIT_S;
 	else
 		hdev->dv_buf_size = HCLGE_DEFAULT_DV;
 
@@ -901,10 +901,10 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev)
 
 	if (hnae3_dev_roce_supported(hdev)) {
 		hdev->roce_base_msix_offset =
-		hnae3_get_field(__le16_to_cpu(req->msixcap_localid_ba_rocee),
+		hnae3_get_field(le16_to_cpu(req->msixcap_localid_ba_rocee),
 				HCLGE_MSIX_OFT_ROCEE_M, HCLGE_MSIX_OFT_ROCEE_S);
 		hdev->num_roce_msi =
-		hnae3_get_field(__le16_to_cpu(req->pf_intr_vector_number),
+		hnae3_get_field(le16_to_cpu(req->pf_intr_vector_number),
 				HCLGE_PF_VEC_NUM_M, HCLGE_PF_VEC_NUM_S);
 
 		/* nic's msix numbers is always equals to the roce's. */
@@ -917,7 +917,7 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev)
 				hdev->roce_base_msix_offset;
 	} else {
 		hdev->num_msi =
-		hnae3_get_field(__le16_to_cpu(req->pf_intr_vector_number),
+		hnae3_get_field(le16_to_cpu(req->pf_intr_vector_number),
 				HCLGE_PF_VEC_NUM_M, HCLGE_PF_VEC_NUM_S);
 
 		hdev->num_nic_msi = hdev->num_msi;
@@ -1331,11 +1331,7 @@ static int hclge_get_cap(struct hclge_dev *hdev)
 	}
 
 	/* get pf resource */
-	ret = hclge_query_pf_resource(hdev);
-	if (ret)
-		dev_err(&hdev->pdev->dev, "query pf resource error %d.\n", ret);
-
-	return ret;
+	return hclge_query_pf_resource(hdev);
 }
 
 static void hclge_init_kdump_kernel_config(struct hclge_dev *hdev)
@@ -2619,30 +2615,21 @@ static int hclge_mac_init(struct hclge_dev *hdev)
 	hdev->hw.mac.duplex = HCLGE_MAC_FULL;
 	ret = hclge_cfg_mac_speed_dup_hw(hdev, hdev->hw.mac.speed,
 					 hdev->hw.mac.duplex);
-	if (ret) {
-		dev_err(&hdev->pdev->dev,
-			"Config mac speed dup fail ret=%d\n", ret);
+	if (ret)
 		return ret;
-	}
 
 	if (hdev->hw.mac.support_autoneg) {
 		ret = hclge_set_autoneg_en(hdev, hdev->hw.mac.autoneg);
-		if (ret) {
-			dev_err(&hdev->pdev->dev,
-				"Config mac autoneg fail ret=%d\n", ret);
+		if (ret)
 			return ret;
-		}
 	}
 
 	mac->link = 0;
 
 	if (mac->user_fec_mode & BIT(HNAE3_FEC_USER_DEF)) {
 		ret = hclge_set_fec_hw(hdev, mac->user_fec_mode);
-		if (ret) {
-			dev_err(&hdev->pdev->dev,
-				"Fec mode init fail, ret = %d\n", ret);
+		if (ret)
 			return ret;
-		}
 	}
 
 	ret = hclge_set_mac_mtu(hdev, hdev->mps);
@@ -2665,31 +2652,27 @@ static int hclge_mac_init(struct hclge_dev *hdev)
 
 static void hclge_mbx_task_schedule(struct hclge_dev *hdev)
 {
-	if (!test_bit(HCLGE_STATE_CMD_DISABLE, &hdev->state) &&
+	if (!test_bit(HCLGE_STATE_REMOVING, &hdev->state) &&
 	    !test_and_set_bit(HCLGE_STATE_MBX_SERVICE_SCHED, &hdev->state))
-		queue_work_on(cpumask_first(&hdev->affinity_mask), system_wq,
-			      &hdev->mbx_service_task);
+		mod_delayed_work_on(cpumask_first(&hdev->affinity_mask),
+				    hclge_wq, &hdev->service_task, 0);
 }
 
 static void hclge_reset_task_schedule(struct hclge_dev *hdev)
 {
 	if (!test_bit(HCLGE_STATE_REMOVING, &hdev->state) &&
 	    !test_and_set_bit(HCLGE_STATE_RST_SERVICE_SCHED, &hdev->state))
-		queue_work_on(cpumask_first(&hdev->affinity_mask), system_wq,
-			      &hdev->rst_service_task);
+		mod_delayed_work_on(cpumask_first(&hdev->affinity_mask),
+				    hclge_wq, &hdev->service_task, 0);
 }
 
 void hclge_task_schedule(struct hclge_dev *hdev, unsigned long delay_time)
 {
-	if (!test_bit(HCLGE_STATE_DOWN, &hdev->state) &&
-	    !test_bit(HCLGE_STATE_REMOVING, &hdev->state) &&
-	    !test_and_set_bit(HCLGE_STATE_SERVICE_SCHED, &hdev->state)) {
-		hdev->hw_stats.stats_timer++;
-		hdev->fd_arfs_expire_timer++;
+	if (!test_bit(HCLGE_STATE_REMOVING, &hdev->state) &&
+	    !test_bit(HCLGE_STATE_RST_FAIL, &hdev->state))
 		mod_delayed_work_on(cpumask_first(&hdev->affinity_mask),
-				    system_wq, &hdev->service_task,
+				    hclge_wq, &hdev->service_task,
 				    delay_time);
-	}
 }
 
 static int hclge_get_mac_link_status(struct hclge_dev *hdev)
@@ -2748,6 +2731,10 @@ static void hclge_update_link_status(struct hclge_dev *hdev)
 
 	if (!client)
 		return;
+
+	if (test_and_set_bit(HCLGE_STATE_LINK_UPDATING, &hdev->state))
+		return;
+
 	state = hclge_get_mac_phy_link(hdev);
 	if (state != hdev->hw.mac.link) {
 		for (i = 0; i < hdev->num_vmdq_vport + 1; i++) {
@@ -2761,6 +2748,8 @@ static void hclge_update_link_status(struct hclge_dev *hdev)
 		}
 		hdev->hw.mac.link = state;
 	}
+
+	clear_bit(HCLGE_STATE_LINK_UPDATING, &hdev->state);
 }
 
 static void hclge_update_port_capability(struct hclge_mac *mac)
@@ -2831,6 +2820,12 @@ static int hclge_get_sfp_info(struct hclge_dev *hdev, struct hclge_mac *mac)
 		return ret;
 	}
 
+	/* In some case, mac speed get from IMP may be 0, it shouldn't be
+	 * set to mac->speed.
+	 */
+	if (!le32_to_cpu(resp->speed))
+		return 0;
+
 	mac->speed = le32_to_cpu(resp->speed);
 	/* if resp->speed_ability is 0, it means it's an old version
 	 * firmware, do not update these params
@@ -2906,7 +2901,7 @@ static int hclge_get_status(struct hnae3_handle *handle)
 
 static struct hclge_vport *hclge_get_vf_vport(struct hclge_dev *hdev, int vf)
 {
-	if (pci_num_vf(hdev->pdev) == 0) {
+	if (!pci_num_vf(hdev->pdev)) {
 		dev_err(&hdev->pdev->dev,
 			"SRIOV is disabled, can not get vport(%d) info.\n", vf);
 		return NULL;
@@ -2940,6 +2935,9 @@ static int hclge_get_vf_config(struct hnae3_handle *handle, int vf,
 	ivf->trusted = vport->vf_info.trusted;
 	ivf->min_tx_rate = 0;
 	ivf->max_tx_rate = vport->vf_info.max_tx_rate;
+	ivf->vlan = vport->port_base_vlan_cfg.vlan_info.vlan_tag;
+	ivf->vlan_proto = htons(vport->port_base_vlan_cfg.vlan_info.vlan_proto);
+	ivf->qos = vport->port_base_vlan_cfg.vlan_info.qos;
 	ether_addr_copy(ivf->mac, vport->vf_info.mac);
 
 	return 0;
@@ -2998,8 +2996,6 @@ static u32 hclge_check_event_cause(struct hclge_dev *hdev, u32 *clearval)
 
 	/* check for vector0 msix event source */
 	if (msix_src_reg & HCLGE_VECTOR0_REG_MSIX_MASK) {
-		dev_info(&hdev->pdev->dev, "received event 0x%x\n",
-			 msix_src_reg);
 		*clearval = msix_src_reg;
 		return HCLGE_VECTOR0_EVENT_ERR;
 	}
@@ -3172,8 +3168,10 @@ static int hclge_misc_irq_init(struct hclge_dev *hdev)
 	hclge_get_misc_vector(hdev);
 
 	/* this would be explicitly freed in the end */
+	snprintf(hdev->misc_vector.name, HNAE3_INT_NAME_LEN, "%s-misc-%s",
+		 HCLGE_NAME, pci_name(hdev->pdev));
 	ret = request_irq(hdev->misc_vector.vector_irq, hclge_misc_irq_handle,
-			  0, "hclge_misc", hdev);
+			  0, hdev->misc_vector.name, hdev);
 	if (ret) {
 		hclge_free_vector(hdev, 0);
 		dev_err(&hdev->pdev->dev, "request misc irq(%d) fail\n",
@@ -3247,7 +3245,8 @@ static int hclge_notify_roce_client(struct hclge_dev *hdev,
 static int hclge_reset_wait(struct hclge_dev *hdev)
 {
 #define HCLGE_RESET_WATI_MS	100
-#define HCLGE_RESET_WAIT_CNT	200
+#define HCLGE_RESET_WAIT_CNT	350
+
 	u32 val, reg, reg_bit;
 	u32 cnt = 0;
 
@@ -3264,8 +3263,6 @@ static int hclge_reset_wait(struct hclge_dev *hdev)
 		reg = HCLGE_FUN_RST_ING;
 		reg_bit = HCLGE_FUN_RST_ING_B;
 		break;
-	case HNAE3_FLR_RESET:
-		break;
 	default:
 		dev_err(&hdev->pdev->dev,
 			"Wait for unsupported reset type: %d\n",
@@ -3273,20 +3270,6 @@ static int hclge_reset_wait(struct hclge_dev *hdev)
 		return -EINVAL;
 	}
 
-	if (hdev->reset_type == HNAE3_FLR_RESET) {
-		while (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state) &&
-		       cnt++ < HCLGE_RESET_WAIT_CNT)
-			msleep(HCLGE_RESET_WATI_MS);
-
-		if (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state)) {
-			dev_err(&hdev->pdev->dev,
-				"flr wait timeout: %u\n", cnt);
-			return -EBUSY;
-		}
-
-		return 0;
-	}
-
 	val = hclge_read_dev(&hdev->hw, reg);
 	while (hnae3_get_bit(val, reg_bit) && cnt < HCLGE_RESET_WAIT_CNT) {
 		msleep(HCLGE_RESET_WATI_MS);
@@ -3352,7 +3335,19 @@ static int hclge_set_all_vf_rst(struct hclge_dev *hdev, bool reset)
 	return 0;
 }
 
