Merge branch 'net-hns3-some-code-optimizations-bugfixes'

Huazhong Tan says:

====================
net: hns3: some code optimizations & bugfixes

This patch-set includes code optimizations and bugfixes for
the HNS3 ethernet controller driver.

[patch 1/11] fixes a selftest issue when doing autoneg.

[patch 2/11 - 3-11] adds two code optimizations about VLAN issue.

[patch 4/11] restores the MAC autoneg state after reset.

[patch 5/11 - 8/11] adds some code optimizations and bugfixes about
HW errors handling.

[patch 9/11 - 11/11] fixes some issues related to driver loading and
unloading.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

drivers/net/ethernet/hisilicon/hns3/hnae3.h

@@ -213,7 +213,6 @@ struct hnae3_ae_dev {
 	const struct hnae3_ae_ops *ops;
 	struct list_head node;
 	u32 flag;
-	u8 override_pci_need_reset; /* fix to stop multiple reset happening */
 	unsigned long hw_err_reset_req;
 	enum hnae3_reset_type reset_type;
 	void *priv;
@@ -264,6 +263,8 @@ struct hnae3_ae_dev {
  *   get auto autonegotiation of pause frame use
  * restart_autoneg()
  *   restart autonegotiation
+ * halt_autoneg()
+ *   halt/resume autonegotiation when autonegotiation on
  * get_coalesce_usecs()
  *   get usecs to delay a TX interrupt after a packet is sent
  * get_rx_max_coalesced_frames()
@@ -383,6 +384,7 @@ struct hnae3_ae_ops {
 	int (*set_autoneg)(struct hnae3_handle *handle, bool enable);
 	int (*get_autoneg)(struct hnae3_handle *handle);
 	int (*restart_autoneg)(struct hnae3_handle *handle);
+	int (*halt_autoneg)(struct hnae3_handle *handle, bool halt);
 
 	void (*get_coalesce_usecs)(struct hnae3_handle *handle,
 				   u32 *tx_usecs, u32 *rx_usecs);

drivers/net/ethernet/hisilicon/hns3/hns3_enet.c

@@ -1950,7 +1950,7 @@ static pci_ers_result_t hns3_slot_reset(struct pci_dev *pdev)
 	ops = ae_dev->ops;
 	/* request the reset */
 	if (ops->reset_event) {
-		if (!ae_dev->override_pci_need_reset) {
+		if (ae_dev->hw_err_reset_req) {
 			reset_type = ops->get_reset_level(ae_dev,
 						&ae_dev->hw_err_reset_req);
 			ops->set_default_reset_request(ae_dev, reset_type);

drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c

@@ -336,6 +336,13 @@ static void hns3_self_test(struct net_device *ndev,
 		h->ae_algo->ops->enable_vlan_filter(h, false);
 #endif
 
+	/* Tell firmware to stop mac autoneg before loopback test start,
+	 * otherwise loopback test may be failed when the port is still
+	 * negotiating.
+	 */
+	if (h->ae_algo->ops->halt_autoneg)
+		h->ae_algo->ops->halt_autoneg(h, true);
+
 	set_bit(HNS3_NIC_STATE_TESTING, &priv->state);
 
 	for (i = 0; i < HNS3_SELF_TEST_TYPE_NUM; i++) {
@@ -358,6 +365,9 @@ static void hns3_self_test(struct net_device *ndev,
 
 	clear_bit(HNS3_NIC_STATE_TESTING, &priv->state);
 
+	if (h->ae_algo->ops->halt_autoneg)
+		h->ae_algo->ops->halt_autoneg(h, false);
+
 #if IS_ENABLED(CONFIG_VLAN_8021Q)
 	if (dis_vlan_filter)
 		h->ae_algo->ops->enable_vlan_filter(h, true);

drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c

@@ -1060,6 +1060,52 @@ static int hclge_config_ssu_hw_err_int(struct hclge_dev *hdev, bool en)
 	return ret;
 }
 
