1 files changed, 287 insertions, 156 deletions

diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
index d13c7fc8fb0a..82e2ce23df3d 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_lib.c
@@ -158,6 +158,8 @@ static void ice_vsi_set_num_qs(struct ice_vsi *vsi, u16 vf_id)
 
 	if (vsi->type == ICE_VSI_VF)
 		vsi->vf_id = vf_id;
+	else
+		vsi->vf_id = ICE_INVAL_VFID;
 
 	switch (vsi->type) {
 	case ICE_VSI_PF:
@@ -343,6 +345,9 @@ static int ice_vsi_clear(struct ice_vsi *vsi)
 	pf->vsi[vsi->idx] = NULL;
 	if (vsi->idx < pf->next_vsi && vsi->type != ICE_VSI_CTRL)
 		pf->next_vsi = vsi->idx;
+	if (vsi->idx < pf->next_vsi && vsi->type == ICE_VSI_CTRL &&
+	    vsi->vf_id != ICE_INVAL_VFID)
+		pf->next_vsi = vsi->idx;
 
 	ice_vsi_free_arrays(vsi);
 	mutex_unlock(&pf->sw_mutex);
@@ -382,6 +387,8 @@ static irqreturn_t ice_msix_clean_rings(int __always_unused irq, void *data)
 	if (!q_vector->tx.ring && !q_vector->rx.ring)
 		return IRQ_HANDLED;
 
+	q_vector->total_events++;
+
 	napi_schedule(&q_vector->napi);
 
 	return IRQ_HANDLED;
@@ -419,7 +426,7 @@ ice_vsi_alloc(struct ice_pf *pf, enum ice_vsi_type vsi_type, u16 vf_id)
 
 	vsi->type = vsi_type;
 	vsi->back = pf;
-	set_bit(__ICE_DOWN, vsi->state);
+	set_bit(ICE_VSI_DOWN, vsi->state);
 
 	if (vsi_type == ICE_VSI_VF)
 		ice_vsi_set_num_qs(vsi, vf_id);
@@ -454,8 +461,8 @@ ice_vsi_alloc(struct ice_pf *pf, enum ice_vsi_type vsi_type, u16 vf_id)
 		goto unlock_pf;
 	}
 
-	if (vsi->type == ICE_VSI_CTRL) {
-		/* Use the last VSI slot as the index for the control VSI */
+	if (vsi->type == ICE_VSI_CTRL && vf_id == ICE_INVAL_VFID) {
+		/* Use the last VSI slot as the index for PF control VSI */
 		vsi->idx = pf->num_alloc_vsi - 1;
 		pf->ctrl_vsi_idx = vsi->idx;
 		pf->vsi[vsi->idx] = vsi;
@@ -468,6 +475,9 @@ ice_vsi_alloc(struct ice_pf *pf, enum ice_vsi_type vsi_type, u16 vf_id)
 		pf->next_vsi = ice_get_free_slot(pf->vsi, pf->num_alloc_vsi,
 						 pf->next_vsi);
 	}
+
+	if (vsi->type == ICE_VSI_CTRL && vf_id != ICE_INVAL_VFID)
+		pf->vf[vf_id].ctrl_vsi_idx = vsi->idx;
 	goto unlock_pf;
 
 err_rings:
@@ -506,7 +516,7 @@ static int ice_alloc_fd_res(struct ice_vsi *vsi)
 	if (!b_val)
 		return -EPERM;
 
-	if (vsi->type != ICE_VSI_PF)
+	if (!(vsi->type == ICE_VSI_PF || vsi->type == ICE_VSI_VF))
 		return -EPERM;
 
 	if (!test_bit(ICE_FLAG_FD_ENA, pf->flags))
@@ -517,6 +527,13 @@ static int ice_alloc_fd_res(struct ice_vsi *vsi)
 	/* each VSI gets same "best_effort" quota */
 	vsi->num_bfltr = b_val;
 
+	if (vsi->type == ICE_VSI_VF) {
+		vsi->num_gfltr = 0;
+
+		/* each VSI gets same "best_effort" quota */
+		vsi->num_bfltr = b_val;
+	}
+
 	return 0;
 }
 
@@ -729,11 +746,10 @@ static void ice_set_dflt_vsi_ctx(struct ice_vsi_ctx *ctxt)
  */
 static void ice_vsi_setup_q_map(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt)
 {
-	u16 offset = 0, qmap = 0, tx_count = 0;
+	u16 offset = 0, qmap = 0, tx_count = 0, pow = 0;
+	u16 num_txq_per_tc, num_rxq_per_tc;
 	u16 qcount_tx = vsi->alloc_txq;
 	u16 qcount_rx = vsi->alloc_rxq;
-	u16 tx_numq_tc, rx_numq_tc;
-	u16 pow = 0, max_rss = 0;
 	bool ena_tc0 = false;
 	u8 netdev_tc = 0;
 	int i;
@@ -751,12 +767,15 @@ static void ice_vsi_setup_q_map(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt)
 		vsi->tc_cfg.ena_tc |= 1;
 	}
 
