ice: Support link events, reset and rebuild

Link events are posted to a PF's admin receive queue (ARQ). This patch
adds the ability to detect and process link events.

This patch also adds the ability to process resets.

The driver can process the following resets:
    1) EMP Reset (EMPR)
    2) Global Reset (GLOBR)
    3) Core Reset (CORER)
    4) Physical Function Reset (PFR)

EMPR is the largest level of reset that the driver can handle. An EMPR
resets the manageability block and also the data path, including PHY and
link for all the PFs. The affected PFs are notified of this event through
a miscellaneous interrupt.

GLOBR is a subset of EMPR. It does everything EMPR does except that it
doesn't reset the manageability block.

CORER is a subset of GLOBR. It does everything GLOBR does but doesn't
reset PHY and link.

PFR is a subset of CORER and affects only the given physical function.
In other words, PFR can be thought of as a CORER for a single PF. Since
only the issuing PF is affected, a PFR doesn't result in the miscellaneous
interrupt being triggered.

All the resets have the following in common:
1) Tx/Rx is halted and all queues are stopped.
2) All the VSIs and filters programmed for the PF are lost and have to be
   reprogrammed.
3) Control queue interfaces are reset and have to be reprogrammed.

In the rebuild flow, control queues are reinitialized, VSIs are reallocated
and filters are restored.
Signed-off-by: NAnirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Tested-by: NTony Brelinski <tonyx.brelinski@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

drivers/net/ethernet/intel/ice/ice.h

@@ -78,6 +78,11 @@ extern const char ice_drv_ver[];
 #define ICE_RX_DESC(R, i) (&(((union ice_32b_rx_flex_desc *)((R)->desc))[i]))
 #define ICE_TX_CTX_DESC(R, i) (&(((struct ice_tx_ctx_desc *)((R)->desc))[i]))
 
+/* Macro for each VSI in a PF */
+#define ice_for_each_vsi(pf, i) \
+	for ((i) = 0; (i) < (pf)->num_alloc_vsi; (i)++)
+
+/* Macros for each tx/rx ring in a VSI */
 #define ice_for_each_txq(vsi, i) \
 	for ((i) = 0; (i) < (vsi)->num_txq; (i)++)
 
@@ -109,7 +114,16 @@ struct ice_sw {
 
 enum ice_state {
 	__ICE_DOWN,
+	__ICE_NEEDS_RESTART,
+	__ICE_RESET_RECOVERY_PENDING,	/* set by driver when reset starts */
 	__ICE_PFR_REQ,			/* set by driver and peers */
+	__ICE_CORER_REQ,		/* set by driver and peers */
+	__ICE_GLOBR_REQ,		/* set by driver and peers */
+	__ICE_CORER_RECV,		/* set by OICR handler */
+	__ICE_GLOBR_RECV,		/* set by OICR handler */
+	__ICE_EMPR_RECV,		/* set by OICR handler */
+	__ICE_SUSPENDED,		/* set on module remove path */
+	__ICE_RESET_FAILED,		/* set by reset/rebuild */
 	__ICE_ADMINQ_EVENT_PENDING,
 	__ICE_CFG_BUSY,
 	__ICE_SERVICE_SCHED,
@@ -226,6 +240,11 @@ struct ice_pf {
 	u16 q_left_rx;		/* remaining num rx queues left unclaimed */
 	u16 next_vsi;		/* Next free slot in pf->vsi[] - 0-based! */
 	u16 num_alloc_vsi;
+	u16 corer_count;	/* Core reset count */
+	u16 globr_count;	/* Global reset count */
+	u16 empr_count;		/* EMP reset count */
+	u16 pfr_count;		/* PF reset count */
+
 	struct ice_hw_port_stats stats;
 	struct ice_hw_port_stats stats_prev;
 	struct ice_hw hw;

drivers/net/ethernet/intel/ice/ice_adminq_cmd.h

@@ -1009,6 +1009,23 @@ struct ice_aqc_get_link_status_data {
 	__le64 reserved4;
 };
 