+/* hclge_query_bd_num: query number of buffer descriptors
+ * @hdev: pointer to struct hclge_dev
+ * @is_ras: true for ras, false for msix
+ * @mpf_bd_num: number of main PF interrupt buffer descriptors
+ * @pf_bd_num: number of not main PF interrupt buffer descriptors
+ *
+ * This function querys number of mpf and pf buffer descriptors.
+ */
+static int hclge_query_bd_num(struct hclge_dev *hdev, bool is_ras,
+			      int *mpf_bd_num, int *pf_bd_num)
+{
+	struct device *dev = &hdev->pdev->dev;
+	u32 mpf_min_bd_num, pf_min_bd_num;
+	enum hclge_opcode_type opcode;
+	struct hclge_desc desc_bd;
+	int ret;
+
+	if (is_ras) {
+		opcode = HCLGE_QUERY_RAS_INT_STS_BD_NUM;
+		mpf_min_bd_num = HCLGE_MPF_RAS_INT_MIN_BD_NUM;
+		pf_min_bd_num = HCLGE_PF_RAS_INT_MIN_BD_NUM;
+	} else {
+		opcode = HCLGE_QUERY_MSIX_INT_STS_BD_NUM;
+		mpf_min_bd_num = HCLGE_MPF_MSIX_INT_MIN_BD_NUM;
+		pf_min_bd_num = HCLGE_PF_MSIX_INT_MIN_BD_NUM;
+	}
+
+	hclge_cmd_setup_basic_desc(&desc_bd, opcode, true);
+	ret = hclge_cmd_send(&hdev->hw, &desc_bd, 1);
+	if (ret) {
+		dev_err(dev, "fail(%d) to query msix int status bd num\n",
+			ret);
+		return ret;
+	}
+
+	*mpf_bd_num = le32_to_cpu(desc_bd.data[0]);
+	*pf_bd_num = le32_to_cpu(desc_bd.data[1]);
+	if (*mpf_bd_num < mpf_min_bd_num || *pf_bd_num < pf_min_bd_num) {
+		dev_err(dev, "Invalid bd num: mpf(%d), pf(%d)\n",
+			*mpf_bd_num, *pf_bd_num);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
 /* hclge_handle_mpf_ras_error: handle all main PF RAS errors
  * @hdev: pointer to struct hclge_dev
  * @desc: descriptor for describing the command
@@ -1291,24 +1337,16 @@ static int hclge_handle_pf_ras_error(struct hclge_dev *hdev,
 
 static int hclge_handle_all_ras_errors(struct hclge_dev *hdev)
 {
-	struct device *dev = &hdev->pdev->dev;
 	u32 mpf_bd_num, pf_bd_num, bd_num;
-	struct hclge_desc desc_bd;
 	struct hclge_desc *desc;
 	int ret;
 
 	/* query the number of registers in the RAS int status */
-	hclge_cmd_setup_basic_desc(&desc_bd, HCLGE_QUERY_RAS_INT_STS_BD_NUM,
-				   true);
-	ret = hclge_cmd_send(&hdev->hw, &desc_bd, 1);
-	if (ret) {
-		dev_err(dev, "fail(%d) to query ras int status bd num\n", ret);
+	ret = hclge_query_bd_num(hdev, true, &mpf_bd_num, &pf_bd_num);
+	if (ret)
 		return ret;
-	}
-	mpf_bd_num = le32_to_cpu(desc_bd.data[0]);
-	pf_bd_num = le32_to_cpu(desc_bd.data[1]);
-	bd_num = max_t(u32, mpf_bd_num, pf_bd_num);
 