-	rx_numq_tc = qcount_rx / vsi->tc_cfg.numtc;
-	if (!rx_numq_tc)
-		rx_numq_tc = 1;
-	tx_numq_tc = qcount_tx / vsi->tc_cfg.numtc;
-	if (!tx_numq_tc)
-		tx_numq_tc = 1;
+	num_rxq_per_tc = min_t(u16, qcount_rx / vsi->tc_cfg.numtc, ICE_MAX_RXQS_PER_TC);
+	if (!num_rxq_per_tc)
+		num_rxq_per_tc = 1;
+	num_txq_per_tc = qcount_tx / vsi->tc_cfg.numtc;
+	if (!num_txq_per_tc)
+		num_txq_per_tc = 1;
+
+	/* find the (rounded up) power-of-2 of qcount */
+	pow = (u16)order_base_2(num_rxq_per_tc);
 
 	/* TC mapping is a function of the number of Rx queues assigned to the
 	 * VSI for each traffic class and the offset of these queues.
@@ -769,26 +788,6 @@ static void ice_vsi_setup_q_map(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt)
 	 *
 	 * Setup number and offset of Rx queues for all TCs for the VSI
 	 */
-
-	qcount_rx = rx_numq_tc;
-
-	/* qcount will change if RSS is enabled */
-	if (test_bit(ICE_FLAG_RSS_ENA, vsi->back->flags)) {
-		if (vsi->type == ICE_VSI_PF || vsi->type == ICE_VSI_VF) {
-			if (vsi->type == ICE_VSI_PF)
-				max_rss = ICE_MAX_LG_RSS_QS;
-			else
-				max_rss = ICE_MAX_RSS_QS_PER_VF;
-			qcount_rx = min_t(u16, rx_numq_tc, max_rss);
-			if (!vsi->req_rxq)
-				qcount_rx = min_t(u16, qcount_rx,
-						  vsi->rss_size);
-		}
-	}
-
-	/* find the (rounded up) power-of-2 of qcount */
-	pow = (u16)order_base_2(qcount_rx);
-
 	ice_for_each_traffic_class(i) {
 		if (!(vsi->tc_cfg.ena_tc & BIT(i))) {
 			/* TC is not enabled */
@@ -802,16 +801,16 @@ static void ice_vsi_setup_q_map(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt)
 
 		/* TC is enabled */
 		vsi->tc_cfg.tc_info[i].qoffset = offset;
-		vsi->tc_cfg.tc_info[i].qcount_rx = qcount_rx;
-		vsi->tc_cfg.tc_info[i].qcount_tx = tx_numq_tc;
+		vsi->tc_cfg.tc_info[i].qcount_rx = num_rxq_per_tc;
+		vsi->tc_cfg.tc_info[i].qcount_tx = num_txq_per_tc;
 		vsi->tc_cfg.tc_info[i].netdev_tc = netdev_tc++;
 
 		qmap = ((offset << ICE_AQ_VSI_TC_Q_OFFSET_S) &
 			ICE_AQ_VSI_TC_Q_OFFSET_M) |
 			((pow << ICE_AQ_VSI_TC_Q_NUM_S) &
 			 ICE_AQ_VSI_TC_Q_NUM_M);
-		offset += qcount_rx;
-		tx_count += tx_numq_tc;
+		offset += num_rxq_per_tc;
+		tx_count += num_txq_per_tc;
 		ctxt->info.tc_mapping[i] = cpu_to_le16(qmap);
 	}
 
@@ -824,7 +823,7 @@ static void ice_vsi_setup_q_map(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt)
 	if (offset)
 		vsi->num_rxq = offset;
 	else
-		vsi->num_rxq = qcount_rx;
+		vsi->num_rxq = num_rxq_per_tc;
 
 	vsi->num_txq = tx_count;
 
@@ -856,7 +855,8 @@ static void ice_set_fd_vsi_ctx(struct ice_vsi_ctx *ctxt, struct ice_vsi *vsi)
 	u8 dflt_q_group, dflt_q_prio;
 	u16 dflt_q, report_q, val;
 
-	if (vsi->type != ICE_VSI_PF && vsi->type != ICE_VSI_CTRL)
+	if (vsi->type != ICE_VSI_PF && vsi->type != ICE_VSI_CTRL &&
+	    vsi->type != ICE_VSI_VF)
 		return;
 
 	val = ICE_AQ_VSI_PROP_FLOW_DIR_VALID;
@@ -1179,7 +1179,24 @@ static int ice_vsi_setup_vector_base(struct ice_vsi *vsi)
 
