1 files changed, 147 insertions, 57 deletions

diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index f02d44455772..c8286adae946 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -64,6 +64,7 @@ struct device *ice_hw_to_dev(struct ice_hw *hw)
 }
 
 static struct workqueue_struct *ice_wq;
+struct workqueue_struct *ice_lag_wq;
 static const struct net_device_ops ice_netdev_safe_mode_ops;
 static const struct net_device_ops ice_netdev_ops;
 
@@ -80,7 +81,7 @@ ice_indr_setup_tc_cb(struct net_device *netdev, struct Qdisc *sch,
 		     void *data,
 		     void (*cleanup)(struct flow_block_cb *block_cb));
 
-bool netif_is_ice(struct net_device *dev)
+bool netif_is_ice(const struct net_device *dev)
 {
 	return dev && (dev->netdev_ops == &ice_netdev_ops);
 }
@@ -635,6 +636,11 @@ static void ice_do_reset(struct ice_pf *pf, enum ice_reset_req reset_type)
 
 	dev_dbg(dev, "reset_type 0x%x requested\n", reset_type);
 
+	if (pf->lag && pf->lag->bonded && reset_type == ICE_RESET_PFR) {
+		dev_dbg(dev, "PFR on a bonded interface, promoting to CORER\n");
+		reset_type = ICE_RESET_CORER;
+	}
+
 	ice_prepare_for_reset(pf, reset_type);
 
 	/* trigger the reset */
@@ -718,8 +724,13 @@ static void ice_reset_subtask(struct ice_pf *pf)
 	}
 
 	/* No pending resets to finish processing. Check for new resets */
-	if (test_bit(ICE_PFR_REQ, pf->state))
+	if (test_bit(ICE_PFR_REQ, pf->state)) {
 		reset_type = ICE_RESET_PFR;
+		if (pf->lag && pf->lag->bonded) {
+			dev_dbg(ice_pf_to_dev(pf), "PFR on a bonded interface, promoting to CORER\n");
+			reset_type = ICE_RESET_CORER;
+		}
+	}
 	if (test_bit(ICE_CORER_REQ, pf->state))
 		reset_type = ICE_RESET_CORER;
 	if (test_bit(ICE_GLOBR_REQ, pf->state))
@@ -1239,64 +1250,63 @@ ice_handle_link_event(struct ice_pf *pf, struct ice_rq_event_info *event)
 	return status;
 }
 
-enum ice_aq_task_state {
-	ICE_AQ_TASK_WAITING = 0,
-	ICE_AQ_TASK_COMPLETE,
-	ICE_AQ_TASK_CANCELED,
-};
-
-struct ice_aq_task {
-	struct hlist_node entry;
+/**
+ * ice_aq_prep_for_event - Prepare to wait for an AdminQ event from firmware
+ * @pf: pointer to the PF private structure
+ * @task: intermediate helper storage and identifier for waiting
+ * @opcode: the opcode to wait for
+ *
+ * Prepares to wait for a specific AdminQ completion event on the ARQ for
+ * a given PF. Actual wait would be done by a call to ice_aq_wait_for_event().
+ *
+ * Calls are separated to allow caller registering for event before sending
+ * the command, which mitigates a race between registering and FW responding.
+ *
+ * To obtain only the descriptor contents, pass an task->event with null
+ * msg_buf. If the complete data buffer is desired, allocate the
+ * task->event.msg_buf with enough space ahead of time.
+ */
+void ice_aq_prep_for_event(struct ice_pf *pf, struct ice_aq_task *task,
+			   u16 opcode)
+{
+	INIT_HLIST_NODE(&task->entry);
+	task->opcode = opcode;
+	task->state = ICE_AQ_TASK_WAITING;
 
-	u16 opcode;
-	struct ice_rq_event_info *event;
-	enum ice_aq_task_state state;
-};
+	spin_lock_bh(&pf->aq_wait_lock);
+	hlist_add_head(&task->entry, &pf->aq_wait_list);
+	spin_unlock_bh(&pf->aq_wait_lock);
+}
 