+/* Set event mask command (direct 0x0613) */
+struct ice_aqc_set_event_mask {
+	u8	lport_num;
+	u8	reserved[7];
+	__le16	event_mask;
+#define ICE_AQ_LINK_EVENT_UPDOWN		BIT(1)
+#define ICE_AQ_LINK_EVENT_MEDIA_NA		BIT(2)
+#define ICE_AQ_LINK_EVENT_LINK_FAULT		BIT(3)
+#define ICE_AQ_LINK_EVENT_PHY_TEMP_ALARM	BIT(4)
+#define ICE_AQ_LINK_EVENT_EXCESSIVE_ERRORS	BIT(5)
+#define ICE_AQ_LINK_EVENT_SIGNAL_DETECT		BIT(6)
+#define ICE_AQ_LINK_EVENT_AN_COMPLETED		BIT(7)
+#define ICE_AQ_LINK_EVENT_MODULE_QUAL_FAIL	BIT(8)
+#define ICE_AQ_LINK_EVENT_PORT_TX_SUSPENDED	BIT(9)
+	u8	reserved1[6];
+};
+
 /* NVM Read command (indirect 0x0701)
  * NVM Erase commands (direct 0x0702)
  * NVM Update commands (indirect 0x0703)
@@ -1215,6 +1232,7 @@ struct ice_aq_desc {
 		struct ice_aqc_dis_txqs dis_txqs;
 		struct ice_aqc_add_get_update_free_vsi vsi_cmd;
 		struct ice_aqc_alloc_free_res_cmd sw_res_ctrl;
+		struct ice_aqc_set_event_mask set_event_mask;
 		struct ice_aqc_get_link_status get_link_status;
 	} params;
 };
@@ -1294,6 +1312,7 @@ enum ice_adminq_opc {
 	ice_aqc_opc_set_phy_cfg				= 0x0601,
 	ice_aqc_opc_restart_an				= 0x0605,
 	ice_aqc_opc_get_link_status			= 0x0607,
+	ice_aqc_opc_set_event_mask			= 0x0613,
 
 	/* NVM commands */
 	ice_aqc_opc_nvm_read				= 0x0701,

drivers/net/ethernet/intel/ice/ice_common.c

@@ -1427,6 +1427,39 @@ ice_set_fc(struct ice_port_info *pi, u8 *aq_failures, bool atomic_restart)
 	return status;
 }
 
+/**
+ * ice_get_link_status - get status of the HW network link
+ * @pi: port information structure
+ * @link_up: pointer to bool (true/false = linkup/linkdown)
+ *
+ * Variable link_up is true if link is up, false if link is down.
+ * The variable link_up is invalid if status is non zero. As a
+ * result of this call, link status reporting becomes enabled
+ */
+enum ice_status ice_get_link_status(struct ice_port_info *pi, bool *link_up)
+{
+	struct ice_phy_info *phy_info;
+	enum ice_status status = 0;
+
+	if (!pi)
+		return ICE_ERR_PARAM;
+
+	phy_info = &pi->phy;
+
+	if (phy_info->get_link_info) {
+		status = ice_update_link_info(pi);
+
+		if (status)
+			ice_debug(pi->hw, ICE_DBG_LINK,
+				  "get link status error, status = %d\n",
+				  status);
+	}
+
+	*link_up = phy_info->link_info.link_info & ICE_AQ_LINK_UP;
+
+	return status;
+}
+
 /**
  * ice_aq_set_link_restart_an
  * @pi: pointer to the port information structure
@@ -1456,6 +1489,33 @@ ice_aq_set_link_restart_an(struct ice_port_info *pi, bool ena_link,
 	return ice_aq_send_cmd(pi->hw, &desc, NULL, 0, cd);
 }
 
+/**
+ * ice_aq_set_event_mask
+ * @hw: pointer to the hw struct
+ * @port_num: port number of the physical function
+ * @mask: event mask to be set
+ * @cd: pointer to command details structure or NULL
+ *
+ * Set event mask (0x0613)
+ */
+enum ice_status
+ice_aq_set_event_mask(struct ice_hw *hw, u8 port_num, u16 mask,
+		      struct ice_sq_cd *cd)
+{
+	struct ice_aqc_set_event_mask *cmd;
+	struct ice_aq_desc desc;
+
+	cmd = &desc.params.set_event_mask;
+
+	ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_set_event_mask);
+
+	cmd->lport_num = port_num;
+
+	cmd->event_mask = cpu_to_le16(mask);
+
+	return ice_aq_send_cmd(hw, &desc, NULL, 0, cd);
+}
+
 /**
  * __ice_aq_get_set_rss_lut
  * @hw: pointer to the hardware structure