 	num_q_vectors = vsi->num_q_vectors;
 	/* reserve slots from OS requested IRQs */
-	base = ice_get_res(pf, pf->irq_tracker, num_q_vectors, vsi->idx);
+	if (vsi->type == ICE_VSI_CTRL && vsi->vf_id != ICE_INVAL_VFID) {
+		struct ice_vf *vf;
+		int i;
+
+		ice_for_each_vf(pf, i) {
+			vf = &pf->vf[i];
+			if (i != vsi->vf_id && vf->ctrl_vsi_idx != ICE_NO_VSI) {
+				base = pf->vsi[vf->ctrl_vsi_idx]->base_vector;
+				break;
+			}
+		}
+		if (i == pf->num_alloc_vfs)
+			base = ice_get_res(pf, pf->irq_tracker, num_q_vectors,
+					   ICE_RES_VF_CTRL_VEC_ID);
+	} else {
+		base = ice_get_res(pf, pf->irq_tracker, num_q_vectors,
+				   vsi->idx);
+	}
 
 	if (base < 0) {
 		dev_err(dev, "%d MSI-X interrupts available. %s %d failed to get %d MSI-X vectors\n",
@@ -1296,14 +1313,13 @@ err_out:
  * LUT, while in the event of enable request for RSS, it will reconfigure RSS
  * LUT.
  */
-int ice_vsi_manage_rss_lut(struct ice_vsi *vsi, bool ena)
+void ice_vsi_manage_rss_lut(struct ice_vsi *vsi, bool ena)
 {
-	int err = 0;
 	u8 *lut;
 
 	lut = kzalloc(vsi->rss_table_size, GFP_KERNEL);
 	if (!lut)
-		return -ENOMEM;
+		return;
 
 	if (ena) {
 		if (vsi->rss_lut_user)
@@ -1313,9 +1329,8 @@ int ice_vsi_manage_rss_lut(struct ice_vsi *vsi, bool ena)
 					 vsi->rss_size);
 	}
 
-	err = ice_set_rss(vsi, NULL, lut, vsi->rss_table_size);
+	ice_set_rss_lut(vsi, lut, vsi->rss_table_size);
 	kfree(lut);
-	return err;
 }
 
 /**
@@ -1324,12 +1339,10 @@ int ice_vsi_manage_rss_lut(struct ice_vsi *vsi, bool ena)
  */
 static int ice_vsi_cfg_rss_lut_key(struct ice_vsi *vsi)
 {
-	struct ice_aqc_get_set_rss_keys *key;
 	struct ice_pf *pf = vsi->back;
-	enum ice_status status;
 	struct device *dev;
-	int err = 0;
-	u8 *lut;
+	u8 *lut, *key;
+	int err;
 
 	dev = ice_pf_to_dev(pf);
 	vsi->rss_size = min_t(u16, vsi->rss_size, vsi->num_rxq);
@@ -1343,37 +1356,26 @@ static int ice_vsi_cfg_rss_lut_key(struct ice_vsi *vsi)
 	else
 		ice_fill_rss_lut(lut, vsi->rss_table_size, vsi->rss_size);
 
-	status = ice_aq_set_rss_lut(&pf->hw, vsi->idx, vsi->rss_lut_type, lut,
-				    vsi->rss_table_size);
-
-	if (status) {
-		dev_err(dev, "set_rss_lut failed, error %s\n",
-			ice_stat_str(status));
-		err = -EIO;
+	err = ice_set_rss_lut(vsi, lut, vsi->rss_table_size);
+	if (err) {
+		dev_err(dev, "set_rss_lut failed, error %d\n", err);
 		goto ice_vsi_cfg_rss_exit;
 	}
 
-	key = kzalloc(sizeof(*key), GFP_KERNEL);
+	key = kzalloc(ICE_GET_SET_RSS_KEY_EXTEND_KEY_SIZE, GFP_KERNEL);
 	if (!key) {
 		err = -ENOMEM;
 		goto ice_vsi_cfg_rss_exit;
 	}
 
 	if (vsi->rss_hkey_user)
-		memcpy(key,
-		       (struct ice_aqc_get_set_rss_keys *)vsi->rss_hkey_user,
-		       ICE_GET_SET_RSS_KEY_EXTEND_KEY_SIZE);
+		memcpy(key, vsi->rss_hkey_user, ICE_GET_SET_RSS_KEY_EXTEND_KEY_SIZE);
 	else
-		netdev_rss_key_fill((void *)key,
-				    ICE_GET_SET_RSS_KEY_EXTEND_KEY_SIZE);
+		netdev_rss_key_fill((void *)key, ICE_GET_SET_RSS_KEY_EXTEND_KEY_SIZE);
 
-	status = ice_aq_set_rss_key(&pf->hw, vsi->idx, key);
-
-	if (status) {
-		dev_err(dev, "set_rss_key failed, error %s\n",
-			ice_stat_str(status));
-		err = -EIO;
-	}
+	err = ice_set_rss_key(vsi, key);
+	if (err)
+		dev_err(dev, "set_rss_key failed, error %d\n", err);
 
 	kfree(key);
 ice_vsi_cfg_rss_exit:
@@ -1502,13 +1504,13 @@ static void ice_vsi_set_rss_flow_fld(struct ice_vsi *vsi)
  */
 bool ice_pf_state_is_nominal(struct ice_pf *pf)
 {
-	DECLARE_BITMAP(check_bits, __ICE_STATE_NBITS) = { 0 };
+	DECLARE_BITMAP(check_bits, ICE_STATE_NBITS) = { 0 };
 
 	if (!pf)
 		return false;
 