-static int hclge_func_reset_sync_vf(struct hclge_dev *hdev)
+static void hclge_mailbox_service_task(struct hclge_dev *hdev)
+{
+	if (!test_and_clear_bit(HCLGE_STATE_MBX_SERVICE_SCHED, &hdev->state) ||
+	    test_bit(HCLGE_STATE_CMD_DISABLE, &hdev->state) ||
+	    test_and_set_bit(HCLGE_STATE_MBX_HANDLING, &hdev->state))
+		return;
+
+	hclge_mbx_handler(hdev);
+
+	clear_bit(HCLGE_STATE_MBX_HANDLING, &hdev->state);
+}
+
+static void hclge_func_reset_sync_vf(struct hclge_dev *hdev)
 {
 	struct hclge_pf_rst_sync_cmd *req;
 	struct hclge_desc desc;
@@ -3363,26 +3358,28 @@ static int hclge_func_reset_sync_vf(struct hclge_dev *hdev)
 	hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_QUERY_VF_RST_RDY, true);
 
 	do {
+		/* vf need to down netdev by mbx during PF or FLR reset */
+		hclge_mailbox_service_task(hdev);
+
 		ret = hclge_cmd_send(&hdev->hw, &desc, 1);
 		/* for compatible with old firmware, wait
 		 * 100 ms for VF to stop IO
 		 */
 		if (ret == -EOPNOTSUPP) {
 			msleep(HCLGE_RESET_SYNC_TIME);
-			return 0;
+			return;
 		} else if (ret) {
-			dev_err(&hdev->pdev->dev, "sync with VF fail %d!\n",
-				ret);
-			return ret;
+			dev_warn(&hdev->pdev->dev, "sync with VF fail %d!\n",
+				 ret);
+			return;
 		} else if (req->all_vf_ready) {
-			return 0;
+			return;
 		}
 		msleep(HCLGE_PF_RESET_SYNC_TIME);
 		hclge_cmd_reuse_desc(&desc, true);
 	} while (cnt++ < HCLGE_PF_RESET_SYNC_CNT);
 
-	dev_err(&hdev->pdev->dev, "sync with VF timeout!\n");
-	return -ETIME;
+	dev_warn(&hdev->pdev->dev, "sync with VF timeout!\n");
 }
 
 void hclge_report_hw_error(struct hclge_dev *hdev,
@@ -3462,12 +3459,6 @@ static void hclge_do_reset(struct hclge_dev *hdev)
 		set_bit(HNAE3_FUNC_RESET, &hdev->reset_pending);
 		hclge_reset_task_schedule(hdev);
 		break;
-	case HNAE3_FLR_RESET:
-		dev_info(&pdev->dev, "FLR requested\n");
-		/* schedule again to check later */
-		set_bit(HNAE3_FLR_RESET, &hdev->reset_pending);
-		hclge_reset_task_schedule(hdev);
-		break;
 	default:
 		dev_warn(&pdev->dev,
 			 "Unsupported reset type: %d\n", hdev->reset_type);
@@ -3483,10 +3474,15 @@ static enum hnae3_reset_type hclge_get_reset_level(struct hnae3_ae_dev *ae_dev,
 
 	/* first, resolve any unknown reset type to the known type(s) */
 	if (test_bit(HNAE3_UNKNOWN_RESET, addr)) {
+		u32 msix_sts_reg = hclge_read_dev(&hdev->hw,
+					HCLGE_VECTOR0_PF_OTHER_INT_STS_REG);
 		/* we will intentionally ignore any errors from this function
 		 *  as we will end up in *some* reset request in any case
 		 */
-		hclge_handle_hw_msix_error(hdev, addr);
+		if (hclge_handle_hw_msix_error(hdev, addr))
+			dev_info(&hdev->pdev->dev, "received msix interrupt 0x%x\n",
+				 msix_sts_reg);
+
 		clear_bit(HNAE3_UNKNOWN_RESET, addr);
 		/* We defered the clearing of the error event which caused
 		 * interrupt since it was not posssible to do that in
@@ -3551,23 +3547,6 @@ static void hclge_clear_reset_cause(struct hclge_dev *hdev)
 	hclge_enable_vector(&hdev->misc_vector, true);
 }
 
-static int hclge_reset_prepare_down(struct hclge_dev *hdev)
-{
-	int ret = 0;
-
-	switch (hdev->reset_type) {
-	case HNAE3_FUNC_RESET:
-		/* fall through */
-	case HNAE3_FLR_RESET:
-		ret = hclge_set_all_vf_rst(hdev, true);
-		break;
-	default:
-		break;
-	}
-
-	return ret;
-}
-
 static void hclge_reset_handshake(struct hclge_dev *hdev, bool enable)
 {
 	u32 reg_val;
@@ -3581,6 +3560,19 @@ static void hclge_reset_handshake(struct hclge_dev *hdev, bool enable)
 	hclge_write_dev(&hdev->hw, HCLGE_NIC_CSQ_DEPTH_REG, reg_val);
 }
 
+static int hclge_func_reset_notify_vf(struct hclge_dev *hdev)
+{
+	int ret;
+
+	ret = hclge_set_all_vf_rst(hdev, true);
+	if (ret)
+		return ret;
+
+	hclge_func_reset_sync_vf(hdev);
+
+	return 0;
+}
+
 static int hclge_reset_prepare_wait(struct hclge_dev *hdev)
 {
 	u32 reg_val;
@@ -3588,10 +3580,7 @@ static int hclge_reset_prepare_wait(struct hclge_dev *hdev)
 
 	switch (hdev->reset_type) {
 	case HNAE3_FUNC_RESET:
-		/* to confirm whether all running VF is ready
-		 * before request PF reset
-		 */
-		ret = hclge_func_reset_sync_vf(hdev);
+		ret = hclge_func_reset_notify_vf(hdev);
 		if (ret)
 			return ret;
 
@@ -3611,16 +3600,9 @@ static int hclge_reset_prepare_wait(struct hclge_dev *hdev)
 		hdev->rst_stats.pf_rst_cnt++;
 		break;
 	case HNAE3_FLR_RESET:
-		/* to confirm whether all running VF is ready
-		 * before request PF reset
-		 */
-		ret = hclge_func_reset_sync_vf(hdev);
+		ret = hclge_func_reset_notify_vf(hdev);
 		if (ret)
 			return ret;
-
-		set_bit(HCLGE_STATE_CMD_DISABLE, &hdev->state);
-		set_bit(HNAE3_FLR_DOWN, &hdev->flr_state);
-		hdev->rst_stats.flr_rst_cnt++;
 		break;
 	case HNAE3_IMP_RESET:
 		hclge_handle_imp_error(hdev);
@@ -3672,6 +3654,8 @@ static bool hclge_reset_err_handle(struct hclge_dev *hdev)
 
 	hclge_dbg_dump_rst_info(hdev);
 
+	set_bit(HCLGE_STATE_RST_FAIL, &hdev->state);
+
 	return false;
 }
 
@@ -3747,10 +3731,9 @@ static int hclge_reset_stack(struct hclge_dev *hdev)
 	return hclge_notify_client(hdev, HNAE3_RESTORE_CLIENT);
 }
 
-static void hclge_reset(struct hclge_dev *hdev)
+static int hclge_reset_prepare(struct hclge_dev *hdev)
 {
 	struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev);
-	enum hnae3_reset_type reset_level;
 	int ret;
 
 	/* Initialize ae_dev reset status as well, in case enet layer wants to
@@ -3761,45 +3744,41 @@ static void hclge_reset(struct hclge_dev *hdev)
 	/* perform reset of the stack & ae device for a client */
 	ret = hclge_notify_roce_client(hdev, HNAE3_DOWN_CLIENT);
 	if (ret)
-		goto err_reset;
-
-	ret = hclge_reset_prepare_down(hdev);
-	if (ret)
-		goto err_reset;
+		return ret;
 
 	rtnl_lock();
 	ret = hclge_notify_client(hdev, HNAE3_DOWN_CLIENT);
-	if (ret)
-		goto err_reset_lock;
-
 	rtnl_unlock();
-
-	ret = hclge_reset_prepare_wait(hdev);
 	if (ret)
-		goto err_reset;
+		return ret;
 
-	if (hclge_reset_wait(hdev))
-		goto err_reset;
+	return hclge_reset_prepare_wait(hdev);
+}
+
+static int hclge_reset_rebuild(struct hclge_dev *hdev)
+{
+	struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev);
+	enum hnae3_reset_type reset_level;
+	int ret;
 
 	hdev->rst_stats.hw_reset_done_cnt++;
 
 	ret = hclge_notify_roce_client(hdev, HNAE3_UNINIT_CLIENT);
 	if (ret)
-		goto err_reset;
+		return ret;
 
 	rtnl_lock();
-
 	ret = hclge_reset_stack(hdev);
+	rtnl_unlock();
 	if (ret)
-		goto err_reset_lock;
+		return ret;
 
 	hclge_clear_reset_cause(hdev);
 
 	ret = hclge_reset_prepare_up(hdev);
 	if (ret)
-		goto err_reset_lock;
+		return ret;
 
-	rtnl_unlock();
 
 	ret = hclge_notify_roce_client(hdev, HNAE3_INIT_CLIENT);
 	/* ignore RoCE notify error if it fails HCLGE_RESET_MAX_FAIL_CNT - 1
@@ -3807,24 +3786,23 @@ static void hclge_reset(struct hclge_dev *hdev)
 	 */
 	if (ret &&
 	    hdev->rst_stats.reset_fail_cnt < HCLGE_RESET_MAX_FAIL_CNT - 1)
-		goto err_reset;
+		return ret;
 
 	rtnl_lock();
-
 	ret = hclge_notify_client(hdev, HNAE3_UP_CLIENT);
-	if (ret)
-		goto err_reset_lock;
-
 	rtnl_unlock();
+	if (ret)
+		return ret;
 
 	ret = hclge_notify_roce_client(hdev, HNAE3_UP_CLIENT);
 	if (ret)
-		goto err_reset;
+		return ret;
 
 	hdev->last_reset_time = jiffies;
 	hdev->rst_stats.reset_fail_cnt = 0;
 	hdev->rst_stats.reset_done_cnt++;
 	ae_dev->reset_type = HNAE3_NONE_RESET;
+	clear_bit(HCLGE_STATE_RST_FAIL, &hdev->state);
 