drivers/net/ethernet/intel/ice/ice_common.h

@@ -20,6 +20,8 @@ enum ice_status
 ice_clean_rq_elem(struct ice_hw *hw, struct ice_ctl_q_info *cq,
 		  struct ice_rq_event_info *e, u16 *pending);
 enum ice_status
+ice_get_link_status(struct ice_port_info *pi, bool *link_up);
+enum ice_status
 ice_acquire_res(struct ice_hw *hw, enum ice_aq_res_ids res,
 		enum ice_aq_res_access_type access);
 void ice_release_res(struct ice_hw *hw, enum ice_aq_res_ids res);
@@ -66,6 +68,9 @@ enum ice_status
 ice_aq_get_link_info(struct ice_port_info *pi, bool ena_lse,
 		     struct ice_link_status *link, struct ice_sq_cd *cd);
 enum ice_status
+ice_aq_set_event_mask(struct ice_hw *hw, u8 port_num, u16 mask,
+		      struct ice_sq_cd *cd);
+enum ice_status
 ice_dis_vsi_txq(struct ice_port_info *pi, u8 num_queues, u16 *q_ids,
 		u32 *q_teids, struct ice_sq_cd *cmd_details);
 enum ice_status

drivers/net/ethernet/intel/ice/ice_hw_autogen.h

@@ -85,6 +85,8 @@
 #define GLGEN_RSTCTL			0x000B8180
 #define GLGEN_RSTCTL_GRSTDEL_S		0
 #define GLGEN_RSTCTL_GRSTDEL_M		ICE_M(0x3F, GLGEN_RSTCTL_GRSTDEL_S)
+#define GLGEN_RSTAT_RESET_TYPE_S	2
+#define GLGEN_RSTAT_RESET_TYPE_M	ICE_M(0x3, GLGEN_RSTAT_RESET_TYPE_S)
 #define GLGEN_RTRIG			0x000B8190
 #define GLGEN_RTRIG_CORER_S		0
 #define GLGEN_RTRIG_CORER_M		BIT(GLGEN_RTRIG_CORER_S)

drivers/net/ethernet/intel/ice/ice_main.c

@@ -29,6 +29,8 @@ MODULE_PARM_DESC(debug, "netif level (0=none,...,16=all)");
 static struct workqueue_struct *ice_wq;
 static const struct net_device_ops ice_netdev_ops;
 
+static void ice_pf_dis_all_vsi(struct ice_pf *pf);
+static void ice_rebuild(struct ice_pf *pf);
 static int ice_vsi_release(struct ice_vsi *vsi);
 static void ice_update_vsi_stats(struct ice_vsi *vsi);
 static void ice_update_pf_stats(struct ice_pf *pf);
@@ -216,6 +218,132 @@ static void ice_free_fltr_list(struct device *dev, struct list_head *h)
 	}
 }
 