-	bitmap_set(check_bits, 0, __ICE_STATE_NOMINAL_CHECK_BITS);
-	if (bitmap_intersects(pf->state, check_bits, __ICE_STATE_NBITS))
+	bitmap_set(check_bits, 0, ICE_STATE_NOMINAL_CHECK_BITS);
+	if (bitmap_intersects(pf->state, check_bits, ICE_STATE_NBITS))
 		return false;
 
 	return true;
@@ -1773,7 +1775,7 @@ int ice_vsi_cfg_xdp_txqs(struct ice_vsi *vsi)
  * This function converts a decimal interrupt rate limit in usecs to the format
  * expected by firmware.
  */
-u32 ice_intrl_usec_to_reg(u8 intrl, u8 gran)
+static u32 ice_intrl_usec_to_reg(u8 intrl, u8 gran)
 {
 	u32 val = intrl / gran;
 
@@ -1783,6 +1785,51 @@ u32 ice_intrl_usec_to_reg(u8 intrl, u8 gran)
 }
 
 /**
+ * ice_write_intrl - write throttle rate limit to interrupt specific register
+ * @q_vector: pointer to interrupt specific structure
+ * @intrl: throttle rate limit in microseconds to write
+ */
+void ice_write_intrl(struct ice_q_vector *q_vector, u8 intrl)
+{
+	struct ice_hw *hw = &q_vector->vsi->back->hw;
+
+	wr32(hw, GLINT_RATE(q_vector->reg_idx),
+	     ice_intrl_usec_to_reg(intrl, ICE_INTRL_GRAN_ABOVE_25));
+}
+
+/**
+ * __ice_write_itr - write throttle rate to register
+ * @q_vector: pointer to interrupt data structure
+ * @rc: pointer to ring container
+ * @itr: throttle rate in microseconds to write
+ */
+static void __ice_write_itr(struct ice_q_vector *q_vector,
+			    struct ice_ring_container *rc, u16 itr)
+{
+	struct ice_hw *hw = &q_vector->vsi->back->hw;
+
+	wr32(hw, GLINT_ITR(rc->itr_idx, q_vector->reg_idx),
+	     ITR_REG_ALIGN(itr) >> ICE_ITR_GRAN_S);
+}
+
+/**
+ * ice_write_itr - write throttle rate to queue specific register
+ * @rc: pointer to ring container
+ * @itr: throttle rate in microseconds to write
+ */
+void ice_write_itr(struct ice_ring_container *rc, u16 itr)
+{
+	struct ice_q_vector *q_vector;
+
+	if (!rc->ring)
+		return;
+
+	q_vector = rc->ring->q_vector;
+
+	__ice_write_itr(q_vector, rc, itr);
+}
+
+/**
  * ice_vsi_cfg_msix - MSIX mode Interrupt Config in the HW
  * @vsi: the VSI being configured
  *
@@ -1802,9 +1849,6 @@ void ice_vsi_cfg_msix(struct ice_vsi *vsi)
 
 		ice_cfg_itr(hw, q_vector);
 
-		wr32(hw, GLINT_RATE(reg_idx),
-		     ice_intrl_usec_to_reg(q_vector->intrl, hw->intrl_gran));
-
 		/* Both Transmit Queue Interrupt Cause Control register
 		 * and Receive Queue Interrupt Cause control register
 		 * expects MSIX_INDX field to be the vector index
@@ -2308,7 +2352,7 @@ ice_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi,
 	struct ice_vsi *vsi;
 	int ret, i;
 
-	if (vsi_type == ICE_VSI_VF)
+	if (vsi_type == ICE_VSI_VF || vsi_type == ICE_VSI_CTRL)
 		vsi = ice_vsi_alloc(pf, vsi_type, vf_id);
 	else
 		vsi = ice_vsi_alloc(pf, vsi_type, ICE_INVAL_VFID);
@@ -2323,7 +2367,7 @@ ice_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi,
 	if (vsi->type == ICE_VSI_PF)
 		vsi->ethtype = ETH_P_PAUSE;
 
-	if (vsi->type == ICE_VSI_VF)
+	if (vsi->type == ICE_VSI_VF || vsi->type == ICE_VSI_CTRL)
 		vsi->vf_id = vf_id;
 
 	ice_alloc_fd_res(vsi);
@@ -2492,11 +2536,10 @@ static void ice_vsi_release_msix(struct ice_vsi *vsi)
 
 	for (i = 0; i < vsi->num_q_vectors; i++) {
 		struct ice_q_vector *q_vector = vsi->q_vectors[i];
-		u16 reg_idx = q_vector->reg_idx;
 