 /**
  * ice_aq_wait_for_event - Wait for an AdminQ event from firmware
  * @pf: pointer to the PF private structure
- * @opcode: the opcode to wait for
+ * @task: ptr prepared by ice_aq_prep_for_event()
  * @timeout: how long to wait, in jiffies
- * @event: storage for the event info
  *
  * Waits for a specific AdminQ completion event on the ARQ for a given PF. The
  * current thread will be put to sleep until the specified event occurs or
  * until the given timeout is reached.
  *
- * To obtain only the descriptor contents, pass an event without an allocated
- * msg_buf. If the complete data buffer is desired, allocate the
- * event->msg_buf with enough space ahead of time.
- *
  * Returns: zero on success, or a negative error code on failure.
  */
-int ice_aq_wait_for_event(struct ice_pf *pf, u16 opcode, unsigned long timeout,
-			  struct ice_rq_event_info *event)
+int ice_aq_wait_for_event(struct ice_pf *pf, struct ice_aq_task *task,
+			  unsigned long timeout)
 {
+	enum ice_aq_task_state *state = &task->state;
 	struct device *dev = ice_pf_to_dev(pf);
-	struct ice_aq_task *task;
-	unsigned long start;
+	unsigned long start = jiffies;
 	long ret;
 	int err;
 
-	task = kzalloc(sizeof(*task), GFP_KERNEL);
-	if (!task)
-		return -ENOMEM;
-
-	INIT_HLIST_NODE(&task->entry);
-	task->opcode = opcode;
-	task->event = event;
-	task->state = ICE_AQ_TASK_WAITING;
-
-	spin_lock_bh(&pf->aq_wait_lock);
-	hlist_add_head(&task->entry, &pf->aq_wait_list);
-	spin_unlock_bh(&pf->aq_wait_lock);
-
-	start = jiffies;
-
-	ret = wait_event_interruptible_timeout(pf->aq_wait_queue, task->state,
+	ret = wait_event_interruptible_timeout(pf->aq_wait_queue,
+					       *state != ICE_AQ_TASK_WAITING,
 					       timeout);
-	switch (task->state) {
+	switch (*state) {
+	case ICE_AQ_TASK_NOT_PREPARED:
+		WARN(1, "call to %s without ice_aq_prep_for_event()", __func__);
+		err = -EINVAL;
+		break;
 	case ICE_AQ_TASK_WAITING:
 		err = ret < 0 ? ret : -ETIMEDOUT;
 		break;
@@ -1307,7 +1317,7 @@ int ice_aq_wait_for_event(struct ice_pf *pf, u16 opcode, unsigned long timeout,
 		err = ret < 0 ? ret : 0;
 		break;
 	default:
-		WARN(1, "Unexpected AdminQ wait task state %u", task->state);
+		WARN(1, "Unexpected AdminQ wait task state %u", *state);
 		err = -EINVAL;
 		break;
 	}
@@ -1315,12 +1325,11 @@ int ice_aq_wait_for_event(struct ice_pf *pf, u16 opcode, unsigned long timeout,
 	dev_dbg(dev, "Waited %u msecs (max %u msecs) for firmware response to op 0x%04x\n",
 		jiffies_to_msecs(jiffies - start),
 		jiffies_to_msecs(timeout),
-		opcode);
+		task->opcode);
 
 	spin_lock_bh(&pf->aq_wait_lock);
 	hlist_del(&task->entry);
 	spin_unlock_bh(&pf->aq_wait_lock);
-	kfree(task);
 
 	return err;
 }
@@ -1346,23 +1355,26 @@ int ice_aq_wait_for_event(struct ice_pf *pf, u16 opcode, unsigned long timeout,
 static void ice_aq_check_events(struct ice_pf *pf, u16 opcode,
 				struct ice_rq_event_info *event)
 {
+	struct ice_rq_event_info *task_ev;
 	struct ice_aq_task *task;
 	bool found = false;
 
 	spin_lock_bh(&pf->aq_wait_lock);
 	hlist_for_each_entry(task, &pf->aq_wait_list, entry) {
-		if (task->state || task->opcode != opcode)
+		if (task->state != ICE_AQ_TASK_WAITING)
+			continue;
+		if (task->opcode != opcode)
 			continue;
 
-		memcpy(&task->event->desc, &event->desc, sizeof(event->desc));
-		task->event->msg_len = event->msg_len;
+		task_ev = &task->event;
+		memcpy(&task_ev->desc, &event->desc, sizeof(event->desc));
+		task_ev->msg_len = event->msg_len;
 