 	/* if default_reset_request has a higher level reset request,
 	 * it should be handled as soon as possible. since some errors
@@ -3835,10 +3813,22 @@ static void hclge_reset(struct hclge_dev *hdev)
 	if (reset_level != HNAE3_NONE_RESET)
 		set_bit(reset_level, &hdev->reset_request);
 
+	return 0;
+}
+
+static void hclge_reset(struct hclge_dev *hdev)
+{
+	if (hclge_reset_prepare(hdev))
+		goto err_reset;
+
+	if (hclge_reset_wait(hdev))
+		goto err_reset;
+
+	if (hclge_reset_rebuild(hdev))
+		goto err_reset;
+
 	return;
 
-err_reset_lock:
-	rtnl_unlock();
 err_reset:
 	if (hclge_reset_err_handle(hdev))
 		hclge_reset_task_schedule(hdev);
@@ -3939,34 +3929,18 @@ static void hclge_reset_subtask(struct hclge_dev *hdev)
 	hdev->reset_type = HNAE3_NONE_RESET;
 }
 
-static void hclge_reset_service_task(struct work_struct *work)
+static void hclge_reset_service_task(struct hclge_dev *hdev)
 {
-	struct hclge_dev *hdev =
-		container_of(work, struct hclge_dev, rst_service_task);
-
-	if (test_and_set_bit(HCLGE_STATE_RST_HANDLING, &hdev->state))
+	if (!test_and_clear_bit(HCLGE_STATE_RST_SERVICE_SCHED, &hdev->state))
 		return;
 
-	clear_bit(HCLGE_STATE_RST_SERVICE_SCHED, &hdev->state);
+	down(&hdev->reset_sem);
+	set_bit(HCLGE_STATE_RST_HANDLING, &hdev->state);
 
 	hclge_reset_subtask(hdev);
 
 	clear_bit(HCLGE_STATE_RST_HANDLING, &hdev->state);
-}
-
-static void hclge_mailbox_service_task(struct work_struct *work)
-{
-	struct hclge_dev *hdev =
-		container_of(work, struct hclge_dev, mbx_service_task);
-
-	if (test_and_set_bit(HCLGE_STATE_MBX_HANDLING, &hdev->state))
-		return;
-
-	clear_bit(HCLGE_STATE_MBX_SERVICE_SCHED, &hdev->state);
-
-	hclge_mbx_handler(hdev);
-
-	clear_bit(HCLGE_STATE_MBX_HANDLING, &hdev->state);
+	up(&hdev->reset_sem);
 }
 
 static void hclge_update_vport_alive(struct hclge_dev *hdev)
@@ -3986,29 +3960,62 @@ static void hclge_update_vport_alive(struct hclge_dev *hdev)
 	}
 }
 
-static void hclge_service_task(struct work_struct *work)
+static void hclge_periodic_service_task(struct hclge_dev *hdev)
 {
-	struct hclge_dev *hdev =
-		container_of(work, struct hclge_dev, service_task.work);
+	unsigned long delta = round_jiffies_relative(HZ);
 
-	clear_bit(HCLGE_STATE_SERVICE_SCHED, &hdev->state);
+	/* Always handle the link updating to make sure link state is
+	 * updated when it is triggered by mbx.
+	 */
+	hclge_update_link_status(hdev);
 
-	if (hdev->hw_stats.stats_timer >= HCLGE_STATS_TIMER_INTERVAL) {
-		hclge_update_stats_for_all(hdev);
-		hdev->hw_stats.stats_timer = 0;
+	if (time_is_after_jiffies(hdev->last_serv_processed + HZ)) {
+		delta = jiffies - hdev->last_serv_processed;
+
+		if (delta < round_jiffies_relative(HZ)) {
+			delta = round_jiffies_relative(HZ) - delta;
+			goto out;
+		}
 	}
 
-	hclge_update_port_info(hdev);
-	hclge_update_link_status(hdev);
+	hdev->serv_processed_cnt++;
 	hclge_update_vport_alive(hdev);
+
+	if (test_bit(HCLGE_STATE_DOWN, &hdev->state)) {
+		hdev->last_serv_processed = jiffies;
+		goto out;
+	}
+
+	if (!(hdev->serv_processed_cnt % HCLGE_STATS_TIMER_INTERVAL))
+		hclge_update_stats_for_all(hdev);
+
+	hclge_update_port_info(hdev);
 	hclge_sync_vlan_filter(hdev);
 
-	if (hdev->fd_arfs_expire_timer >= HCLGE_FD_ARFS_EXPIRE_TIMER_INTERVAL) {
+	if (!(hdev->serv_processed_cnt % HCLGE_ARFS_EXPIRE_INTERVAL))
 		hclge_rfs_filter_expire(hdev);
-		hdev->fd_arfs_expire_timer = 0;
-	}
 
-	hclge_task_schedule(hdev, round_jiffies_relative(HZ));
+	hdev->last_serv_processed = jiffies;
+
+out:
+	hclge_task_schedule(hdev, delta);
+}
+
+static void hclge_service_task(struct work_struct *work)
+{
+	struct hclge_dev *hdev =
+		container_of(work, struct hclge_dev, service_task.work);
+
+	hclge_reset_service_task(hdev);
+	hclge_mailbox_service_task(hdev);
+	hclge_periodic_service_task(hdev);
+
+	/* Handle reset and mbx again in case periodical task delays the
+	 * handling by calling hclge_task_schedule() in
+	 * hclge_periodic_service_task().
+	 */
+	hclge_reset_service_task(hdev);
+	hclge_mailbox_service_task(hdev);
 }
 
 struct hclge_vport *hclge_get_vport(struct hnae3_handle *handle)
@@ -4079,7 +4086,7 @@ static int hclge_put_vector(struct hnae3_handle *handle, int vector)
 	vector_id = hclge_get_vector_index(hdev, vector);
 	if (vector_id < 0) {
 		dev_err(&hdev->pdev->dev,
-			"Get vector index fail. vector_id =%d\n", vector_id);
+			"Get vector index fail. vector = %d\n", vector);
 		return vector_id;
 	}
 
@@ -4654,7 +4661,7 @@ static int hclge_map_ring_to_vector(struct hnae3_handle *handle, int vector,
 	vector_id = hclge_get_vector_index(hdev, vector);
 	if (vector_id < 0) {
 		dev_err(&hdev->pdev->dev,
-			"Get vector index fail. vector_id =%d\n", vector_id);
+			"failed to get vector index. vector=%d\n", vector);
 		return vector_id;
 	}
 
@@ -6562,7 +6569,7 @@ static int hclge_set_serdes_loopback(struct hclge_dev *hdev, bool en,
 
 	hclge_cfg_mac_mode(hdev, en);
 
-	ret = hclge_mac_phy_link_status_wait(hdev, en, FALSE);
+	ret = hclge_mac_phy_link_status_wait(hdev, en, false);
 	if (ret)
 		dev_err(&hdev->pdev->dev,
 			"serdes loopback config mac mode timeout\n");
@@ -6620,7 +6627,7 @@ static int hclge_set_phy_loopback(struct hclge_dev *hdev, bool en)
 
 	hclge_cfg_mac_mode(hdev, en);
 
-	ret = hclge_mac_phy_link_status_wait(hdev, en, TRUE);
+	ret = hclge_mac_phy_link_status_wait(hdev, en, true);
 	if (ret)
 		dev_err(&hdev->pdev->dev,
 			"phy loopback config mac mode timeout\n");
@@ -6734,6 +6741,19 @@ static void hclge_reset_tqp_stats(struct hnae3_handle *handle)
 	}
 }
 
+static void hclge_flush_link_update(struct hclge_dev *hdev)
+{
+#define HCLGE_FLUSH_LINK_TIMEOUT	100000
+
+	unsigned long last = hdev->serv_processed_cnt;
+	int i = 0;
+
+	while (test_bit(HCLGE_STATE_LINK_UPDATING, &hdev->state) &&
+	       i++ < HCLGE_FLUSH_LINK_TIMEOUT &&
+	       last == hdev->serv_processed_cnt)
+		usleep_range(1, 1);
+}
+
 static void hclge_set_timer_task(struct hnae3_handle *handle, bool enable)
 {
 	struct hclge_vport *vport = hclge_get_vport(handle);
@@ -6742,12 +6762,12 @@ static void hclge_set_timer_task(struct hnae3_handle *handle, bool enable)
 	if (enable) {
 		hclge_task_schedule(hdev, round_jiffies_relative(HZ));
 	} else {
-		/* Set the DOWN flag here to disable the service to be
-		 * scheduled again
-		 */
+		/* Set the DOWN flag here to disable link updating */
 		set_bit(HCLGE_STATE_DOWN, &hdev->state);
-		cancel_delayed_work_sync(&hdev->service_task);
-		clear_bit(HCLGE_STATE_SERVICE_SCHED, &hdev->state);
+
+		/* flush memory to make sure DOWN is seen by service task */
+		smp_mb__before_atomic();
+		hclge_flush_link_update(hdev);
 	}
 }
 
@@ -7483,7 +7503,6 @@ void hclge_uninit_vport_mac_table(struct hclge_dev *hdev)
 	struct hclge_vport *vport;
 	int i;
 
-	mutex_lock(&hdev->vport_cfg_mutex);
 	for (i = 0; i < hdev->num_alloc_vport; i++) {
 		vport = &hdev->vport[i];
 		list_for_each_entry_safe(mac, tmp, &vport->uc_mac_list, node) {
@@ -7496,7 +7515,6 @@ void hclge_uninit_vport_mac_table(struct hclge_dev *hdev)
 			kfree(mac);
 		}
 	}
-	mutex_unlock(&hdev->vport_cfg_mutex);
 }
 
 static int hclge_get_mac_ethertype_cmd_status(struct hclge_dev *hdev,
@@ -8257,7 +8275,6 @@ void hclge_uninit_vport_vlan_table(struct hclge_dev *hdev)
 	struct hclge_vport *vport;
 	int i;
 
-	mutex_lock(&hdev->vport_cfg_mutex);
 	for (i = 0; i < hdev->num_alloc_vport; i++) {
 		vport = &hdev->vport[i];
 		list_for_each_entry_safe(vlan, tmp, &vport->vlan_list, node) {
@@ -8265,7 +8282,6 @@ void hclge_uninit_vport_vlan_table(struct hclge_dev *hdev)
 			kfree(vlan);
 		}
 	}
-	mutex_unlock(&hdev->vport_cfg_mutex);
 }
 
 static void hclge_restore_vlan_table(struct hnae3_handle *handle)
@@ -8277,7 +8293,6 @@ static void hclge_restore_vlan_table(struct hnae3_handle *handle)
 	u16 state, vlan_id;
 	int i;
 