+	bd_num = max_t(u32, mpf_bd_num, pf_bd_num);
 	desc = kcalloc(bd_num, sizeof(struct hclge_desc), GFP_KERNEL);
 	if (!desc)
 		return -ENOMEM;
@@ -1606,6 +1644,8 @@ pci_ers_result_t hclge_handle_hw_ras_error(struct hnae3_ae_dev *ae_dev)
 	if (status & HCLGE_RAS_REG_NFE_MASK ||
 	    status & HCLGE_RAS_REG_ROCEE_ERR_MASK)
 		ae_dev->hw_err_reset_req = 0;
+	else
+		goto out;
 
 	/* Handling Non-fatal HNS RAS errors */
 	if (status & HCLGE_RAS_REG_NFE_MASK) {
@@ -1613,27 +1653,22 @@ pci_ers_result_t hclge_handle_hw_ras_error(struct hnae3_ae_dev *ae_dev)
 			 "HNS Non-Fatal RAS error(status=0x%x) identified\n",
 			 status);
 		hclge_handle_all_ras_errors(hdev);
-	} else {
-		if (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state) ||
-		    hdev->pdev->revision < 0x21) {
-			ae_dev->override_pci_need_reset = 1;
-			return PCI_ERS_RESULT_RECOVERED;
-		}
 	}
 
-	if (status & HCLGE_RAS_REG_ROCEE_ERR_MASK) {
-		dev_warn(dev, "ROCEE uncorrected RAS error identified\n");
+	/* Handling Non-fatal Rocee RAS errors */
+	if (hdev->pdev->revision >= 0x21 &&
+	    status & HCLGE_RAS_REG_ROCEE_ERR_MASK) {
+		dev_warn(dev, "ROCEE Non-Fatal RAS error identified\n");
 		hclge_handle_rocee_ras_error(ae_dev);
 	}
 
-	if ((status & HCLGE_RAS_REG_NFE_MASK ||
-	     status & HCLGE_RAS_REG_ROCEE_ERR_MASK) &&
-	     ae_dev->hw_err_reset_req) {
-		ae_dev->override_pci_need_reset = 0;
+	if (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state))
+		goto out;
+
+	if (ae_dev->hw_err_reset_req)
 		return PCI_ERS_RESULT_NEED_RESET;
-	}
-	ae_dev->override_pci_need_reset = 1;
 
+out:
 	return PCI_ERS_RESULT_RECOVERED;
 }
 
@@ -1847,28 +1882,21 @@ static int hclge_handle_all_hw_msix_error(struct hclge_dev *hdev,
 	struct hclge_mac_tnl_stats mac_tnl_stats;
 	struct device *dev = &hdev->pdev->dev;
 	u32 mpf_bd_num, pf_bd_num, bd_num;
-	struct hclge_desc desc_bd;
 	struct hclge_desc *desc;
 	u32 status;
 	int ret;
 
 	/* query the number of bds for the MSIx int status */
-	hclge_cmd_setup_basic_desc(&desc_bd, HCLGE_QUERY_MSIX_INT_STS_BD_NUM,
-				   true);
-	ret = hclge_cmd_send(&hdev->hw, &desc_bd, 1);
-	if (ret) {
-		dev_err(dev, "fail(%d) to query msix int status bd num\n",
-			ret);
-		return ret;
-	}
+	ret = hclge_query_bd_num(hdev, false, &mpf_bd_num, &pf_bd_num);
+	if (ret)
+		goto out;
 
-	mpf_bd_num = le32_to_cpu(desc_bd.data[0]);
-	pf_bd_num = le32_to_cpu(desc_bd.data[1]);
 	bd_num = max_t(u32, mpf_bd_num, pf_bd_num);
-
 	desc = kcalloc(bd_num, sizeof(struct hclge_desc), GFP_KERNEL);
-	if (!desc)
+	if (!desc) {
+		ret = -ENOMEM;
 		goto out;
+	}
 
 	ret = hclge_handle_mpf_msix_error(hdev, desc, mpf_bd_num,
 					  reset_requests);
@@ -1931,7 +1959,6 @@ void hclge_handle_all_hns_hw_errors(struct hnae3_ae_dev *ae_dev)
 	struct hclge_dev *hdev = ae_dev->priv;
 	struct device *dev = &hdev->pdev->dev;
 	u32 mpf_bd_num, pf_bd_num, bd_num;
-	struct hclge_desc desc_bd;
 	struct hclge_desc *desc;
 	u32 status;
 	int ret;
@@ -1940,19 +1967,11 @@ void hclge_handle_all_hns_hw_errors(struct hnae3_ae_dev *ae_dev)
 	status = hclge_read_dev(&hdev->hw, HCLGE_RAS_PF_OTHER_INT_STS_REG);
 