+/**
+ * ice_is_reset_recovery_pending - schedule a reset
+ * @state: pf state field
+ */
+static bool ice_is_reset_recovery_pending(unsigned long int *state)
+{
+	return test_bit(__ICE_RESET_RECOVERY_PENDING, state);
+}
+
+/**
+ * ice_prepare_for_reset - prep for the core to reset
+ * @pf: board private structure
+ *
+ * Inform or close all dependent features in prep for reset.
+ */
+static void
+ice_prepare_for_reset(struct ice_pf *pf)
+{
+	struct ice_hw *hw = &pf->hw;
+	u32 v;
+
+	ice_for_each_vsi(pf, v)
+		if (pf->vsi[v])
+			ice_remove_vsi_fltr(hw, pf->vsi[v]->vsi_num);
+
+	dev_dbg(&pf->pdev->dev, "Tearing down internal switch for reset\n");
+
+	/* disable the VSIs and their queues that are not already DOWN */
+	/* pf_dis_all_vsi modifies netdev structures -rtnl_lock needed */
+	ice_pf_dis_all_vsi(pf);
+
+	ice_for_each_vsi(pf, v)
+		if (pf->vsi[v])
+			pf->vsi[v]->vsi_num = 0;
+
+	ice_shutdown_all_ctrlq(hw);
+}
+
+/**
+ * ice_do_reset - Initiate one of many types of resets
+ * @pf: board private structure
+ * @reset_type: reset type requested
+ * before this function was called.
+ */
+static void ice_do_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
+{
+	struct device *dev = &pf->pdev->dev;
+	struct ice_hw *hw = &pf->hw;
+
+	dev_dbg(dev, "reset_type 0x%x requested\n", reset_type);
+	WARN_ON(in_interrupt());
+
+	/* PFR is a bit of a special case because it doesn't result in an OICR
+	 * interrupt. So for PFR, we prepare for reset, issue the reset and
+	 * rebuild sequentially.
+	 */
+	if (reset_type == ICE_RESET_PFR) {
+		set_bit(__ICE_RESET_RECOVERY_PENDING, pf->state);
+		ice_prepare_for_reset(pf);
+	}
+
+	/* trigger the reset */
+	if (ice_reset(hw, reset_type)) {
+		dev_err(dev, "reset %d failed\n", reset_type);
+		set_bit(__ICE_RESET_FAILED, pf->state);
+		clear_bit(__ICE_RESET_RECOVERY_PENDING, pf->state);
+		return;
+	}
+
+	if (reset_type == ICE_RESET_PFR) {
+		pf->pfr_count++;
+		ice_rebuild(pf);
+		clear_bit(__ICE_RESET_RECOVERY_PENDING, pf->state);
+	}
+}
+
+/**
+ * ice_reset_subtask - Set up for resetting the device and driver
+ * @pf: board private structure
+ */
+static void ice_reset_subtask(struct ice_pf *pf)
+{
+	enum ice_reset_req reset_type;
+
+	rtnl_lock();
+
+	/* When a CORER/GLOBR/EMPR is about to happen, the hardware triggers an
+	 * OICR interrupt. The OICR handler (ice_misc_intr) determines what
+	 * type of reset happened and sets __ICE_RESET_RECOVERY_PENDING bit in
+	 * pf->state. So if reset/recovery is pending (as indicated by this bit)
+	 * we do a rebuild and return.
+	 */
+	if (ice_is_reset_recovery_pending(pf->state)) {
+		clear_bit(__ICE_GLOBR_RECV, pf->state);
+		clear_bit(__ICE_CORER_RECV, pf->state);
+		ice_prepare_for_reset(pf);
+
+		/* make sure we are ready to rebuild */
+		if (ice_check_reset(&pf->hw))
+			set_bit(__ICE_RESET_FAILED, pf->state);
+		else
+			ice_rebuild(pf);
+		clear_bit(__ICE_RESET_RECOVERY_PENDING, pf->state);
+		goto unlock;
+	}
+
+	/* No pending resets to finish processing. Check for new resets */
+	if (test_and_clear_bit(__ICE_GLOBR_REQ, pf->state))
+		reset_type = ICE_RESET_GLOBR;
+	else if (test_and_clear_bit(__ICE_CORER_REQ, pf->state))
+		reset_type = ICE_RESET_CORER;
+	else if (test_and_clear_bit(__ICE_PFR_REQ, pf->state))
+		reset_type = ICE_RESET_PFR;
+	else
+		goto unlock;
+
+	/* reset if not already down or resetting */
+	if (!test_bit(__ICE_DOWN, pf->state) &&
+	    !test_bit(__ICE_CFG_BUSY, pf->state)) {
+		ice_do_reset(pf, reset_type);
+	}
+
+unlock:
+	rtnl_unlock();
+}
+
 /**
  * ice_watchdog_subtask - periodic tasks not using event driven scheduling
  * @pf: board private structure
@@ -314,6 +442,144 @@ void ice_print_link_msg(struct ice_vsi *vsi, bool isup)
 		    speed, fc);
 }
 