-		wr32(hw, GLINT_ITR(ICE_IDX_ITR0, reg_idx), 0);
-		wr32(hw, GLINT_ITR(ICE_IDX_ITR1, reg_idx), 0);
+		ice_write_intrl(q_vector, 0);
 		for (q = 0; q < q_vector->num_ring_tx; q++) {
+			ice_write_itr(&q_vector->tx, 0);
 			wr32(hw, QINT_TQCTL(vsi->txq_map[txq]), 0);
 			if (ice_is_xdp_ena_vsi(vsi)) {
 				u32 xdp_txq = txq + vsi->num_xdp_txq;
@@ -2507,6 +2550,7 @@ static void ice_vsi_release_msix(struct ice_vsi *vsi)
 		}
 
 		for (q = 0; q < q_vector->num_ring_rx; q++) {
+			ice_write_itr(&q_vector->rx, 0);
 			wr32(hw, QINT_RQCTL(vsi->rxq_map[rxq]), 0);
 			rxq++;
 		}
@@ -2593,7 +2637,7 @@ void ice_vsi_free_rx_rings(struct ice_vsi *vsi)
  */
 void ice_vsi_close(struct ice_vsi *vsi)
 {
-	if (!test_and_set_bit(__ICE_DOWN, vsi->state))
+	if (!test_and_set_bit(ICE_VSI_DOWN, vsi->state))
 		ice_down(vsi);
 
 	ice_vsi_free_irq(vsi);
@@ -2610,10 +2654,10 @@ int ice_ena_vsi(struct ice_vsi *vsi, bool locked)
 {
 	int err = 0;
 
-	if (!test_bit(__ICE_NEEDS_RESTART, vsi->state))
+	if (!test_bit(ICE_VSI_NEEDS_RESTART, vsi->state))
 		return 0;
 
-	clear_bit(__ICE_NEEDS_RESTART, vsi->state);
+	clear_bit(ICE_VSI_NEEDS_RESTART, vsi->state);
 
 	if (vsi->netdev && vsi->type == ICE_VSI_PF) {
 		if (netif_running(vsi->netdev)) {
@@ -2639,10 +2683,10 @@ int ice_ena_vsi(struct ice_vsi *vsi, bool locked)
  */
 void ice_dis_vsi(struct ice_vsi *vsi, bool locked)
 {
-	if (test_bit(__ICE_DOWN, vsi->state))
+	if (test_bit(ICE_VSI_DOWN, vsi->state))
 		return;
 
-	set_bit(__ICE_NEEDS_RESTART, vsi->state);
+	set_bit(ICE_VSI_NEEDS_RESTART, vsi->state);
 
 	if (vsi->type == ICE_VSI_PF && vsi->netdev) {
 		if (netif_running(vsi->netdev)) {
@@ -2752,11 +2796,14 @@ int ice_vsi_release(struct ice_vsi *vsi)
 	 * PF that is running the work queue items currently. This is done to
 	 * avoid check_flush_dependency() warning on this wq
 	 */
-	if (vsi->netdev && !ice_is_reset_in_progress(pf->state)) {
+	if (vsi->netdev && !ice_is_reset_in_progress(pf->state) &&
+	    (test_bit(ICE_VSI_NETDEV_REGISTERED, vsi->state))) {
 		unregister_netdev(vsi->netdev);
-		ice_devlink_destroy_port(vsi);
+		clear_bit(ICE_VSI_NETDEV_REGISTERED, vsi->state);
 	}
 
+	ice_devlink_destroy_port(vsi);
+
 	if (test_bit(ICE_FLAG_RSS_ENA, pf->flags))
 		ice_rss_clean(vsi);
 
@@ -2770,7 +2817,24 @@ int ice_vsi_release(struct ice_vsi *vsi)
 	 * many interrupts each VF needs. SR-IOV MSIX resources are also
 	 * cleared in the same manner.
 	 */
-	if (vsi->type != ICE_VSI_VF) {
+	if (vsi->type == ICE_VSI_CTRL && vsi->vf_id != ICE_INVAL_VFID) {
+		struct ice_vf *vf;
+		int i;
+
+		ice_for_each_vf(pf, i) {
+			vf = &pf->vf[i];
+			if (i != vsi->vf_id && vf->ctrl_vsi_idx != ICE_NO_VSI)
+				break;
+		}
+		if (i == pf->num_alloc_vfs) {
+			/* No other VFs left that have control VSI, reclaim SW
+			 * interrupts back to the common pool
+			 */
+			ice_free_res(pf->irq_tracker, vsi->base_vector,
+				     ICE_RES_VF_CTRL_VEC_ID);
+			pf->num_avail_sw_msix += vsi->num_q_vectors;
+		}
+	} else if (vsi->type != ICE_VSI_VF) {
 		/* reclaim SW interrupts back to the common pool */
 		ice_free_res(pf->irq_tracker, vsi->base_vector, vsi->idx);
 		pf->num_avail_sw_msix += vsi->num_q_vectors;
@@ -2794,10 +2858,16 @@ int ice_vsi_release(struct ice_vsi *vsi)
 	ice_vsi_delete(vsi);
 	ice_vsi_free_q_vectors(vsi);
 