-	mutex_lock(&hdev->vport_cfg_mutex);
 	for (i = 0; i < hdev->num_alloc_vport; i++) {
 		vport = &hdev->vport[i];
 		vlan_proto = vport->port_base_vlan_cfg.vlan_info.vlan_proto;
@@ -8303,8 +8318,6 @@ static void hclge_restore_vlan_table(struct hnae3_handle *handle)
 				break;
 		}
 	}
-
-	mutex_unlock(&hdev->vport_cfg_mutex);
 }
 
 int hclge_en_hw_strip_rxvtag(struct hnae3_handle *handle, bool enable)
@@ -9256,6 +9269,7 @@ static void hclge_state_init(struct hclge_dev *hdev)
 	set_bit(HCLGE_STATE_DOWN, &hdev->state);
 	clear_bit(HCLGE_STATE_RST_SERVICE_SCHED, &hdev->state);
 	clear_bit(HCLGE_STATE_RST_HANDLING, &hdev->state);
+	clear_bit(HCLGE_STATE_RST_FAIL, &hdev->state);
 	clear_bit(HCLGE_STATE_MBX_SERVICE_SCHED, &hdev->state);
 	clear_bit(HCLGE_STATE_MBX_HANDLING, &hdev->state);
 }
@@ -9269,38 +9283,57 @@ static void hclge_state_uninit(struct hclge_dev *hdev)
 		del_timer_sync(&hdev->reset_timer);
 	if (hdev->service_task.work.func)
 		cancel_delayed_work_sync(&hdev->service_task);
-	if (hdev->rst_service_task.func)
-		cancel_work_sync(&hdev->rst_service_task);
-	if (hdev->mbx_service_task.func)
-		cancel_work_sync(&hdev->mbx_service_task);
 }
 
 static void hclge_flr_prepare(struct hnae3_ae_dev *ae_dev)
 {
-#define HCLGE_FLR_WAIT_MS	100
-#define HCLGE_FLR_WAIT_CNT	50
-	struct hclge_dev *hdev = ae_dev->priv;
-	int cnt = 0;
+#define HCLGE_FLR_RETRY_WAIT_MS	500
+#define HCLGE_FLR_RETRY_CNT	5
 
-	clear_bit(HNAE3_FLR_DOWN, &hdev->flr_state);
-	clear_bit(HNAE3_FLR_DONE, &hdev->flr_state);
-	set_bit(HNAE3_FLR_RESET, &hdev->default_reset_request);
-	hclge_reset_event(hdev->pdev, NULL);
+	struct hclge_dev *hdev = ae_dev->priv;
+	int retry_cnt = 0;
+	int ret;
 
-	while (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state) &&
-	       cnt++ < HCLGE_FLR_WAIT_CNT)
-		msleep(HCLGE_FLR_WAIT_MS);
+retry:
+	down(&hdev->reset_sem);
+	set_bit(HCLGE_STATE_RST_HANDLING, &hdev->state);
+	hdev->reset_type = HNAE3_FLR_RESET;
+	ret = hclge_reset_prepare(hdev);
+	if (ret) {
+		dev_err(&hdev->pdev->dev, "fail to prepare FLR, ret=%d\n",
+			ret);
+		if (hdev->reset_pending ||
+		    retry_cnt++ < HCLGE_FLR_RETRY_CNT) {
+			dev_err(&hdev->pdev->dev,
+				"reset_pending:0x%lx, retry_cnt:%d\n",
+				hdev->reset_pending, retry_cnt);
+			clear_bit(HCLGE_STATE_RST_HANDLING, &hdev->state);
+			up(&hdev->reset_sem);
+			msleep(HCLGE_FLR_RETRY_WAIT_MS);
+			goto retry;
+		}
+	}
 
-	if (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state))
-		dev_err(&hdev->pdev->dev,
-			"flr wait down timeout: %d\n", cnt);
+	/* disable misc vector before FLR done */
+	hclge_enable_vector(&hdev->misc_vector, false);
+	set_bit(HCLGE_STATE_CMD_DISABLE, &hdev->state);
+	hdev->rst_stats.flr_rst_cnt++;
 }
 
 static void hclge_flr_done(struct hnae3_ae_dev *ae_dev)
 {
 	struct hclge_dev *hdev = ae_dev->priv;
+	int ret;
+
+	hclge_enable_vector(&hdev->misc_vector, true);
+
+	ret = hclge_reset_rebuild(hdev);
+	if (ret)
+		dev_err(&hdev->pdev->dev, "fail to rebuild, ret=%d\n", ret);
 
-	set_bit(HNAE3_FLR_DONE, &hdev->flr_state);
+	hdev->reset_type = HNAE3_NONE_RESET;
+	clear_bit(HCLGE_STATE_RST_HANDLING, &hdev->state);
+	up(&hdev->reset_sem);
 }
 
 static void hclge_clear_resetting_state(struct hclge_dev *hdev)
@@ -9342,21 +9375,17 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev)
 	hdev->mps = ETH_FRAME_LEN + ETH_FCS_LEN + 2 * VLAN_HLEN;
 
 	mutex_init(&hdev->vport_lock);
-	mutex_init(&hdev->vport_cfg_mutex);
 	spin_lock_init(&hdev->fd_rule_lock);
+	sema_init(&hdev->reset_sem, 1);
 
 	ret = hclge_pci_init(hdev);
-	if (ret) {
-		dev_err(&pdev->dev, "PCI init failed\n");
+	if (ret)
 		goto out;
-	}
 
 	/* Firmware command queue initialize */
 	ret = hclge_cmd_queue_init(hdev);
-	if (ret) {
-		dev_err(&pdev->dev, "Cmd queue init failed, ret = %d.\n", ret);
+	if (ret)
 		goto err_pci_uninit;
-	}
 
 	/* Firmware command initialize */
 	ret = hclge_cmd_init(hdev);
@@ -9364,11 +9393,8 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev)
 		goto err_cmd_uninit;
 
 	ret = hclge_get_cap(hdev);
-	if (ret) {
-		dev_err(&pdev->dev, "get hw capability error, ret = %d.\n",
-			ret);
+	if (ret)
 		goto err_cmd_uninit;
-	}
 
 	ret = hclge_configure(hdev);
 	if (ret) {
@@ -9383,12 +9409,8 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev)
 	}
 
 	ret = hclge_misc_irq_init(hdev);
-	if (ret) {
-		dev_err(&pdev->dev,
-			"Misc IRQ(vector0) init error, ret = %d.\n",
-			ret);
+	if (ret)
 		goto err_msi_uninit;
-	}
 
 	ret = hclge_alloc_tqps(hdev);
 	if (ret) {
@@ -9397,31 +9419,22 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev)
 	}
 
 	ret = hclge_alloc_vport(hdev);
-	if (ret) {
-		dev_err(&pdev->dev, "Allocate vport error, ret = %d.\n", ret);
+	if (ret)
 		goto err_msi_irq_uninit;
-	}
 
 	ret = hclge_map_tqp(hdev);
-	if (ret) {
-		dev_err(&pdev->dev, "Map tqp error, ret = %d.\n", ret);
+	if (ret)
 		goto err_msi_irq_uninit;
-	}
 
 	if (hdev->hw.mac.media_type == HNAE3_MEDIA_TYPE_COPPER) {
 		ret = hclge_mac_mdio_config(hdev);
-		if (ret) {
-			dev_err(&hdev->pdev->dev,
-				"mdio config fail ret=%d\n", ret);
+		if (ret)
 			goto err_msi_irq_uninit;
-		}
 	}
 
 	ret = hclge_init_umv_space(hdev);
-	if (ret) {
-		dev_err(&pdev->dev, "umv space init error, ret=%d.\n", ret);
+	if (ret)
 		goto err_mdiobus_unreg;
-	}
 
 	ret = hclge_mac_init(hdev);
 	if (ret) {
@@ -9477,8 +9490,6 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev)
 
 	timer_setup(&hdev->reset_timer, hclge_reset_timer, 0);
 	INIT_DELAYED_WORK(&hdev->service_task, hclge_service_task);
-	INIT_WORK(&hdev->rst_service_task, hclge_reset_service_task);
-	INIT_WORK(&hdev->mbx_service_task, hclge_mailbox_service_task);
 
 	/* Setup affinity after service timer setup because add_timer_on
 	 * is called in affinity notify.
@@ -9512,6 +9523,8 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev)
 	dev_info(&hdev->pdev->dev, "%s driver initialization finished.\n",
 		 HCLGE_DRIVER_NAME);
 
+	hclge_task_schedule(hdev, round_jiffies_relative(HZ));
+
 	return 0;
 
 err_mdiobus_unreg:
@@ -9534,7 +9547,7 @@ out:
 
 static void hclge_stats_clear(struct hclge_dev *hdev)
 {
-	memset(&hdev->hw_stats, 0, sizeof(hdev->hw_stats));
+	memset(&hdev->mac_stats, 0, sizeof(hdev->mac_stats));
 }
 
 static int hclge_set_mac_spoofchk(struct hclge_dev *hdev, int vf, bool enable)
@@ -9895,7 +9908,6 @@ static void hclge_uninit_ae_dev(struct hnae3_ae_dev *ae_dev)
 	mutex_destroy(&hdev->vport_lock);
 	hclge_uninit_vport_mac_table(hdev);
 	hclge_uninit_vport_vlan_table(hdev);
-	mutex_destroy(&hdev->vport_cfg_mutex);
 	ae_dev->priv = NULL;
 }
 
@@ -10157,10 +10169,8 @@ static int hclge_get_dfx_reg_bd_num(struct hclge_dev *hdev,
 				    int *bd_num_list,
 				    u32 type_num)
 {
-#define HCLGE_DFX_REG_BD_NUM	4
-
 	u32 entries_per_desc, desc_index, index, offset, i;
-	struct hclge_desc desc[HCLGE_DFX_REG_BD_NUM];
+	struct hclge_desc desc[HCLGE_GET_DFX_REG_TYPE_CNT];
 	int ret;
 
 	ret = hclge_query_bd_num_cmd_send(hdev, desc);
@@ -10273,10 +10283,8 @@ static int hclge_get_dfx_reg(struct hclge_dev *hdev, void *data)
 
 	buf_len = sizeof(*desc_src) * bd_num_max;
 	desc_src = kzalloc(buf_len, GFP_KERNEL);
-	if (!desc_src) {
-		dev_err(&hdev->pdev->dev, "%s kzalloc failed\n", __func__);
+	if (!desc_src)
 		return -ENOMEM;
-	}
 
 	for (i = 0; i < dfx_reg_type_num; i++) {
 		bd_num = bd_num_list[i];
@@ -10611,6 +10619,12 @@ static int hclge_init(void)
 {
 	pr_info("%s is initializing\n", HCLGE_NAME);
 