+/**
+ * ice_init_link_events - enable/initialize link events
+ * @pi: pointer to the port_info instance
+ *
+ * Returns -EIO on failure, 0 on success
+ */
+static int ice_init_link_events(struct ice_port_info *pi)
+{
+	u16 mask;
+
+	mask = ~((u16)(ICE_AQ_LINK_EVENT_UPDOWN | ICE_AQ_LINK_EVENT_MEDIA_NA |
+		       ICE_AQ_LINK_EVENT_MODULE_QUAL_FAIL));
+
+	if (ice_aq_set_event_mask(pi->hw, pi->lport, mask, NULL)) {
+		dev_dbg(ice_hw_to_dev(pi->hw),
+			"Failed to set link event mask for port %d\n",
+			pi->lport);
+		return -EIO;
+	}
+
+	if (ice_aq_get_link_info(pi, true, NULL, NULL)) {
+		dev_dbg(ice_hw_to_dev(pi->hw),
+			"Failed to enable link events for port %d\n",
+			pi->lport);
+		return -EIO;
+	}
+
+	return 0;
+}
+
+/**
+ * ice_vsi_link_event - update the vsi's netdev
+ * @vsi: the vsi on which the link event occurred
+ * @link_up: whether or not the vsi needs to be set up or down
+ */
+static void ice_vsi_link_event(struct ice_vsi *vsi, bool link_up)
+{
+	if (!vsi || test_bit(__ICE_DOWN, vsi->state))
+		return;
+
+	if (vsi->type == ICE_VSI_PF) {
+		if (!vsi->netdev) {
+			dev_dbg(&vsi->back->pdev->dev,
+				"vsi->netdev is not initialized!\n");
+			return;
+		}
+		if (link_up) {
+			netif_carrier_on(vsi->netdev);
+			netif_tx_wake_all_queues(vsi->netdev);
+		} else {
+			netif_carrier_off(vsi->netdev);
+			netif_tx_stop_all_queues(vsi->netdev);
+		}
+	}
+}
+
+/**
+ * ice_link_event - process the link event
+ * @pf: pf that the link event is associated with
+ * @pi: port_info for the port that the link event is associated with
+ *
+ * Returns -EIO if ice_get_link_status() fails
+ * Returns 0 on success
+ */
+static int
+ice_link_event(struct ice_pf *pf, struct ice_port_info *pi)
+{
+	u8 new_link_speed, old_link_speed;
+	struct ice_phy_info *phy_info;
+	bool new_link_same_as_old;
+	bool new_link, old_link;
+	u8 lport;
+	u16 v;
+
+	phy_info = &pi->phy;
+	phy_info->link_info_old = phy_info->link_info;
+	/* Force ice_get_link_status() to update link info */
+	phy_info->get_link_info = true;
+
+	old_link = (phy_info->link_info_old.link_info & ICE_AQ_LINK_UP);
+	old_link_speed = phy_info->link_info_old.link_speed;
+
+	lport = pi->lport;
+	if (ice_get_link_status(pi, &new_link)) {
+		dev_dbg(&pf->pdev->dev,
+			"Could not get link status for port %d\n", lport);
+		return -EIO;
+	}
+
+	new_link_speed = phy_info->link_info.link_speed;
+
+	new_link_same_as_old = (new_link == old_link &&
+				new_link_speed == old_link_speed);
+
+	ice_for_each_vsi(pf, v) {
+		struct ice_vsi *vsi = pf->vsi[v];
+
+		if (!vsi || !vsi->port_info)
+			continue;
+
+		if (new_link_same_as_old &&
+		    (test_bit(__ICE_DOWN, vsi->state) ||
+		    new_link == netif_carrier_ok(vsi->netdev)))
+			continue;
+
+		if (vsi->port_info->lport == lport) {
+			ice_print_link_msg(vsi, new_link);
+			ice_vsi_link_event(vsi, new_link);
+		}
+	}
+
+	return 0;
+}
+
+/**
+ * ice_handle_link_event - handle link event via ARQ
+ * @pf: pf that the link event is associated with
+ *
+ * Return -EINVAL if port_info is null
+ * Return status on succes
+ */
+static int ice_handle_link_event(struct ice_pf *pf)
+{
+	struct ice_port_info *port_info;
+	int status;
+
+	port_info = pf->hw.port_info;
+	if (!port_info)
+		return -EINVAL;
+
+	status = ice_link_event(pf, port_info);
+	if (status)
+		dev_dbg(&pf->pdev->dev,
+			"Could not process link event, error %d\n", status);
+
+	return status;
+}
+
 /**
  * __ice_clean_ctrlq - helper function to clean controlq rings
  * @pf: ptr to struct ice_pf
@@ -328,6 +594,10 @@ static int __ice_clean_ctrlq(struct ice_pf *pf, enum ice_ctl_q q_type)
 	const char *qtype;
 	u32 oldval, val;
 
+	/* Do not clean control queue if/when PF reset fails */
+	if (test_bit(__ICE_RESET_FAILED, pf->state))
+		return 0;
+
 	switch (q_type) {
 	case ICE_CTL_Q_ADMIN:
 		cq = &hw->adminq;
@@ -394,6 +664,7 @@ static int __ice_clean_ctrlq(struct ice_pf *pf, enum ice_ctl_q q_type)
 
 	do {
 		enum ice_status ret;
+		u16 opcode;
 
 		ret = ice_clean_rq_elem(hw, cq, &event, &pending);
 		if (ret == ICE_ERR_AQ_NO_WORK)
@@ -404,6 +675,21 @@ static int __ice_clean_ctrlq(struct ice_pf *pf, enum ice_ctl_q q_type)
 				ret);
 			break;
 		}
+
+		opcode = le16_to_cpu(event.desc.opcode);
+
+		switch (opcode) {
+		case ice_aqc_opc_get_link_status:
+			if (ice_handle_link_event(pf))
+				dev_err(&pf->pdev->dev,
+					"Could not handle link event");
+			break;
+		default:
+			dev_dbg(&pf->pdev->dev,
+				"%s Receive Queue unknown event 0x%04x ignored\n",
+				qtype, opcode);
+			break;
+		}
 	} while (pending && (i++ < ICE_DFLT_IRQ_WORK));
 
 	devm_kfree(&pf->pdev->dev, event.msg_buf);
@@ -483,6 +769,17 @@ static void ice_service_task(struct work_struct *work)
 	unsigned long start_time = jiffies;
 