 		/* Only copy the data buffer if a destination was set */
-		if (task->event->msg_buf &&
-		    task->event->buf_len > event->buf_len) {
-			memcpy(task->event->msg_buf, event->msg_buf,
+		if (task_ev->msg_buf && task_ev->buf_len >= event->buf_len) {
+			memcpy(task_ev->msg_buf, event->msg_buf,
 			       event->buf_len);
-			task->event->buf_len = event->buf_len;
+			task_ev->buf_len = event->buf_len;
 		}
 
 		task->state = ICE_AQ_TASK_COMPLETE;
@@ -3392,6 +3404,7 @@ static void ice_set_ops(struct ice_vsi *vsi)
 	netdev->xdp_features = NETDEV_XDP_ACT_BASIC | NETDEV_XDP_ACT_REDIRECT |
 			       NETDEV_XDP_ACT_XSK_ZEROCOPY |
 			       NETDEV_XDP_ACT_RX_SG;
+	netdev->xdp_zc_max_segs = ICE_MAX_BUF_TXD;
 }
 
 /**
@@ -3794,6 +3807,7 @@ u16 ice_get_avail_rxq_count(struct ice_pf *pf)
 static void ice_deinit_pf(struct ice_pf *pf)
 {
 	ice_service_task_stop(pf);
+	mutex_destroy(&pf->lag_mutex);
 	mutex_destroy(&pf->adev_mutex);
 	mutex_destroy(&pf->sw_mutex);
 	mutex_destroy(&pf->tc_mutex);
@@ -3874,6 +3888,7 @@ static int ice_init_pf(struct ice_pf *pf)
 	mutex_init(&pf->sw_mutex);
 	mutex_init(&pf->tc_mutex);
 	mutex_init(&pf->adev_mutex);
+	mutex_init(&pf->lag_mutex);
 
 	INIT_HLIST_HEAD(&pf->aq_wait_list);
 	spin_lock_init(&pf->aq_wait_lock);
@@ -4506,6 +4521,31 @@ static void ice_deinit_eth(struct ice_pf *pf)
 	ice_decfg_netdev(vsi);
 }
 
+/**
+ * ice_wait_for_fw - wait for full FW readiness
+ * @hw: pointer to the hardware structure
+ * @timeout: milliseconds that can elapse before timing out
+ */
+static int ice_wait_for_fw(struct ice_hw *hw, u32 timeout)
+{
+	int fw_loading;
+	u32 elapsed = 0;
+
+	while (elapsed <= timeout) {
+		fw_loading = rd32(hw, GL_MNG_FWSM) & GL_MNG_FWSM_FW_LOADING_M;
+
+		/* firmware was not yet loaded, we have to wait more */
+		if (fw_loading) {
+			elapsed += 100;
+			msleep(100);
+			continue;
+		}
+		return 0;
+	}
+
+	return -ETIMEDOUT;
+}
+
 static int ice_init_dev(struct ice_pf *pf)
 {
 	struct device *dev = ice_pf_to_dev(pf);
@@ -4518,6 +4558,18 @@ static int ice_init_dev(struct ice_pf *pf)
 		return err;
 	}
 
+	/* Some cards require longer initialization times
+	 * due to necessity of loading FW from an external source.
+	 * This can take even half a minute.
+	 */
+	if (ice_is_pf_c827(hw)) {
+		err = ice_wait_for_fw(hw, 30000);
+		if (err) {
+			dev_err(dev, "ice_wait_for_fw timed out");
+			return err;
+		}
+	}
+
 	ice_init_feature_support(pf);
 
 	ice_request_fw(pf);
@@ -5570,7 +5622,7 @@ static struct pci_driver ice_driver = {
  */
 static int __init ice_module_init(void)
 {
-	int status;
+	int status = -ENOMEM;
 
 	pr_info("%s\n", ice_driver_string);
 	pr_info("%s\n", ice_copyright);
@@ -5578,15 +5630,27 @@ static int __init ice_module_init(void)
 	ice_wq = alloc_workqueue("%s", 0, 0, KBUILD_MODNAME);
 	if (!ice_wq) {
 		pr_err("Failed to create workqueue\n");
-		return -ENOMEM;
+		return status;
+	}
+
+	ice_lag_wq = alloc_ordered_workqueue("ice_lag_wq", 0);
+	if (!ice_lag_wq) {
+		pr_err("Failed to create LAG workqueue\n");
+		goto err_dest_wq;
 	}
 
 	status = pci_register_driver(&ice_driver);
 	if (status) {
 		pr_err("failed to register PCI driver, err %d\n", status);
-		destroy_workqueue(ice_wq);
+		goto err_dest_lag_wq;
 	}
 