+	hclge_wq = alloc_workqueue("%s", WQ_MEM_RECLAIM, 0, HCLGE_NAME);
+	if (!hclge_wq) {
+		pr_err("%s: failed to create workqueue\n", HCLGE_NAME);
+		return -ENOMEM;
+	}
+
 	hnae3_register_ae_algo(&ae_algo);
 
 	return 0;
@@ -10619,6 +10633,7 @@ static int hclge_init(void)
 static void hclge_exit(void)
 {
 	hnae3_unregister_ae_algo(&ae_algo);
+	destroy_workqueue(hclge_wq);
 }
 module_init(hclge_init);
 module_exit(hclge_exit);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
index ebb4c6e9aed3..f78cbb4cc85e 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h
@@ -139,6 +139,8 @@
 #define HCLGE_PHY_MDIX_STATUS_B		6
 #define HCLGE_PHY_SPEED_DUP_RESOLVE_B	11
 
+#define HCLGE_GET_DFX_REG_TYPE_CNT	4
+
 /* Factor used to calculate offset and bitmap of VF num */
 #define HCLGE_VF_NUM_PER_CMD           64
 
@@ -208,13 +210,14 @@ enum HCLGE_DEV_STATE {
 	HCLGE_STATE_NIC_REGISTERED,
 	HCLGE_STATE_ROCE_REGISTERED,
 	HCLGE_STATE_SERVICE_INITED,
-	HCLGE_STATE_SERVICE_SCHED,
 	HCLGE_STATE_RST_SERVICE_SCHED,
 	HCLGE_STATE_RST_HANDLING,
 	HCLGE_STATE_MBX_SERVICE_SCHED,
 	HCLGE_STATE_MBX_HANDLING,
 	HCLGE_STATE_STATISTICS_UPDATING,
 	HCLGE_STATE_CMD_DISABLE,
+	HCLGE_STATE_LINK_UPDATING,
+	HCLGE_STATE_RST_FAIL,
 	HCLGE_STATE_MAX
 };
 
@@ -454,11 +457,7 @@ struct hclge_mac_stats {
 	u64 mac_rx_ctrl_pkt_num;
 };
 
-#define HCLGE_STATS_TIMER_INTERVAL	(60 * 5)
-struct hclge_hw_stats {
-	struct hclge_mac_stats      mac_stats;
-	u32 stats_timer;
-};
+#define HCLGE_STATS_TIMER_INTERVAL	300UL
 
 struct hclge_vlan_type_cfg {
 	u16 rx_ot_fst_vlan_type;
@@ -549,7 +548,7 @@ struct key_info {
 
 /* assigned by firmware, the real filter number for each pf may be less */
 #define MAX_FD_FILTER_NUM	4096
-#define HCLGE_FD_ARFS_EXPIRE_TIMER_INTERVAL	5
+#define HCLGE_ARFS_EXPIRE_INTERVAL	5UL
 
 enum HCLGE_FD_ACTIVE_RULE_TYPE {
 	HCLGE_FD_RULE_NONE,
@@ -712,7 +711,7 @@ struct hclge_dev {
 	struct hnae3_ae_dev *ae_dev;
 	struct hclge_hw hw;
 	struct hclge_misc_vector misc_vector;
-	struct hclge_hw_stats hw_stats;
+	struct hclge_mac_stats mac_stats;
 	unsigned long state;
 	unsigned long flr_state;
 	unsigned long last_reset_time;
@@ -723,6 +722,7 @@ struct hclge_dev {
 	unsigned long reset_request;	/* reset has been requested */
 	unsigned long reset_pending;	/* client rst is pending to be served */
 	struct hclge_rst_stats rst_stats;
+	struct semaphore reset_sem;	/* protect reset process */
 	u32 fw_version;
 	u16 num_vmdq_vport;		/* Num vmdq vport this PF has set up */
 	u16 num_tqps;			/* Num task queue pairs of this PF */
@@ -774,8 +774,6 @@ struct hclge_dev {
 	unsigned long service_timer_previous;
 	struct timer_list reset_timer;
 	struct delayed_work service_task;
-	struct work_struct rst_service_task;
-	struct work_struct mbx_service_task;
 
 	bool cur_promisc;
 	int num_alloc_vfs;	/* Actual number of VFs allocated */
@@ -811,7 +809,8 @@ struct hclge_dev {
 	struct hlist_head fd_rule_list;
 	spinlock_t fd_rule_lock; /* protect fd_rule_list and fd_bmap */
 	u16 hclge_fd_rule_num;
-	u16 fd_arfs_expire_timer;
+	unsigned long serv_processed_cnt;
+	unsigned long last_serv_processed;
 	unsigned long fd_bmap[BITS_TO_LONGS(MAX_FD_FILTER_NUM)];
 	enum HCLGE_FD_ACTIVE_RULE_TYPE fd_active_type;
 	u8 fd_en;
@@ -825,8 +824,6 @@ struct hclge_dev {
 	u16 share_umv_size;
 	struct mutex umv_mutex; /* protect share_umv_size */
 
-	struct mutex vport_cfg_mutex;   /* Protect stored vf table */
-
 	DECLARE_KFIFO(mac_tnl_log, struct hclge_mac_tnl_stats,
 		      HCLGE_MAC_TNL_LOG_SIZE);
 
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
index 0b433ebe6a2d..a3c0822191a9 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
@@ -86,10 +86,12 @@ static int hclge_send_mbx_msg(struct hclge_vport *vport, u8 *msg, u16 msg_len,
 int hclge_inform_reset_assert_to_vf(struct hclge_vport *vport)
 {
 	struct hclge_dev *hdev = vport->back;
-	enum hnae3_reset_type reset_type;
+	u16 reset_type;
 	u8 msg_data[2];
 	u8 dest_vfid;
 
+	BUILD_BUG_ON(HNAE3_MAX_RESET > U16_MAX);
+
 	dest_vfid = (u8)vport->vport_id;
 
 	if (hdev->reset_type == HNAE3_FUNC_RESET)
@@ -635,7 +637,6 @@ static void hclge_handle_link_change_event(struct hclge_dev *hdev,
 #define LINK_STATUS_OFFSET	1
 #define LINK_FAIL_CODE_OFFSET	2
 
-	clear_bit(HCLGE_STATE_SERVICE_SCHED, &hdev->state);
 	hclge_task_schedule(hdev, 0);
 
 	if (!req->msg[LINK_STATUS_OFFSET])
@@ -798,13 +799,11 @@ void hclge_mbx_handler(struct hclge_dev *hdev)
 			hclge_get_link_mode(vport, req);
 			break;
 		case HCLGE_MBX_GET_VF_FLR_STATUS:
-			mutex_lock(&hdev->vport_cfg_mutex);
 			hclge_rm_vport_all_mac_table(vport, true,
 						     HCLGE_MAC_ADDR_UC);
 			hclge_rm_vport_all_mac_table(vport, true,
 						     HCLGE_MAC_ADDR_MC);
 			hclge_rm_vport_all_vlan_table(vport, true);
-			mutex_unlock(&hdev->vport_cfg_mutex);
 			break;
 		case HCLGE_MBX_GET_MEDIA_TYPE:
 			ret = hclge_get_vf_media_type(vport, req);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c
index af2245e3bb95..f38d236ebf4f 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c
@@ -443,7 +443,7 @@ void hclgevf_cmd_uninit(struct hclgevf_dev *hdev)
 {
 	spin_lock_bh(&hdev->hw.cmq.csq.lock);
 	spin_lock(&hdev->hw.cmq.crq.lock);
-	clear_bit(HCLGEVF_STATE_CMD_DISABLE, &hdev->state);
+	set_bit(HCLGEVF_STATE_CMD_DISABLE, &hdev->state);
 	hclgevf_cmd_uninit_regs(&hdev->hw);
 	spin_unlock(&hdev->hw.cmq.crq.lock);
 	spin_unlock_bh(&hdev->hw.cmq.csq.lock);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
index 25d78a5aaa34..d6597206e692 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
@@ -16,6 +16,8 @@
 static int hclgevf_reset_hdev(struct hclgevf_dev *hdev);
 static struct hnae3_ae_algo ae_algovf;
 
+static struct workqueue_struct *hclgevf_wq;
+
 static const struct pci_device_id ae_algovf_pci_tbl[] = {
 	{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_100G_VF), 0},
 	{PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_100G_RDMA_DCB_PFC_VF), 0},
@@ -440,6 +442,9 @@ void hclgevf_update_link_status(struct hclgevf_dev *hdev, int link_state)
 	struct hnae3_client *rclient;
 	struct hnae3_client *client;
 
+	if (test_and_set_bit(HCLGEVF_STATE_LINK_UPDATING, &hdev->state))
+		return;
+
 	client = handle->client;
 	rclient = hdev->roce_client;
 
@@ -452,6 +457,8 @@ void hclgevf_update_link_status(struct hclgevf_dev *hdev, int link_state)
 			rclient->ops->link_status_change(rhandle, !!link_state);
 		hdev->hw.mac.link = link_state;
 	}
+
+	clear_bit(HCLGEVF_STATE_LINK_UPDATING, &hdev->state);
 }
 
 static void hclgevf_update_link_mode(struct hclgevf_dev *hdev)
@@ -1309,14 +1316,13 @@ static int hclgevf_set_vlan_filter(struct hnae3_handle *handle,
 	msg_data[0] = is_kill;
 	memcpy(&msg_data[1], &vlan_id, sizeof(vlan_id));
 	memcpy(&msg_data[3], &proto, sizeof(proto));
-	ret = hclgevf_send_mbx_msg(hdev, HCLGE_MBX_SET_VLAN,
-				   HCLGE_MBX_VLAN_FILTER, msg_data,
-				   HCLGEVF_VLAN_MBX_MSG_LEN, true, NULL, 0);
-
 	/* when remove hw vlan filter failed, record the vlan id,
 	 * and try to remove it from hw later, to be consistence
 	 * with stack.
 	 */
+	ret = hclgevf_send_mbx_msg(hdev, HCLGE_MBX_SET_VLAN,
+				   HCLGE_MBX_VLAN_FILTER, msg_data,
+				   HCLGEVF_VLAN_MBX_MSG_LEN, true, NULL, 0);
 	if (is_kill && ret)
 		set_bit(vlan_id, hdev->vlan_del_fail_bmap);
 
@@ -1404,32 +1410,6 @@ static int hclgevf_notify_client(struct hclgevf_dev *hdev,
 	return ret;
 }
 