-	/* make sure unregister_netdev() was called by checking __ICE_DOWN */
-	if (vsi->netdev && test_bit(__ICE_DOWN, vsi->state)) {
-		free_netdev(vsi->netdev);
-		vsi->netdev = NULL;
+	if (vsi->netdev) {
+		if (test_bit(ICE_VSI_NETDEV_REGISTERED, vsi->state)) {
+			unregister_netdev(vsi->netdev);
+			clear_bit(ICE_VSI_NETDEV_REGISTERED, vsi->state);
+		}
+		if (test_bit(ICE_VSI_NETDEV_ALLOCD, vsi->state)) {
+			free_netdev(vsi->netdev);
+			vsi->netdev = NULL;
+			clear_bit(ICE_VSI_NETDEV_ALLOCD, vsi->state);
+		}
 	}
 
 	if (vsi->type == ICE_VSI_VF &&
@@ -2818,39 +2888,6 @@ int ice_vsi_release(struct ice_vsi *vsi)
 }
 
 /**
- * ice_vsi_rebuild_update_coalesce - set coalesce for a q_vector
- * @q_vector: pointer to q_vector which is being updated
- * @coalesce: pointer to array of struct with stored coalesce
- *
- * Set coalesce param in q_vector and update these parameters in HW.
- */
-static void
-ice_vsi_rebuild_update_coalesce(struct ice_q_vector *q_vector,
-				struct ice_coalesce_stored *coalesce)
-{
-	struct ice_ring_container *rx_rc = &q_vector->rx;
-	struct ice_ring_container *tx_rc = &q_vector->tx;
-	struct ice_hw *hw = &q_vector->vsi->back->hw;
-
-	tx_rc->itr_setting = coalesce->itr_tx;
-	rx_rc->itr_setting = coalesce->itr_rx;
-
-	/* dynamic ITR values will be updated during Tx/Rx */
-	if (!ITR_IS_DYNAMIC(tx_rc->itr_setting))
-		wr32(hw, GLINT_ITR(tx_rc->itr_idx, q_vector->reg_idx),
-		     ITR_REG_ALIGN(tx_rc->itr_setting) >>
-		     ICE_ITR_GRAN_S);
-	if (!ITR_IS_DYNAMIC(rx_rc->itr_setting))
-		wr32(hw, GLINT_ITR(rx_rc->itr_idx, q_vector->reg_idx),
-		     ITR_REG_ALIGN(rx_rc->itr_setting) >>
-		     ICE_ITR_GRAN_S);
-
-	q_vector->intrl = coalesce->intrl;
-	wr32(hw, GLINT_RATE(q_vector->reg_idx),
-	     ice_intrl_usec_to_reg(q_vector->intrl, hw->intrl_gran));
-}
-
-/**
  * ice_vsi_rebuild_get_coalesce - get coalesce from all q_vectors
  * @vsi: VSI connected with q_vectors
  * @coalesce: array of struct with stored coalesce
@@ -2869,6 +2906,11 @@ ice_vsi_rebuild_get_coalesce(struct ice_vsi *vsi,
 		coalesce[i].itr_tx = q_vector->tx.itr_setting;
 		coalesce[i].itr_rx = q_vector->rx.itr_setting;
 		coalesce[i].intrl = q_vector->intrl;
+
+		if (i < vsi->num_txq)
+			coalesce[i].tx_valid = true;
+		if (i < vsi->num_rxq)
+			coalesce[i].rx_valid = true;
 	}
 
 	return vsi->num_q_vectors;
@@ -2888,22 +2930,75 @@ static void
 ice_vsi_rebuild_set_coalesce(struct ice_vsi *vsi,
 			     struct ice_coalesce_stored *coalesce, int size)
 {
+	struct ice_ring_container *rc;
 	int i;
 
 	if ((size && !coalesce) || !vsi)
 		return;
 