 	/* subtasks */
+
+	/* process reset requests first */
+	ice_reset_subtask(pf);
+
+	/* bail if a reset/recovery cycle is pending */
+	if (ice_is_reset_recovery_pending(pf->state) ||
+	    test_bit(__ICE_SUSPENDED, pf->state)) {
+		ice_service_task_complete(pf);
+		return;
+	}
+
 	ice_watchdog_subtask(pf);
 	ice_clean_adminq_subtask(pf);
 
@@ -1208,6 +1505,37 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
 	if (!(oicr & PFINT_OICR_INTEVENT_M))
 		goto ena_intr;
 
+	if (oicr & PFINT_OICR_GRST_M) {
+		u32 reset;
+		/* we have a reset warning */
+		ena_mask &= ~PFINT_OICR_GRST_M;
+		reset = (rd32(hw, GLGEN_RSTAT) & GLGEN_RSTAT_RESET_TYPE_M) >>
+			GLGEN_RSTAT_RESET_TYPE_S;
+
+		if (reset == ICE_RESET_CORER)
+			pf->corer_count++;
+		else if (reset == ICE_RESET_GLOBR)
+			pf->globr_count++;
+		else
+			pf->empr_count++;
+
+		/* If a reset cycle isn't already in progress, we set a bit in
+		 * pf->state so that the service task can start a reset/rebuild.
+		 * We also make note of which reset happened so that peer
+		 * devices/drivers can be informed.
+		 */
+		if (!test_bit(__ICE_RESET_RECOVERY_PENDING, pf->state)) {
+			if (reset == ICE_RESET_CORER)
+				set_bit(__ICE_CORER_RECV, pf->state);
+			else if (reset == ICE_RESET_GLOBR)
+				set_bit(__ICE_GLOBR_RECV, pf->state);
+			else
+				set_bit(__ICE_EMPR_RECV, pf->state);
+
+			set_bit(__ICE_RESET_RECOVERY_PENDING, pf->state);
+		}
+	}
+
 	if (oicr & PFINT_OICR_HMC_ERR_M) {
 		ena_mask &= ~PFINT_OICR_HMC_ERR_M;
 		dev_dbg(&pf->pdev->dev,
@@ -1226,9 +1554,10 @@ static irqreturn_t ice_misc_intr(int __always_unused irq, void *data)
 		 */
 		if (oicr & (PFINT_OICR_PE_CRITERR_M |
 			    PFINT_OICR_PCI_EXCEPTION_M |
-			    PFINT_OICR_ECC_ERR_M))
+			    PFINT_OICR_ECC_ERR_M)) {
 			set_bit(__ICE_PFR_REQ, pf->state);
-
+			ice_service_task_schedule(pf);
+		}
 		ena_mask &= ~oicr;
 	}
 	ret = IRQ_HANDLED;
@@ -1485,6 +1814,13 @@ static int ice_req_irq_msix_misc(struct ice_pf *pf)
 			 dev_driver_string(&pf->pdev->dev),
 			 dev_name(&pf->pdev->dev));
 
+	/* Do not request IRQ but do enable OICR interrupt since settings are
+	 * lost during reset. Note that this function is called only during
+	 * rebuild path and not while reset is in progress.
+	 */
+	if (ice_is_reset_recovery_pending(pf->state))
+		goto skip_req_irq;
+
 	/* reserve one vector in irq_tracker for misc interrupts */
 	oicr_idx = ice_get_res(pf, pf->irq_tracker, 1, ICE_RES_MISC_VEC_ID);
 	if (oicr_idx < 0)
@@ -1503,6 +1839,7 @@ static int ice_req_irq_msix_misc(struct ice_pf *pf)
 		return err;
 	}
 
+skip_req_irq:
 	ice_ena_misc_vector(pf);
 
 	val = (pf->oicr_idx & PFINT_OICR_CTL_MSIX_INDX_M) |
@@ -2070,6 +2407,100 @@ static int ice_vsi_cfg_rss(struct ice_vsi *vsi)
 	return err;
 }
 