-static void hclgevf_flr_done(struct hnae3_ae_dev *ae_dev)
-{
-	struct hclgevf_dev *hdev = ae_dev->priv;
-
-	set_bit(HNAE3_FLR_DONE, &hdev->flr_state);
-}
-
-static int hclgevf_flr_poll_timeout(struct hclgevf_dev *hdev,
-				    unsigned long delay_us,
-				    unsigned long wait_cnt)
-{
-	unsigned long cnt = 0;
-
-	while (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state) &&
-	       cnt++ < wait_cnt)
-		usleep_range(delay_us, delay_us * 2);
-
-	if (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state)) {
-		dev_err(&hdev->pdev->dev,
-			"flr wait timeout\n");
-		return -ETIMEDOUT;
-	}
-
-	return 0;
-}
-
 static int hclgevf_reset_wait(struct hclgevf_dev *hdev)
 {
 #define HCLGEVF_RESET_WAIT_US	20000
@@ -1440,11 +1420,7 @@ static int hclgevf_reset_wait(struct hclgevf_dev *hdev)
 	u32 val;
 	int ret;
 
-	if (hdev->reset_type == HNAE3_FLR_RESET)
-		return hclgevf_flr_poll_timeout(hdev,
-						HCLGEVF_RESET_WAIT_US,
-						HCLGEVF_RESET_WAIT_CNT);
-	else if (hdev->reset_type == HNAE3_VF_RESET)
+	if (hdev->reset_type == HNAE3_VF_RESET)
 		ret = readl_poll_timeout(hdev->hw.io_base +
 					 HCLGEVF_VF_RST_ING, val,
 					 !(val & HCLGEVF_VF_RST_ING_BIT),
@@ -1516,7 +1492,8 @@ static int hclgevf_reset_stack(struct hclgevf_dev *hdev)
 	/* clear handshake status with IMP */
 	hclgevf_reset_handshake(hdev, false);
 
-	return 0;
+	/* bring up the nic to enable TX/RX again */
+	return hclgevf_notify_client(hdev, HNAE3_UP_CLIENT);
 }
 
 static int hclgevf_reset_prepare_wait(struct hclgevf_dev *hdev)
@@ -1525,18 +1502,10 @@ static int hclgevf_reset_prepare_wait(struct hclgevf_dev *hdev)
 
 	int ret = 0;
 
-	switch (hdev->reset_type) {
-	case HNAE3_VF_FUNC_RESET:
+	if (hdev->reset_type == HNAE3_VF_FUNC_RESET) {
 		ret = hclgevf_send_mbx_msg(hdev, HCLGE_MBX_RESET, 0, NULL,
 					   0, true, NULL, sizeof(u8));
 		hdev->rst_stats.vf_func_rst_cnt++;
-		break;
-	case HNAE3_FLR_RESET:
-		set_bit(HNAE3_FLR_DOWN, &hdev->flr_state);
-		hdev->rst_stats.flr_rst_cnt++;
-		break;
-	default:
-		break;
 	}
 
 	set_bit(HCLGEVF_STATE_CMD_DISABLE, &hdev->state);
@@ -1591,11 +1560,12 @@ static void hclgevf_reset_err_handle(struct hclgevf_dev *hdev)
 		set_bit(HCLGEVF_RESET_PENDING, &hdev->reset_state);
 		hclgevf_reset_task_schedule(hdev);
 	} else {
+		set_bit(HCLGEVF_STATE_RST_FAIL, &hdev->state);
 		hclgevf_dump_rst_info(hdev);
 	}
 }
 
-static int hclgevf_reset(struct hclgevf_dev *hdev)
+static int hclgevf_reset_prepare(struct hclgevf_dev *hdev)
 {
 	struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev);
 	int ret;
@@ -1605,61 +1575,64 @@ static int hclgevf_reset(struct hclgevf_dev *hdev)
 	 */
 	ae_dev->reset_type = hdev->reset_type;
 	hdev->rst_stats.rst_cnt++;
-	rtnl_lock();
 
+	rtnl_lock();
 	/* bring down the nic to stop any ongoing TX/RX */
 	ret = hclgevf_notify_client(hdev, HNAE3_DOWN_CLIENT);
-	if (ret)
-		goto err_reset_lock;
-
 	rtnl_unlock();
-
-	ret = hclgevf_reset_prepare_wait(hdev);
 	if (ret)
-		goto err_reset;
+		return ret;
 
-	/* check if VF could successfully fetch the hardware reset completion
-	 * status from the hardware
-	 */
-	ret = hclgevf_reset_wait(hdev);
-	if (ret) {
-		/* can't do much in this situation, will disable VF */
-		dev_err(&hdev->pdev->dev,
-			"VF failed(=%d) to fetch H/W reset completion status\n",
-			ret);
-		goto err_reset;
-	}
+	return hclgevf_reset_prepare_wait(hdev);
+}
+
+static int hclgevf_reset_rebuild(struct hclgevf_dev *hdev)
+{
+	struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev);
+	int ret;
 
 	hdev->rst_stats.hw_rst_done_cnt++;
 
 	rtnl_lock();
-
 	/* now, re-initialize the nic client and ae device */
 	ret = hclgevf_reset_stack(hdev);
+	rtnl_unlock();
 	if (ret) {
 		dev_err(&hdev->pdev->dev, "failed to reset VF stack\n");
-		goto err_reset_lock;
+		return ret;
 	}
 
-	/* bring up the nic to enable TX/RX again */
-	ret = hclgevf_notify_client(hdev, HNAE3_UP_CLIENT);
-	if (ret)
-		goto err_reset_lock;
-
-	rtnl_unlock();
-
 	hdev->last_reset_time = jiffies;
 	ae_dev->reset_type = HNAE3_NONE_RESET;
 	hdev->rst_stats.rst_done_cnt++;
 	hdev->rst_stats.rst_fail_cnt = 0;
+	clear_bit(HCLGEVF_STATE_RST_FAIL, &hdev->state);
+
+	return 0;
+}
+
+static void hclgevf_reset(struct hclgevf_dev *hdev)
+{
+	if (hclgevf_reset_prepare(hdev))
+		goto err_reset;
+
+	/* check if VF could successfully fetch the hardware reset completion
+	 * status from the hardware
+	 */
+	if (hclgevf_reset_wait(hdev)) {
+		/* can't do much in this situation, will disable VF */
+		dev_err(&hdev->pdev->dev,
+			"failed to fetch H/W reset completion status\n");
+		goto err_reset;
+	}
+
+	if (hclgevf_reset_rebuild(hdev))
+		goto err_reset;
+
+	return;
 
-	return ret;
-err_reset_lock:
-	rtnl_unlock();
 err_reset:
 	hclgevf_reset_err_handle(hdev);
-
-	return ret;
 }
 
 static enum hnae3_reset_type hclgevf_get_reset_level(struct hclgevf_dev *hdev,
@@ -1722,25 +1695,60 @@ static void hclgevf_set_def_reset_request(struct hnae3_ae_dev *ae_dev,
 	set_bit(rst_type, &hdev->default_reset_request);
 }
 
+static void hclgevf_enable_vector(struct hclgevf_misc_vector *vector, bool en)
+{
+	writel(en ? 1 : 0, vector->addr);
+}
+
 static void hclgevf_flr_prepare(struct hnae3_ae_dev *ae_dev)
 {
-#define HCLGEVF_FLR_WAIT_MS	100
-#define HCLGEVF_FLR_WAIT_CNT	50
+#define HCLGEVF_FLR_RETRY_WAIT_MS	500
+#define HCLGEVF_FLR_RETRY_CNT		5
+
 	struct hclgevf_dev *hdev = ae_dev->priv;
-	int cnt = 0;
+	int retry_cnt = 0;
+	int ret;
 
-	clear_bit(HNAE3_FLR_DOWN, &hdev->flr_state);
-	clear_bit(HNAE3_FLR_DONE, &hdev->flr_state);
-	set_bit(HNAE3_FLR_RESET, &hdev->default_reset_request);
-	hclgevf_reset_event(hdev->pdev, NULL);
+retry:
+	down(&hdev->reset_sem);
+	set_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state);
+	hdev->reset_type = HNAE3_FLR_RESET;
+	ret = hclgevf_reset_prepare(hdev);
+	if (ret) {
+		dev_err(&hdev->pdev->dev, "fail to prepare FLR, ret=%d\n",
+			ret);
+		if (hdev->reset_pending ||
+		    retry_cnt++ < HCLGEVF_FLR_RETRY_CNT) {
+			dev_err(&hdev->pdev->dev,
+				"reset_pending:0x%lx, retry_cnt:%d\n",
+				hdev->reset_pending, retry_cnt);
+			clear_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state);
+			up(&hdev->reset_sem);
+			msleep(HCLGEVF_FLR_RETRY_WAIT_MS);
+			goto retry;
+		}
+	}
 
-	while (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state) &&
-	       cnt++ < HCLGEVF_FLR_WAIT_CNT)
-		msleep(HCLGEVF_FLR_WAIT_MS);
+	/* disable misc vector before FLR done */
+	hclgevf_enable_vector(&hdev->misc_vector, false);
+	hdev->rst_stats.flr_rst_cnt++;
+}
 
-	if (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state))
-		dev_err(&hdev->pdev->dev,
-			"flr wait down timeout: %d\n", cnt);
+static void hclgevf_flr_done(struct hnae3_ae_dev *ae_dev)
+{
+	struct hclgevf_dev *hdev = ae_dev->priv;
+	int ret;
+
+	hclgevf_enable_vector(&hdev->misc_vector, true);
+
+	ret = hclgevf_reset_rebuild(hdev);
+	if (ret)
+		dev_warn(&hdev->pdev->dev, "fail to rebuild, ret=%d\n",
+			 ret);
+
+	hdev->reset_type = HNAE3_NONE_RESET;
+	clear_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state);
+	up(&hdev->reset_sem);
 }
 
 static u32 hclgevf_get_fw_version(struct hnae3_handle *handle)
@@ -1767,62 +1775,37 @@ static void hclgevf_get_misc_vector(struct hclgevf_dev *hdev)
 
 void hclgevf_reset_task_schedule(struct hclgevf_dev *hdev)
 {
-	if (!test_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state) &&
-	    !test_bit(HCLGEVF_STATE_REMOVING, &hdev->state)) {
-		set_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state);
-		schedule_work(&hdev->rst_service_task);
-	}
+	if (!test_bit(HCLGEVF_STATE_REMOVING, &hdev->state) &&
+	    !test_and_set_bit(HCLGEVF_STATE_RST_SERVICE_SCHED,
+			      &hdev->state))
+		mod_delayed_work(hclgevf_wq, &hdev->service_task, 0);
 }
 
 void hclgevf_mbx_task_schedule(struct hclgevf_dev *hdev)
 {
-	if (!test_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state) &&
-	    !test_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state)) {
-		set_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state);
-		schedule_work(&hdev->mbx_service_task);
-	}
+	if (!test_bit(HCLGEVF_STATE_REMOVING, &hdev->state) &&
+	    !test_and_set_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED,
+			      &hdev->state))
+		mod_delayed_work(hclgevf_wq, &hdev->service_task, 0);
 }
 