-	for (i = 0; i < size && i < vsi->num_q_vectors; i++)
-		ice_vsi_rebuild_update_coalesce(vsi->q_vectors[i],
-						&coalesce[i]);
+	/* There are a couple of cases that have to be handled here:
+	 *   1. The case where the number of queue vectors stays the same, but
+	 *      the number of Tx or Rx rings changes (the first for loop)
+	 *   2. The case where the number of queue vectors increased (the
+	 *      second for loop)
+	 */
+	for (i = 0; i < size && i < vsi->num_q_vectors; i++) {
+		/* There are 2 cases to handle here and they are the same for
+		 * both Tx and Rx:
+		 *   if the entry was valid previously (coalesce[i].[tr]x_valid
+		 *   and the loop variable is less than the number of rings
+		 *   allocated, then write the previous values
+		 *
+		 *   if the entry was not valid previously, but the number of
+		 *   rings is less than are allocated (this means the number of
+		 *   rings increased from previously), then write out the
+		 *   values in the first element
+		 *
+		 *   Also, always write the ITR, even if in ITR_IS_DYNAMIC
+		 *   as there is no harm because the dynamic algorithm
+		 *   will just overwrite.
+		 */
+		if (i < vsi->alloc_rxq && coalesce[i].rx_valid) {
+			rc = &vsi->q_vectors[i]->rx;
+			rc->itr_setting = coalesce[i].itr_rx;
+			ice_write_itr(rc, rc->itr_setting);
+		} else if (i < vsi->alloc_rxq) {
+			rc = &vsi->q_vectors[i]->rx;
+			rc->itr_setting = coalesce[0].itr_rx;
+			ice_write_itr(rc, rc->itr_setting);
+		}
+
+		if (i < vsi->alloc_txq && coalesce[i].tx_valid) {
+			rc = &vsi->q_vectors[i]->tx;
+			rc->itr_setting = coalesce[i].itr_tx;
+			ice_write_itr(rc, rc->itr_setting);
+		} else if (i < vsi->alloc_txq) {
+			rc = &vsi->q_vectors[i]->tx;
+			rc->itr_setting = coalesce[0].itr_tx;
+			ice_write_itr(rc, rc->itr_setting);
+		}
+
+		vsi->q_vectors[i]->intrl = coalesce[i].intrl;
+		ice_write_intrl(vsi->q_vectors[i], coalesce[i].intrl);
+	}
 
-	/* number of q_vectors increased, so assume coalesce settings were
-	 * changed globally (i.e. ethtool -C eth0 instead of per-queue) and use
-	 * the previous settings from q_vector 0 for all of the new q_vectors
+	/* the number of queue vectors increased so write whatever is in
+	 * the first element
 	 */
-	for (; i < vsi->num_q_vectors; i++)
-		ice_vsi_rebuild_update_coalesce(vsi->q_vectors[i],
-						&coalesce[0]);
+	for (; i < vsi->num_q_vectors; i++) {
+		/* transmit */
+		rc = &vsi->q_vectors[i]->tx;
+		rc->itr_setting = coalesce[0].itr_tx;
+		ice_write_itr(rc, rc->itr_setting);
+
+		/* receive */
+		rc = &vsi->q_vectors[i]->rx;
+		rc->itr_setting = coalesce[0].itr_rx;
+		ice_write_itr(rc, rc->itr_setting);
+
+		vsi->q_vectors[i]->intrl = coalesce[0].intrl;
+		ice_write_intrl(vsi->q_vectors[i], coalesce[0].intrl);
+	}
 }
 
 /**
@@ -2919,6 +3014,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi)
 	struct ice_coalesce_stored *coalesce;
 	int prev_num_q_vectors = 0;
 	struct ice_vf *vf = NULL;
+	enum ice_vsi_type vtype;
 	enum ice_status status;
 	struct ice_pf *pf;
 	int ret, i;
@@ -2927,14 +3023,17 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi)
 		return -EINVAL;
 
 	pf = vsi->back;
-	if (vsi->type == ICE_VSI_VF)
+	vtype = vsi->type;
+	if (vtype == ICE_VSI_VF)
 		vf = &pf->vf[vsi->vf_id];
 
 	coalesce = kcalloc(vsi->num_q_vectors,
 			   sizeof(struct ice_coalesce_stored), GFP_KERNEL);
-	if (coalesce)
-		prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi,
-								  coalesce);
+	if (!coalesce)
+		return -ENOMEM;
+
+	prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi, coalesce);
+
 	ice_rm_vsi_lan_cfg(vsi->port_info, vsi->idx);
 	ice_vsi_free_q_vectors(vsi);
 
@@ -2943,7 +3042,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi)
 	 * many interrupts each VF needs. SR-IOV MSIX resources are also
 	 * cleared in the same manner.
 	 */
-	if (vsi->type != ICE_VSI_VF) {
+	if (vtype != ICE_VSI_VF) {
 		/* reclaim SW interrupts back to the common pool */
 		ice_free_res(pf->irq_tracker, vsi->base_vector, vsi->idx);
 		pf->num_avail_sw_msix += vsi->num_q_vectors;
@@ -2958,7 +3057,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi)
 	ice_vsi_put_qs(vsi);
 	ice_vsi_clear_rings(vsi);
 	ice_vsi_free_arrays(vsi);
-	if (vsi->type == ICE_VSI_VF)
+	if (vtype == ICE_VSI_VF)
 		ice_vsi_set_num_qs(vsi, vf->vf_id);
 	else
 		ice_vsi_set_num_qs(vsi, ICE_INVAL_VFID);
@@ -2977,7 +3076,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi)
 	if (ret < 0)
 		goto err_vsi;
 