+/**
+ * ice_vsi_reinit_setup - return resource and reallocate resource for a VSI
+ * @vsi: pointer to the ice_vsi
+ *
+ * This reallocates the VSIs queue resources
+ *
+ * Returns 0 on success and negative value on failure
+ */
+static int ice_vsi_reinit_setup(struct ice_vsi *vsi)
+{
+	u16 max_txqs[ICE_MAX_TRAFFIC_CLASS] = { 0 };
+	int ret, i;
+
+	if (!vsi)
+		return -EINVAL;
+
+	ice_vsi_free_q_vectors(vsi);
+	ice_free_res(vsi->back->irq_tracker, vsi->base_vector, vsi->idx);
+	vsi->base_vector = 0;
+	ice_vsi_clear_rings(vsi);
+	ice_vsi_free_arrays(vsi, false);
+	ice_vsi_set_num_qs(vsi);
+
+	/* Initialize VSI struct elements and create VSI in FW */
+	ret = ice_vsi_add(vsi);
+	if (ret < 0)
+		goto err_vsi;
+
+	ret = ice_vsi_alloc_arrays(vsi, false);
+	if (ret < 0)
+		goto err_vsi;
+
+	switch (vsi->type) {
+	case ICE_VSI_PF:
+		if (!vsi->netdev) {
+			ret = ice_cfg_netdev(vsi);
+			if (ret)
+				goto err_rings;
+
+			ret = register_netdev(vsi->netdev);
+			if (ret)
+				goto err_rings;
+
+			netif_carrier_off(vsi->netdev);
+			netif_tx_stop_all_queues(vsi->netdev);
+		}
+
+		ret = ice_vsi_alloc_q_vectors(vsi);
+		if (ret)
+			goto err_rings;
+
+		ret = ice_vsi_setup_vector_base(vsi);
+		if (ret)
+			goto err_vectors;
+
+		ret = ice_vsi_alloc_rings(vsi);
+		if (ret)
+			goto err_vectors;
+
+		ice_vsi_map_rings_to_vectors(vsi);
+		break;
+	default:
+		break;
+	}
+
+	ice_vsi_set_tc_cfg(vsi);
+
+	/* configure VSI nodes based on number of queues and TC's */
+	for (i = 0; i < vsi->tc_cfg.numtc; i++)
+		max_txqs[i] = vsi->num_txq;
+
+	ret = ice_cfg_vsi_lan(vsi->port_info, vsi->vsi_num,
+			      vsi->tc_cfg.ena_tc, max_txqs);
+	if (ret) {
+		dev_info(&vsi->back->pdev->dev,
+			 "Failed VSI lan queue config\n");
+		goto err_vectors;
+	}
+	return 0;
+
+err_vectors:
+	ice_vsi_free_q_vectors(vsi);
+err_rings:
+	if (vsi->netdev) {
+		unregister_netdev(vsi->netdev);
+		free_netdev(vsi->netdev);
+		vsi->netdev = NULL;
+	}
+err_vsi:
+	ice_vsi_clear(vsi);
+	set_bit(__ICE_RESET_FAILED, vsi->back->state);
+	return ret;
+}
+
 /**
  * ice_vsi_setup - Set up a VSI by a given type
  * @pf: board private structure
@@ -2345,10 +2776,17 @@ static int ice_setup_pf_sw(struct ice_pf *pf)
 	struct ice_vsi *vsi;
 	int status = 0;
 
-	vsi = ice_vsi_setup(pf, ICE_VSI_PF, pf->hw.port_info);
-	if (!vsi) {
-		status = -ENOMEM;
-		goto error_exit;
+	if (!ice_is_reset_recovery_pending(pf->state)) {
+		vsi = ice_vsi_setup(pf, ICE_VSI_PF, pf->hw.port_info);
+		if (!vsi) {
+			status = -ENOMEM;
+			goto error_exit;
+		}
+	} else {
+		vsi = pf->vsi[0];
+		status = ice_vsi_reinit_setup(vsi);
+		if (status < 0)
+			return -EIO;
 	}
 
 	/* tmp_add_list contains a list of MAC addresses for which MAC
@@ -2737,6 +3175,12 @@ static int ice_probe(struct pci_dev *pdev,
 	/* since everything is good, start the service timer */
 	mod_timer(&pf->serv_tmr, round_jiffies(jiffies + pf->serv_tmr_period));
 
+	err = ice_init_link_events(pf->hw.port_info);
+	if (err) {
+		dev_err(&pdev->dev, "ice_init_link_events failed: %d\n", err);
+		goto err_alloc_sw_unroll;
+	}
+
 	return 0;
 
 err_alloc_sw_unroll:
@@ -4217,6 +4661,131 @@ static int ice_vsi_release(struct ice_vsi *vsi)
 	return 0;
 }
 