 	/* query the number of bds for the MSIx int status */
-	hclge_cmd_setup_basic_desc(&desc_bd, HCLGE_QUERY_MSIX_INT_STS_BD_NUM,
-				   true);
-	ret = hclge_cmd_send(&hdev->hw, &desc_bd, 1);
-	if (ret) {
-		dev_err(dev, "fail(%d) to query msix int status bd num\n",
-			ret);
+	ret = hclge_query_bd_num(hdev, false, &mpf_bd_num, &pf_bd_num);
+	if (ret)
 		return;
-	}
 
-	mpf_bd_num = le32_to_cpu(desc_bd.data[0]);
-	pf_bd_num = le32_to_cpu(desc_bd.data[1]);
 	bd_num = max_t(u32, mpf_bd_num, pf_bd_num);
-
 	desc = kcalloc(bd_num, sizeof(struct hclge_desc), GFP_KERNEL);
 	if (!desc)
 		return;

drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.h

@@ -6,6 +6,11 @@
 
 #include "hclge_main.h"
 
+#define HCLGE_MPF_RAS_INT_MIN_BD_NUM	10
+#define HCLGE_PF_RAS_INT_MIN_BD_NUM	4
+#define HCLGE_MPF_MSIX_INT_MIN_BD_NUM	10
+#define HCLGE_PF_MSIX_INT_MIN_BD_NUM	4
+
 #define HCLGE_RAS_PF_OTHER_INT_STS_REG   0x20B00
 #define HCLGE_RAS_REG_NFE_MASK   0xFF00
 #define HCLGE_RAS_REG_ROCEE_ERR_MASK   0x3000000

drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c

@@ -35,6 +35,7 @@
 
 static int hclge_set_mac_mtu(struct hclge_dev *hdev, int new_mps);
 static int hclge_init_vlan_config(struct hclge_dev *hdev);
+static void hclge_sync_vlan_filter(struct hclge_dev *hdev);
 static int hclge_reset_ae_dev(struct hnae3_ae_dev *ae_dev);
 static bool hclge_get_hw_reset_stat(struct hnae3_handle *handle);
 static int hclge_set_umv_space(struct hclge_dev *hdev, u16 space_size,
@@ -2315,6 +2316,17 @@ static int hclge_restart_autoneg(struct hnae3_handle *handle)
 	return hclge_notify_client(hdev, HNAE3_UP_CLIENT);
 }
 
+static int hclge_halt_autoneg(struct hnae3_handle *handle, bool halt)
+{
+	struct hclge_vport *vport = hclge_get_vport(handle);
+	struct hclge_dev *hdev = vport->back;
+
+	if (hdev->hw.mac.support_autoneg && hdev->hw.mac.autoneg)
+		return hclge_set_autoneg_en(hdev, !halt);
+
+	return 0;
+}
+
 static int hclge_set_fec_hw(struct hclge_dev *hdev, u32 fec_mode)
 {
 	struct hclge_config_fec_cmd *req;
@@ -2388,6 +2400,15 @@ static int hclge_mac_init(struct hclge_dev *hdev)
 		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);
+			return ret;
+		}
+	}
+
 	mac->link = 0;
 