-	switch (vsi->type) {
+	switch (vtype) {
 	case ICE_VSI_CTRL:
 	case ICE_VSI_PF:
 		ret = ice_vsi_alloc_q_vectors(vsi);
@@ -3004,7 +3103,7 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi)
 				goto err_vectors;
 		}
 		/* ICE_VSI_CTRL does not need RSS so skip RSS processing */
-		if (vsi->type != ICE_VSI_CTRL)
+		if (vtype != ICE_VSI_CTRL)
 			/* Do not exit if configuring RSS had an issue, at
 			 * least receive traffic on first queue. Hence no
 			 * need to capture return value
@@ -3066,7 +3165,7 @@ err_rings:
 	}
 err_vsi:
 	ice_vsi_clear(vsi);
-	set_bit(__ICE_RESET_FAILED, pf->state);
+	set_bit(ICE_RESET_FAILED, pf->state);
 	kfree(coalesce);
 	return ret;
 }
@@ -3077,10 +3176,10 @@ err_vsi:
  */
 bool ice_is_reset_in_progress(unsigned long *state)
 {
-	return test_bit(__ICE_RESET_OICR_RECV, state) ||
-	       test_bit(__ICE_PFR_REQ, state) ||
-	       test_bit(__ICE_CORER_REQ, state) ||
-	       test_bit(__ICE_GLOBR_REQ, state);
+	return test_bit(ICE_RESET_OICR_RECV, state) ||
+	       test_bit(ICE_PFR_REQ, state) ||
+	       test_bit(ICE_CORER_REQ, state) ||
+	       test_bit(ICE_GLOBR_REQ, state);
 }
 
 #ifdef CONFIG_DCB
@@ -3168,20 +3267,15 @@ out:
 /**
  * ice_update_ring_stats - Update ring statistics
  * @ring: ring to update
- * @cont: used to increment per-vector counters
  * @pkts: number of processed packets
  * @bytes: number of processed bytes
  *
  * This function assumes that caller has acquired a u64_stats_sync lock.
  */
-static void
-ice_update_ring_stats(struct ice_ring *ring, struct ice_ring_container *cont,
-		      u64 pkts, u64 bytes)
+static void ice_update_ring_stats(struct ice_ring *ring, u64 pkts, u64 bytes)
 {
 	ring->stats.bytes += bytes;
 	ring->stats.pkts += pkts;
-	cont->total_bytes += bytes;
-	cont->total_pkts += pkts;
 }
 
 /**
@@ -3193,7 +3287,7 @@ ice_update_ring_stats(struct ice_ring *ring, struct ice_ring_container *cont,
 void ice_update_tx_ring_stats(struct ice_ring *tx_ring, u64 pkts, u64 bytes)
 {
 	u64_stats_update_begin(&tx_ring->syncp);
-	ice_update_ring_stats(tx_ring, &tx_ring->q_vector->tx, pkts, bytes);
+	ice_update_ring_stats(tx_ring, pkts, bytes);
 	u64_stats_update_end(&tx_ring->syncp);
 }
 
@@ -3206,7 +3300,7 @@ void ice_update_tx_ring_stats(struct ice_ring *tx_ring, u64 pkts, u64 bytes)
 void ice_update_rx_ring_stats(struct ice_ring *rx_ring, u64 pkts, u64 bytes)
 {
 	u64_stats_update_begin(&rx_ring->syncp);
-	ice_update_ring_stats(rx_ring, &rx_ring->q_vector->rx, pkts, bytes);
+	ice_update_ring_stats(rx_ring, pkts, bytes);
 	u64_stats_update_end(&rx_ring->syncp);
 }
 
@@ -3348,3 +3442,40 @@ int ice_clear_dflt_vsi(struct ice_sw *sw)
 
 	return 0;
 }
+
+/**
+ * ice_set_link - turn on/off physical link
+ * @vsi: VSI to modify physical link on
+ * @ena: turn on/off physical link
+ */
+int ice_set_link(struct ice_vsi *vsi, bool ena)
+{
+	struct device *dev = ice_pf_to_dev(vsi->back);
+	struct ice_port_info *pi = vsi->port_info;
+	struct ice_hw *hw = pi->hw;
+	enum ice_status status;
+
+	if (vsi->type != ICE_VSI_PF)
+		return -EINVAL;
+
+	status = ice_aq_set_link_restart_an(pi, ena, NULL);
+
+	/* if link is owned by manageability, FW will return ICE_AQ_RC_EMODE.
+	 * this is not a fatal error, so print a warning message and return
+	 * a success code. Return an error if FW returns an error code other
+	 * than ICE_AQ_RC_EMODE
+	 */
+	if (status == ICE_ERR_AQ_ERROR) {
+		if (hw->adminq.sq_last_status == ICE_AQ_RC_EMODE)
+			dev_warn(dev, "can't set link to %s, err %s aq_err %s. not fatal, continuing\n",
+				 (ena ? "ON" : "OFF"), ice_stat_str(status),
+				 ice_aq_str(hw->adminq.sq_last_status));
+	} else if (status) {
+		dev_err(dev, "can't set link to %s, err %s aq_err %s\n",
+			(ena ? "ON" : "OFF"), ice_stat_str(status),
+			ice_aq_str(hw->adminq.sq_last_status));
+		return -EIO;
+	}
+
+	return 0;
+}