+/**
+ * ice_dis_vsi - pause a VSI
+ * @vsi: the VSI being paused
+ */
+static void ice_dis_vsi(struct ice_vsi *vsi)
+{
+	if (test_bit(__ICE_DOWN, vsi->state))
+		return;
+
+	set_bit(__ICE_NEEDS_RESTART, vsi->state);
+
+	if (vsi->netdev && netif_running(vsi->netdev) &&
+	    vsi->type == ICE_VSI_PF)
+		vsi->netdev->netdev_ops->ndo_stop(vsi->netdev);
+
+	ice_vsi_close(vsi);
+}
+
+/**
+ * ice_ena_vsi - resume a VSI
+ * @vsi: the VSI being resume
+ */
+static void ice_ena_vsi(struct ice_vsi *vsi)
+{
+	if (!test_and_clear_bit(__ICE_NEEDS_RESTART, vsi->state))
+		return;
+
+	if (vsi->netdev && netif_running(vsi->netdev))
+		vsi->netdev->netdev_ops->ndo_open(vsi->netdev);
+	else if (ice_vsi_open(vsi))
+		/* this clears the DOWN bit */
+		dev_dbg(&vsi->back->pdev->dev, "Failed open VSI 0x%04X on switch 0x%04X\n",
+			vsi->vsi_num, vsi->vsw->sw_id);
+}
+
+/**
+ * ice_pf_dis_all_vsi - Pause all VSIs on a PF
+ * @pf: the PF
+ */
+static void ice_pf_dis_all_vsi(struct ice_pf *pf)
+{
+	int v;
+
+	ice_for_each_vsi(pf, v)
+		if (pf->vsi[v])
+			ice_dis_vsi(pf->vsi[v]);
+}
+
+/**
+ * ice_pf_ena_all_vsi - Resume all VSIs on a PF
+ * @pf: the PF
+ */
+static void ice_pf_ena_all_vsi(struct ice_pf *pf)
+{
+	int v;
+
+	ice_for_each_vsi(pf, v)
+		if (pf->vsi[v])
+			ice_ena_vsi(pf->vsi[v]);
+}
+
+/**
+ * ice_rebuild - rebuild after reset
+ * @pf: pf to rebuild
+ */
+static void ice_rebuild(struct ice_pf *pf)
+{
+	struct device *dev = &pf->pdev->dev;
+	struct ice_hw *hw = &pf->hw;
+	enum ice_status ret;
+	int err;
+
+	if (test_bit(__ICE_DOWN, pf->state))
+		goto clear_recovery;
+
+	dev_dbg(dev, "rebuilding pf\n");
+
+	ret = ice_init_all_ctrlq(hw);
+	if (ret) {
+		dev_err(dev, "control queues init failed %d\n", ret);
+		goto fail_reset;
+	}
+
+	ret = ice_clear_pf_cfg(hw);
+	if (ret) {
+		dev_err(dev, "clear PF configuration failed %d\n", ret);
+		goto fail_reset;
+	}
+
+	ice_clear_pxe_mode(hw);
+
+	ret = ice_get_caps(hw);
+	if (ret) {
+		dev_err(dev, "ice_get_caps failed %d\n", ret);
+		goto fail_reset;
+	}
+
+	/* basic nic switch setup */
+	err = ice_setup_pf_sw(pf);
+	if (err) {
+		dev_err(dev, "ice_setup_pf_sw failed\n");
+		goto fail_reset;
+	}
+
+	/* start misc vector */
+	if (test_bit(ICE_FLAG_MSIX_ENA, pf->flags)) {
+		err = ice_req_irq_msix_misc(pf);
+		if (err) {
+			dev_err(dev, "misc vector setup failed: %d\n", err);
+			goto fail_reset;
+		}
+	}
+
+	/* restart the VSIs that were rebuilt and running before the reset */
+	ice_pf_ena_all_vsi(pf);
+
+	return;
+
+fail_reset:
+	ice_shutdown_all_ctrlq(hw);
+	set_bit(__ICE_RESET_FAILED, pf->state);
+clear_recovery:
+	set_bit(__ICE_RESET_RECOVERY_PENDING, pf->state);
+}
+
 /**
  * ice_set_rss - Set RSS keys and lut
  * @vsi: Pointer to VSI structure

drivers/net/ethernet/intel/ice/ice_type.h

@@ -17,6 +17,7 @@ static inline bool ice_is_tc_ena(u8 bitmap, u8 tc)
 
 /* debug masks - set these bits in hw->debug_mask to control output */
 #define ICE_DBG_INIT		BIT_ULL(1)
+#define ICE_DBG_LINK		BIT_ULL(4)
 #define ICE_DBG_QCTX		BIT_ULL(6)
 #define ICE_DBG_NVM		BIT_ULL(7)
 #define ICE_DBG_LAN		BIT_ULL(8)