 	if (mac->user_fec_mode & BIT(HNAE3_FEC_USER_DEF)) {
@@ -3528,6 +3549,7 @@ static void hclge_service_task(struct work_struct *work)
 	hclge_update_port_info(hdev);
 	hclge_update_link_status(hdev);
 	hclge_update_vport_alive(hdev);
+	hclge_sync_vlan_filter(hdev);
 	if (hdev->fd_arfs_expire_timer >= HCLGE_FD_ARFS_EXPIRE_TIMER_INTERVAL) {
 		hclge_rfs_filter_expire(hdev);
 		hdev->fd_arfs_expire_timer = 0;
@@ -7101,12 +7123,13 @@ static int hclge_set_vf_vlan_common(struct hclge_dev *hdev, u16 vfid,
 		if (!req0->resp_code)
 			return 0;
 
-		if (req0->resp_code == HCLGE_VF_VLAN_DEL_NO_FOUND) {
-			dev_warn(&hdev->pdev->dev,
-				 "vlan %d filter is not in vf vlan table\n",
-				 vlan);
+		/* vf vlan filter is disabled when vf vlan table is full,
+		 * then new vlan id will not be added into vf vlan table.
+		 * Just return 0 without warning, avoid massive verbose
+		 * print logs when unload.
+		 */
+		if (req0->resp_code == HCLGE_VF_VLAN_DEL_NO_FOUND)
 			return 0;
-		}
 
 		dev_err(&hdev->pdev->dev,
 			"Kill vf vlan filter fail, ret =%d.\n",
@@ -7730,11 +7753,20 @@ int hclge_set_vlan_filter(struct hnae3_handle *handle, __be16 proto,
 	bool writen_to_tbl = false;
 	int ret = 0;
 
-	/* when port based VLAN enabled, we use port based VLAN as the VLAN
-	 * filter entry. In this case, we don't update VLAN filter table
-	 * when user add new VLAN or remove exist VLAN, just update the vport
-	 * VLAN list. The VLAN id in VLAN list won't be writen in VLAN filter
-	 * table until port based VLAN disabled
+	/* When device is resetting, firmware is unable to handle
+	 * mailbox. Just record the vlan id, and remove it after
+	 * reset finished.
+	 */
+	if (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state) && is_kill) {
+		set_bit(vlan_id, vport->vlan_del_fail_bmap);
+		return -EBUSY;
+	}
+
+	/* When port base vlan enabled, we use port base vlan as the vlan
+	 * filter entry. In this case, we don't update vlan filter table
+	 * when user add new vlan or remove exist vlan, just update the vport
+	 * vlan list. The vlan id in vlan list will be writen in vlan filter
+	 * table until port base vlan disabled
 	 */
 	if (handle->port_base_vlan_state == HNAE3_PORT_BASE_VLAN_DISABLE) {
 		ret = hclge_set_vlan_filter_hw(hdev, proto, vport->vport_id,
@@ -7742,16 +7774,53 @@ int hclge_set_vlan_filter(struct hnae3_handle *handle, __be16 proto,
 		writen_to_tbl = true;
 	}
 
-	if (ret)
-		return ret;
+	if (!ret) {
+		if (is_kill)
+			hclge_rm_vport_vlan_table(vport, vlan_id, false);
+		else
+			hclge_add_vport_vlan_table(vport, vlan_id,
+						   writen_to_tbl);
+	} else if (is_kill) {
+		/* When remove hw vlan filter failed, record the vlan id,
+		 * and try to remove it from hw later, to be consistence
+		 * with stack
+		 */
+		set_bit(vlan_id, vport->vlan_del_fail_bmap);
+	}
+	return ret;
+}
 
-	if (is_kill)
-		hclge_rm_vport_vlan_table(vport, vlan_id, false);
-	else
-		hclge_add_vport_vlan_table(vport, vlan_id,
-					   writen_to_tbl);
+static void hclge_sync_vlan_filter(struct hclge_dev *hdev)
+{
+#define HCLGE_MAX_SYNC_COUNT	60
 