+	return 0;
+
+err_dest_lag_wq:
+	destroy_workqueue(ice_lag_wq);
+err_dest_wq:
+	destroy_workqueue(ice_wq);
 	return status;
 }
 module_init(ice_module_init);
@@ -5601,6 +5665,7 @@ static void __exit ice_module_exit(void)
 {
 	pci_unregister_driver(&ice_driver);
 	destroy_workqueue(ice_wq);
+	destroy_workqueue(ice_lag_wq);
 	pr_info("module unloaded\n");
 }
 module_exit(ice_module_exit);
@@ -5703,7 +5768,7 @@ static void ice_set_rx_mode(struct net_device *netdev)
 	struct ice_netdev_priv *np = netdev_priv(netdev);
 	struct ice_vsi *vsi = np->vsi;
 
-	if (!vsi)
+	if (!vsi || ice_is_switchdev_running(vsi->back))
 		return;
 
 	/* Set the flags to synchronize filters
@@ -6255,7 +6320,7 @@ static void ice_tx_dim_work(struct work_struct *work)
 	u16 itr;
 
 	dim = container_of(work, struct dim, work);
-	rc = (struct ice_ring_container *)dim->priv;
+	rc = dim->priv;
 
 	WARN_ON(dim->profile_ix >= ARRAY_SIZE(tx_profile));
 
@@ -6275,7 +6340,7 @@ static void ice_rx_dim_work(struct work_struct *work)
 	u16 itr;
 
 	dim = container_of(work, struct dim, work);
-	rc = (struct ice_ring_container *)dim->priv;
+	rc = dim->priv;
 
 	WARN_ON(dim->profile_ix >= ARRAY_SIZE(rx_profile));
 
@@ -7356,6 +7421,8 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)
 	clear_bit(ICE_RESET_FAILED, pf->state);
 
 	ice_plug_aux_dev(pf);
+	if (ice_is_feature_supported(pf, ICE_F_SRIOV_LAG))
+		ice_lag_rebuild(pf);
 	return;
 
 err_vsi_rebuild:
@@ -8813,6 +8880,7 @@ ice_setup_tc(struct net_device *netdev, enum tc_setup_type type,
 {
 	struct ice_netdev_priv *np = netdev_priv(netdev);
 	struct ice_pf *pf = np->vsi->back;
+	bool locked = false;
 	int err;
 
 	switch (type) {
@@ -8822,10 +8890,32 @@ ice_setup_tc(struct net_device *netdev, enum tc_setup_type type,
 						  ice_setup_tc_block_cb,
 						  np, np, true);
 	case TC_SETUP_QDISC_MQPRIO:
+		if (ice_is_eswitch_mode_switchdev(pf)) {
+			netdev_err(netdev, "TC MQPRIO offload not supported, switchdev is enabled\n");
+			return -EOPNOTSUPP;
+		}
+
+		if (pf->adev) {
+			mutex_lock(&pf->adev_mutex);
+			device_lock(&pf->adev->dev);
+			locked = true;
+			if (pf->adev->dev.driver) {
+				netdev_err(netdev, "Cannot change qdisc when RDMA is active\n");
+				err = -EBUSY;
+				goto adev_unlock;
+			}
+		}
+
 		/* setup traffic classifier for receive side */
 		mutex_lock(&pf->tc_mutex);
 		err = ice_setup_tc_mqprio_qdisc(netdev, type_data);
 		mutex_unlock(&pf->tc_mutex);
+
+adev_unlock:
+		if (locked) {
+			device_unlock(&pf->adev->dev);
+			mutex_unlock(&pf->adev_mutex);
+		}
 		return err;
 	default:
 		return -EOPNOTSUPP;