-static void hclgevf_task_schedule(struct hclgevf_dev *hdev)
+static void hclgevf_task_schedule(struct hclgevf_dev *hdev,
+				  unsigned long delay)
 {
-	if (!test_bit(HCLGEVF_STATE_DOWN, &hdev->state)  &&
-	    !test_and_set_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state))
-		schedule_work(&hdev->service_task);
+	if (!test_bit(HCLGEVF_STATE_REMOVING, &hdev->state) &&
+	    !test_bit(HCLGEVF_STATE_RST_FAIL, &hdev->state))
+		mod_delayed_work(hclgevf_wq, &hdev->service_task, delay);
 }
 
-static void hclgevf_deferred_task_schedule(struct hclgevf_dev *hdev)
-{
-	/* if we have any pending mailbox event then schedule the mbx task */
-	if (hdev->mbx_event_pending)
-		hclgevf_mbx_task_schedule(hdev);
-
-	if (test_bit(HCLGEVF_RESET_PENDING, &hdev->reset_state))
-		hclgevf_reset_task_schedule(hdev);
-}
-
-static void hclgevf_service_timer(struct timer_list *t)
-{
-	struct hclgevf_dev *hdev = from_timer(hdev, t, service_timer);
-
-	mod_timer(&hdev->service_timer, jiffies +
-		  HCLGEVF_GENERAL_TASK_INTERVAL * HZ);
-
-	hdev->stats_timer++;
-	hclgevf_task_schedule(hdev);
-}
-
-static void hclgevf_reset_service_task(struct work_struct *work)
+static void hclgevf_reset_service_task(struct hclgevf_dev *hdev)
 {
 #define	HCLGEVF_MAX_RESET_ATTEMPTS_CNT	3
 
-	struct hclgevf_dev *hdev =
-		container_of(work, struct hclgevf_dev, rst_service_task);
-	int ret;
-
-	if (test_and_set_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state))
+	if (!test_and_clear_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state))
 		return;
 
-	clear_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state);
+	down(&hdev->reset_sem);
+	set_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state);
 
 	if (test_and_clear_bit(HCLGEVF_RESET_PENDING,
 			       &hdev->reset_state)) {
@@ -1836,12 +1819,8 @@ static void hclgevf_reset_service_task(struct work_struct *work)
 		hdev->last_reset_time = jiffies;
 		while ((hdev->reset_type =
 			hclgevf_get_reset_level(hdev, &hdev->reset_pending))
-		       != HNAE3_NONE_RESET) {
-			ret = hclgevf_reset(hdev);
-			if (ret)
-				dev_err(&hdev->pdev->dev,
-					"VF stack reset failed %d.\n", ret);
-		}
+		       != HNAE3_NONE_RESET)
+			hclgevf_reset(hdev);
 	} else if (test_and_clear_bit(HCLGEVF_RESET_REQUESTED,
 				      &hdev->reset_state)) {
 		/* we could be here when either of below happens:
@@ -1882,42 +1861,29 @@ static void hclgevf_reset_service_task(struct work_struct *work)
 		hclgevf_reset_task_schedule(hdev);
 	}
 
+	hdev->reset_type = HNAE3_NONE_RESET;
 	clear_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state);
+	up(&hdev->reset_sem);
 }
 
-static void hclgevf_mailbox_service_task(struct work_struct *work)
+static void hclgevf_mailbox_service_task(struct hclgevf_dev *hdev)
 {
-	struct hclgevf_dev *hdev;
-
-	hdev = container_of(work, struct hclgevf_dev, mbx_service_task);
+	if (!test_and_clear_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state))
+		return;
 
 	if (test_and_set_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state))
 		return;
 
-	clear_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state);
-
 	hclgevf_mbx_async_handler(hdev);
 
 	clear_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state);
 }
 
-static void hclgevf_keep_alive_timer(struct timer_list *t)
-{
-	struct hclgevf_dev *hdev = from_timer(hdev, t, keep_alive_timer);
-
-	schedule_work(&hdev->keep_alive_task);
-	mod_timer(&hdev->keep_alive_timer, jiffies +
-		  HCLGEVF_KEEP_ALIVE_TASK_INTERVAL * HZ);
-}
-
-static void hclgevf_keep_alive_task(struct work_struct *work)
+static void hclgevf_keep_alive(struct hclgevf_dev *hdev)
 {
-	struct hclgevf_dev *hdev;
 	u8 respmsg;
 	int ret;
 
-	hdev = container_of(work, struct hclgevf_dev, keep_alive_task);
-
 	if (test_bit(HCLGEVF_STATE_CMD_DISABLE, &hdev->state))
 		return;
 
@@ -1928,19 +1894,32 @@ static void hclgevf_keep_alive_task(struct work_struct *work)
 			"VF sends keep alive cmd failed(=%d)\n", ret);
 }
 
-static void hclgevf_service_task(struct work_struct *work)
+static void hclgevf_periodic_service_task(struct hclgevf_dev *hdev)
 {
-	struct hnae3_handle *handle;
-	struct hclgevf_dev *hdev;
+	unsigned long delta = round_jiffies_relative(HZ);
+	struct hnae3_handle *handle = &hdev->nic;
 
-	hdev = container_of(work, struct hclgevf_dev, service_task);
-	handle = &hdev->nic;
+	if (time_is_after_jiffies(hdev->last_serv_processed + HZ)) {
+		delta = jiffies - hdev->last_serv_processed;
 
-	if (hdev->stats_timer >= HCLGEVF_STATS_TIMER_INTERVAL) {
-		hclgevf_tqps_update_stats(handle);
-		hdev->stats_timer = 0;
+		if (delta < round_jiffies_relative(HZ)) {
+			delta = round_jiffies_relative(HZ) - delta;
+			goto out;
+		}
 	}
 
+	hdev->serv_processed_cnt++;
+	if (!(hdev->serv_processed_cnt % HCLGEVF_KEEP_ALIVE_TASK_INTERVAL))
+		hclgevf_keep_alive(hdev);
+
+	if (test_bit(HCLGEVF_STATE_DOWN, &hdev->state)) {
+		hdev->last_serv_processed = jiffies;
+		goto out;
+	}
+
+	if (!(hdev->serv_processed_cnt % HCLGEVF_STATS_TIMER_INTERVAL))
+		hclgevf_tqps_update_stats(handle);
+
 	/* request the link status from the PF. PF would be able to tell VF
 	 * about such updates in future so we might remove this later
 	 */
@@ -1950,9 +1929,27 @@ static void hclgevf_service_task(struct work_struct *work)
 
 	hclgevf_sync_vlan_filter(hdev);
 
-	hclgevf_deferred_task_schedule(hdev);
+	hdev->last_serv_processed = jiffies;
 
-	clear_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state);
+out:
+	hclgevf_task_schedule(hdev, delta);
+}
+
+static void hclgevf_service_task(struct work_struct *work)
+{
+	struct hclgevf_dev *hdev = container_of(work, struct hclgevf_dev,
+						service_task.work);
+
+	hclgevf_reset_service_task(hdev);
+	hclgevf_mailbox_service_task(hdev);
+	hclgevf_periodic_service_task(hdev);
+
+	/* Handle reset and mbx again in case periodical task delays the
+	 * handling by calling hclgevf_task_schedule() in
+	 * hclgevf_periodic_service_task()
+	 */
+	hclgevf_reset_service_task(hdev);
+	hclgevf_mailbox_service_task(hdev);
 }
 
 static void hclgevf_clear_event_cause(struct hclgevf_dev *hdev, u32 regclr)
@@ -2010,11 +2007,6 @@ static enum hclgevf_evt_cause hclgevf_check_evt_cause(struct hclgevf_dev *hdev,
 	return HCLGEVF_VECTOR0_EVENT_OTHER;
 }
 
-static void hclgevf_enable_vector(struct hclgevf_misc_vector *vector, bool en)
-{
-	writel(en ? 1 : 0, vector->addr);
-}
-
 static irqreturn_t hclgevf_misc_irq_handle(int irq, void *data)
 {
 	enum hclgevf_evt_cause event_cause;
@@ -2189,16 +2181,31 @@ static int hclgevf_init_vlan_config(struct hclgevf_dev *hdev)
 				       false);
 }
 
+static void hclgevf_flush_link_update(struct hclgevf_dev *hdev)
+{
+#define HCLGEVF_FLUSH_LINK_TIMEOUT	100000
+
+	unsigned long last = hdev->serv_processed_cnt;
+	int i = 0;
+
+	while (test_bit(HCLGEVF_STATE_LINK_UPDATING, &hdev->state) &&
+	       i++ < HCLGEVF_FLUSH_LINK_TIMEOUT &&
+	       last == hdev->serv_processed_cnt)
+		usleep_range(1, 1);
+}
+
 static void hclgevf_set_timer_task(struct hnae3_handle *handle, bool enable)
 {
 	struct hclgevf_dev *hdev = hclgevf_ae_get_hdev(handle);
 
 	if (enable) {
-		mod_timer(&hdev->service_timer, jiffies + HZ);
+		hclgevf_task_schedule(hdev, 0);
 	} else {
-		del_timer_sync(&hdev->service_timer);
-		cancel_work_sync(&hdev->service_task);
-		clear_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state);
+		set_bit(HCLGEVF_STATE_DOWN, &hdev->state);
+
+		/* flush memory to make sure DOWN is seen by service task */
+		smp_mb__before_atomic();
+		hclgevf_flush_link_update(hdev);
 	}
 }
 
@@ -2245,16 +2252,12 @@ static int hclgevf_set_alive(struct hnae3_handle *handle, bool alive)
 
 static int hclgevf_client_start(struct hnae3_handle *handle)
 {
-	struct hclgevf_dev *hdev = hclgevf_ae_get_hdev(handle);
 	int ret;
 
 	ret = hclgevf_set_alive(handle, true);
 	if (ret)
 		return ret;
 
-	mod_timer(&hdev->keep_alive_timer, jiffies +
-		  HCLGEVF_KEEP_ALIVE_TASK_INTERVAL * HZ);
-
 	return 0;
 }
 
@@ -2267,27 +2270,18 @@ static void hclgevf_client_stop(struct hnae3_handle *handle)
 	if (ret)
 		dev_warn(&hdev->pdev->dev,
 			 "%s failed %d\n", __func__, ret);
-
-	del_timer_sync(&hdev->keep_alive_timer);
-	cancel_work_sync(&hdev->keep_alive_task);
 }
 
 static void hclgevf_state_init(struct hclgevf_dev *hdev)
 {
-	/* setup tasks for the MBX */
-	INIT_WORK(&hdev->mbx_service_task, hclgevf_mailbox_service_task);
 	clear_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state);
 	clear_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state);
+	clear_bit(HCLGEVF_STATE_RST_FAIL, &hdev->state);
 