-	return 0;
+	int i, ret, sync_cnt = 0;
+	u16 vlan_id;
+
+	/* start from vport 1 for PF is always alive */
+	for (i = 0; i < hdev->num_alloc_vport; i++) {
+		struct hclge_vport *vport = &hdev->vport[i];
+
+		vlan_id = find_first_bit(vport->vlan_del_fail_bmap,
+					 VLAN_N_VID);
+		while (vlan_id != VLAN_N_VID) {
+			ret = hclge_set_vlan_filter_hw(hdev, htons(ETH_P_8021Q),
+						       vport->vport_id, vlan_id,
+						       0, true);
+			if (ret && ret != -EINVAL)
+				return;
+
+			clear_bit(vlan_id, vport->vlan_del_fail_bmap);
+			hclge_rm_vport_vlan_table(vport, vlan_id, false);
+
+			sync_cnt++;
+			if (sync_cnt >= HCLGE_MAX_SYNC_COUNT)
+				return;
+
+			vlan_id = find_first_bit(vport->vlan_del_fail_bmap,
+						 VLAN_N_VID);
+		}
+	}
 }
 
 static int hclge_set_mac_mtu(struct hclge_dev *hdev, int new_mps)
@@ -8213,25 +8282,44 @@ static int hclge_init_nic_client_instance(struct hnae3_ae_dev *ae_dev,
 {
 	struct hnae3_client *client = vport->nic.client;
 	struct hclge_dev *hdev = ae_dev->priv;
+	int rst_cnt;
 	int ret;
 
+	rst_cnt = hdev->rst_stats.reset_cnt;
 	ret = client->ops->init_instance(&vport->nic);
 	if (ret)
 		return ret;
 
 	set_bit(HCLGE_STATE_NIC_REGISTERED, &hdev->state);
-	hnae3_set_client_init_flag(client, ae_dev, 1);
+	if (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state) ||
+	    rst_cnt != hdev->rst_stats.reset_cnt) {
+		ret = -EBUSY;
+		goto init_nic_err;
+	}
 
 	/* Enable nic hw error interrupts */
 	ret = hclge_config_nic_hw_error(hdev, true);
-	if (ret)
+	if (ret) {
 		dev_err(&ae_dev->pdev->dev,
 			"fail(%d) to enable hw error interrupts\n", ret);
+		goto init_nic_err;
+	}
+
+	hnae3_set_client_init_flag(client, ae_dev, 1);
 
 	if (netif_msg_drv(&hdev->vport->nic))
 		hclge_info_show(hdev);
 
 	return ret;
+
+init_nic_err:
+	clear_bit(HCLGE_STATE_NIC_REGISTERED, &hdev->state);
+	while (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state))
+		msleep(HCLGE_WAIT_RESET_DONE);
+
+	client->ops->uninit_instance(&vport->nic, 0);
+
+	return ret;
 }
 
 static int hclge_init_roce_client_instance(struct hnae3_ae_dev *ae_dev,
@@ -8239,6 +8327,7 @@ static int hclge_init_roce_client_instance(struct hnae3_ae_dev *ae_dev,
 {
 	struct hnae3_client *client = vport->roce.client;
 	struct hclge_dev *hdev = ae_dev->priv;
+	int rst_cnt;
 	int ret;
 
 	if (!hnae3_dev_roce_supported(hdev) || !hdev->roce_client ||
@@ -8250,14 +8339,38 @@ static int hclge_init_roce_client_instance(struct hnae3_ae_dev *ae_dev,
 	if (ret)
 		return ret;
 
+	rst_cnt = hdev->rst_stats.reset_cnt;
 	ret = client->ops->init_instance(&vport->roce);
 	if (ret)
 		return ret;
 
 	set_bit(HCLGE_STATE_ROCE_REGISTERED, &hdev->state);
+	if (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state) ||
+	    rst_cnt != hdev->rst_stats.reset_cnt) {
+		ret = -EBUSY;
+		goto init_roce_err;
+	}
+
+	/* Enable roce ras interrupts */
+	ret = hclge_config_rocee_ras_interrupt(hdev, true);
+	if (ret) {
+		dev_err(&ae_dev->pdev->dev,
+			"fail(%d) to enable roce ras interrupts\n", ret);
+		goto init_roce_err;
+	}
+
 	hnae3_set_client_init_flag(client, ae_dev, 1);
 
 	return 0;
+
+init_roce_err:
+	clear_bit(HCLGE_STATE_ROCE_REGISTERED, &hdev->state);
+	while (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state))
+		msleep(HCLGE_WAIT_RESET_DONE);
+
+	hdev->roce_client->ops->uninit_instance(&vport->roce, 0);
+
+	return ret;
 }
 
 static int hclge_init_client_instance(struct hnae3_client *client,
@@ -8300,12 +8413,6 @@ static int hclge_init_client_instance(struct hnae3_client *client,
 		}
 	}
 
-	/* Enable roce ras interrupts */
-	ret = hclge_config_rocee_ras_interrupt(hdev, true);
-	if (ret)
-		dev_err(&ae_dev->pdev->dev,
-			"fail(%d) to enable roce ras interrupts\n", ret);
-
 	return ret;
 
 clear_nic:
@@ -8329,6 +8436,9 @@ static void hclge_uninit_client_instance(struct hnae3_client *client,
 		vport = &hdev->vport[i];
 		if (hdev->roce_client) {
 			clear_bit(HCLGE_STATE_ROCE_REGISTERED, &hdev->state);
+			while (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state))
+				msleep(HCLGE_WAIT_RESET_DONE);
+
 			hdev->roce_client->ops->uninit_instance(&vport->roce,
 								0);
 			hdev->roce_client = NULL;
@@ -8338,6 +8448,9 @@ static void hclge_uninit_client_instance(struct hnae3_client *client,
 			return;
 		if (hdev->nic_client && client->ops->uninit_instance) {
 			clear_bit(HCLGE_STATE_NIC_REGISTERED, &hdev->state);
+			while (test_bit(HCLGE_STATE_RST_HANDLING, &hdev->state))
+				msleep(HCLGE_WAIT_RESET_DONE);
+
 			client->ops->uninit_instance(&vport->nic, 0);
 			hdev->nic_client = NULL;
 			vport->nic.client = NULL;
@@ -9265,6 +9378,7 @@ static const struct hnae3_ae_ops hclge_ops = {
 	.set_autoneg = hclge_set_autoneg,
 	.get_autoneg = hclge_get_autoneg,
 	.restart_autoneg = hclge_restart_autoneg,
+	.halt_autoneg = hclge_halt_autoneg,
 	.get_pauseparam = hclge_get_pauseparam,
 	.set_pauseparam = hclge_set_pauseparam,
 	.set_mtu = hclge_set_mtu,

drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h

@@ -700,6 +700,7 @@ struct hclge_mac_tnl_stats {
 };
 
 #define HCLGE_RESET_INTERVAL	(10 * HZ)
+#define HCLGE_WAIT_RESET_DONE	100
 
 #pragma pack(1)
 struct hclge_vf_vlan_cfg {
@@ -929,6 +930,7 @@ struct hclge_vport {
 	u32 bw_limit;		/* VSI BW Limit (0 = disabled) */
 	u8  dwrr;
 
+	unsigned long vlan_del_fail_bmap[BITS_TO_LONGS(VLAN_N_VID)];
 	struct hclge_port_base_vlan_config port_base_vlan_cfg;
 	struct hclge_tx_vtag_cfg  txvlan_cfg;
 	struct hclge_rx_vtag_cfg  rxvlan_cfg;

drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c

@@ -1244,6 +1244,7 @@ static int hclgevf_set_vlan_filter(struct hnae3_handle *handle,
 #define HCLGEVF_VLAN_MBX_MSG_LEN 5
 	struct hclgevf_dev *hdev = hclgevf_ae_get_hdev(handle);
 	u8 msg_data[HCLGEVF_VLAN_MBX_MSG_LEN];
+	int ret;
 
 	if (vlan_id > HCLGEVF_MAX_VLAN_ID)
 		return -EINVAL;
@@ -1251,12 +1252,53 @@ static int hclgevf_set_vlan_filter(struct hnae3_handle *handle,
 	if (proto != htons(ETH_P_8021Q))
 		return -EPROTONOSUPPORT;
 