-	/* setup tasks for service timer */
-	timer_setup(&hdev->service_timer, hclgevf_service_timer, 0);
-
-	INIT_WORK(&hdev->service_task, hclgevf_service_task);
-	clear_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state);
-
-	INIT_WORK(&hdev->rst_service_task, hclgevf_reset_service_task);
+	INIT_DELAYED_WORK(&hdev->service_task, hclgevf_service_task);
 
 	mutex_init(&hdev->mbx_resp.mbx_mutex);
+	sema_init(&hdev->reset_sem, 1);
 
 	/* bring the device down */
 	set_bit(HCLGEVF_STATE_DOWN, &hdev->state);
@@ -2298,18 +2292,8 @@ static void hclgevf_state_uninit(struct hclgevf_dev *hdev)
 	set_bit(HCLGEVF_STATE_DOWN, &hdev->state);
 	set_bit(HCLGEVF_STATE_REMOVING, &hdev->state);
 
-	if (hdev->keep_alive_timer.function)
-		del_timer_sync(&hdev->keep_alive_timer);
-	if (hdev->keep_alive_task.func)
-		cancel_work_sync(&hdev->keep_alive_task);
-	if (hdev->service_timer.function)
-		del_timer_sync(&hdev->service_timer);
-	if (hdev->service_task.func)
-		cancel_work_sync(&hdev->service_task);
-	if (hdev->mbx_service_task.func)
-		cancel_work_sync(&hdev->mbx_service_task);
-	if (hdev->rst_service_task.func)
-		cancel_work_sync(&hdev->rst_service_task);
+	if (hdev->service_task.work.func)
+		cancel_delayed_work_sync(&hdev->service_task);
 
 	mutex_destroy(&hdev->mbx_resp.mbx_mutex);
 }
@@ -2383,8 +2367,10 @@ static int hclgevf_misc_irq_init(struct hclgevf_dev *hdev)
 
 	hclgevf_get_misc_vector(hdev);
 
+	snprintf(hdev->misc_vector.name, HNAE3_INT_NAME_LEN, "%s-misc-%s",
+		 HCLGEVF_NAME, pci_name(hdev->pdev));
 	ret = request_irq(hdev->misc_vector.vector_irq, hclgevf_misc_irq_handle,
-			  0, "hclgevf_cmd", hdev);
+			  0, hdev->misc_vector.name, hdev);
 	if (ret) {
 		dev_err(&hdev->pdev->dev, "VF failed to request misc irq(%d)\n",
 			hdev->misc_vector.vector_irq);
@@ -2611,11 +2597,11 @@ static int hclgevf_query_vf_resource(struct hclgevf_dev *hdev)
 
 	if (hnae3_dev_roce_supported(hdev)) {
 		hdev->roce_base_msix_offset =
-		hnae3_get_field(__le16_to_cpu(req->msixcap_localid_ba_rocee),
+		hnae3_get_field(le16_to_cpu(req->msixcap_localid_ba_rocee),
 				HCLGEVF_MSIX_OFT_ROCEE_M,
 				HCLGEVF_MSIX_OFT_ROCEE_S);
 		hdev->num_roce_msix =
-		hnae3_get_field(__le16_to_cpu(req->vf_intr_vector_number),
+		hnae3_get_field(le16_to_cpu(req->vf_intr_vector_number),
 				HCLGEVF_VEC_NUM_M, HCLGEVF_VEC_NUM_S);
 
 		/* nic's msix numbers is always equals to the roce's. */
@@ -2628,7 +2614,7 @@ static int hclgevf_query_vf_resource(struct hclgevf_dev *hdev)
 				hdev->roce_base_msix_offset;
 	} else {
 		hdev->num_msi =
-		hnae3_get_field(__le16_to_cpu(req->vf_intr_vector_number),
+		hnae3_get_field(le16_to_cpu(req->vf_intr_vector_number),
 				HCLGEVF_VEC_NUM_M, HCLGEVF_VEC_NUM_S);
 
 		hdev->num_nic_msix = hdev->num_msi;
@@ -2725,16 +2711,12 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev)
 	int ret;
 
 	ret = hclgevf_pci_init(hdev);
-	if (ret) {
-		dev_err(&pdev->dev, "PCI initialization failed\n");
+	if (ret)
 		return ret;
-	}
 
 	ret = hclgevf_cmd_queue_init(hdev);
-	if (ret) {
-		dev_err(&pdev->dev, "Cmd queue init failed: %d\n", ret);
+	if (ret)
 		goto err_cmd_queue_init;
-	}
 
 	ret = hclgevf_cmd_init(hdev);
 	if (ret)
@@ -2742,11 +2724,8 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev)
 
 	/* Get vf resource */
 	ret = hclgevf_query_vf_resource(hdev);
-	if (ret) {
-		dev_err(&hdev->pdev->dev,
-			"Query vf status error, ret = %d.\n", ret);
+	if (ret)
 		goto err_cmd_init;
-	}
 
 	ret = hclgevf_init_msi(hdev);
 	if (ret) {
@@ -2756,13 +2735,11 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev)
 
 	hclgevf_state_init(hdev);
 	hdev->reset_level = HNAE3_VF_FUNC_RESET;
+	hdev->reset_type = HNAE3_NONE_RESET;
 
 	ret = hclgevf_misc_irq_init(hdev);
-	if (ret) {
-		dev_err(&pdev->dev, "failed(%d) to init Misc IRQ(vector0)\n",
-			ret);
+	if (ret)
 		goto err_misc_irq_init;
-	}
 
 	set_bit(HCLGEVF_STATE_IRQ_INITED, &hdev->state);
 
@@ -2779,10 +2756,8 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev)
 	}
 
 	ret = hclgevf_set_handle_info(hdev);
-	if (ret) {
-		dev_err(&pdev->dev, "failed(%d) to set handle info\n", ret);
+	if (ret)
 		goto err_config;
-	}
 
 	ret = hclgevf_config_gro(hdev, true);
 	if (ret)
@@ -2807,6 +2782,8 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev)
 	dev_info(&hdev->pdev->dev, "finished initializing %s driver\n",
 		 HCLGEVF_DRIVER_NAME);
 
+	hclgevf_task_schedule(hdev, round_jiffies_relative(HZ));
+
 	return 0;
 
 err_config:
@@ -2838,7 +2815,6 @@ static void hclgevf_uninit_hdev(struct hclgevf_dev *hdev)
 static int hclgevf_init_ae_dev(struct hnae3_ae_dev *ae_dev)
 {
 	struct pci_dev *pdev = ae_dev->pdev;
-	struct hclgevf_dev *hdev;
 	int ret;
 
 	ret = hclgevf_alloc_hdev(ae_dev);
@@ -2853,10 +2829,6 @@ static int hclgevf_init_ae_dev(struct hnae3_ae_dev *ae_dev)
 		return ret;
 	}
 
-	hdev = ae_dev->priv;
-	timer_setup(&hdev->keep_alive_timer, hclgevf_keep_alive_timer, 0);
-	INIT_WORK(&hdev->keep_alive_task, hclgevf_keep_alive_task);
-
 	return 0;
 }
 
@@ -3213,6 +3185,12 @@ static int hclgevf_init(void)
 {
 	pr_info("%s is initializing\n", HCLGEVF_NAME);
 
+	hclgevf_wq = alloc_workqueue("%s", WQ_MEM_RECLAIM, 0, HCLGEVF_NAME);
+	if (!hclgevf_wq) {
+		pr_err("%s: failed to create workqueue\n", HCLGEVF_NAME);
+		return -ENOMEM;
+	}
+
 	hnae3_register_ae_algo(&ae_algovf);
 
 	return 0;
@@ -3221,6 +3199,7 @@ static int hclgevf_init(void)
 static void hclgevf_exit(void)
 {
 	hnae3_unregister_ae_algo(&ae_algovf);
+	destroy_workqueue(hclgevf_wq);
 }
 module_init(hclgevf_init);
 module_exit(hclgevf_exit);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h
index 2f4c81bf4169..fee8d97f323c 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h
@@ -142,12 +142,13 @@ enum hclgevf_states {
 	HCLGEVF_STATE_REMOVING,
 	HCLGEVF_STATE_NIC_REGISTERED,
 	/* task states */
-	HCLGEVF_STATE_SERVICE_SCHED,
 	HCLGEVF_STATE_RST_SERVICE_SCHED,
 	HCLGEVF_STATE_RST_HANDLING,
 	HCLGEVF_STATE_MBX_SERVICE_SCHED,
 	HCLGEVF_STATE_MBX_HANDLING,
 	HCLGEVF_STATE_CMD_DISABLE,
+	HCLGEVF_STATE_LINK_UPDATING,
+	HCLGEVF_STATE_RST_FAIL,
 };
 
 struct hclgevf_mac {
@@ -220,6 +221,7 @@ struct hclgevf_rss_cfg {
 struct hclgevf_misc_vector {
 	u8 __iomem *addr;
 	int vector_irq;
+	char name[HNAE3_INT_NAME_LEN];
 };
 
 struct hclgevf_rst_stats {
@@ -251,6 +253,7 @@ struct hclgevf_dev {
 	unsigned long reset_state;	/* requested, pending */
 	struct hclgevf_rst_stats rst_stats;
 	u32 reset_attempts;
+	struct semaphore reset_sem;	/* protect reset process */
 
 	u32 fw_version;
 	u16 num_tqps;		/* num task queue pairs of this PF */
@@ -283,12 +286,7 @@ struct hclgevf_dev {
 	struct hclgevf_mbx_resp_status mbx_resp; /* mailbox response */
 	struct hclgevf_mbx_arq_ring arq; /* mailbox async rx queue */
 
-	struct timer_list service_timer;
-	struct timer_list keep_alive_timer;
-	struct work_struct service_task;
-	struct work_struct keep_alive_task;
-	struct work_struct rst_service_task;
-	struct work_struct mbx_service_task;
+	struct delayed_work service_task;
 
 	struct hclgevf_tqp *htqp;
 
@@ -298,7 +296,8 @@ struct hclgevf_dev {
 	struct hnae3_client *nic_client;
 	struct hnae3_client *roce_client;
 	u32 flag;
-	u32 stats_timer;
+	unsigned long serv_processed_cnt;
+	unsigned long last_serv_processed;
 };
 
 static inline bool hclgevf_is_reset_pending(struct hclgevf_dev *hdev)