+	/* When device is resetting, firmware is unable to handle
+	 * mailbox. Just record the vlan id, and remove it after
+	 * reset finished.
+	 */
+	if (test_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state) && is_kill) {
+		set_bit(vlan_id, hdev->vlan_del_fail_bmap);
+		return -EBUSY;
+	}
+
 	msg_data[0] = is_kill;
 	memcpy(&msg_data[1], &vlan_id, sizeof(vlan_id));
 	memcpy(&msg_data[3], &proto, sizeof(proto));
-	return hclgevf_send_mbx_msg(hdev, HCLGE_MBX_SET_VLAN,
-				    HCLGE_MBX_VLAN_FILTER, msg_data,
-				    HCLGEVF_VLAN_MBX_MSG_LEN, false, NULL, 0);
+	ret = hclgevf_send_mbx_msg(hdev, HCLGE_MBX_SET_VLAN,
+				   HCLGE_MBX_VLAN_FILTER, msg_data,
+				   HCLGEVF_VLAN_MBX_MSG_LEN, false, 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.
+	 */
+	if (is_kill && ret)
+		set_bit(vlan_id, hdev->vlan_del_fail_bmap);
+
+	return ret;
+}
+
+static void hclgevf_sync_vlan_filter(struct hclgevf_dev *hdev)
+{
+#define HCLGEVF_MAX_SYNC_COUNT	60
+	struct hnae3_handle *handle = &hdev->nic;
+	int ret, sync_cnt = 0;
+	u16 vlan_id;
+
+	vlan_id = find_first_bit(hdev->vlan_del_fail_bmap, VLAN_N_VID);
+	while (vlan_id != VLAN_N_VID) {
+		ret = hclgevf_set_vlan_filter(handle, htons(ETH_P_8021Q),
+					      vlan_id, true);
+		if (ret)
+			return;
+
+		clear_bit(vlan_id, hdev->vlan_del_fail_bmap);
+		sync_cnt++;
+		if (sync_cnt >= HCLGEVF_MAX_SYNC_COUNT)
+			return;
+
+		vlan_id = find_first_bit(hdev->vlan_del_fail_bmap, VLAN_N_VID);
+	}
 }
 
 static int hclgevf_en_hw_strip_rxvtag(struct hnae3_handle *handle, bool enable)
@@ -1797,6 +1839,8 @@ static void hclgevf_service_task(struct work_struct *work)
 
 	hclgevf_update_link_mode(hdev);
 
+	hclgevf_sync_vlan_filter(hdev);
+
 	hclgevf_deferred_task_schedule(hdev);
 
 	clear_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state);

drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h

@@ -4,6 +4,7 @@
 #ifndef __HCLGEVF_MAIN_H
 #define __HCLGEVF_MAIN_H
 #include <linux/fs.h>
+#include <linux/if_vlan.h>
 #include <linux/types.h>
 #include "hclge_mbx.h"
 #include "hclgevf_cmd.h"
@@ -270,6 +271,8 @@ struct hclgevf_dev {
 	u16 *vector_status;
 	int *vector_irq;
 
+	unsigned long vlan_del_fail_bmap[BITS_TO_LONGS(VLAN_N_VID)];
+
 	bool mbx_event_pending;
 	struct hclgevf_mbx_resp_status mbx_resp; /* mailbox response */
 	struct hclgevf_mbx_arq_ring arq; /* mailbox async rx queue */