1 files changed, 348 insertions, 119 deletions

diff --git a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
index 7108fb41b604..add90e75f05c 100644
--- a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c
@@ -5,61 +5,97 @@
 #include "ice_dcb_nl.h"
 
 /**
- * ice_vsi_cfg_netdev_tc - Setup the netdev TC configuration
- * @vsi: the VSI being configured
- * @ena_tc: TC map to be enabled
+ * ice_dcb_get_ena_tc - return bitmap of enabled TCs
+ * @dcbcfg: DCB config to evaluate for enabled TCs
  */
-void ice_vsi_cfg_netdev_tc(struct ice_vsi *vsi, u8 ena_tc)
+static u8 ice_dcb_get_ena_tc(struct ice_dcbx_cfg *dcbcfg)
 {
-	struct net_device *netdev = vsi->netdev;
-	struct ice_pf *pf = vsi->back;
-	struct ice_dcbx_cfg *dcbcfg;
-	u8 netdev_tc;
-	int i;
+	u8 i, num_tc, ena_tc = 1;
 
-	if (!netdev)
-		return;
+	num_tc = ice_dcb_get_num_tc(dcbcfg);
 
-	if (!ena_tc) {
-		netdev_reset_tc(netdev);
-		return;
-	}
+	for (i = 0; i < num_tc; i++)
+		ena_tc |= BIT(i);
 
-	if (netdev_set_num_tc(netdev, vsi->tc_cfg.numtc))
-		return;
+	return ena_tc;
+}
+
+/**
+ * ice_is_pfc_causing_hung_q
+ * @pf: pointer to PF structure
+ * @txqueue: Tx queue which is supposedly hung queue
+ *
+ * find if PFC is causing the hung queue, if yes return true else false
+ */
+bool ice_is_pfc_causing_hung_q(struct ice_pf *pf, unsigned int txqueue)
+{
+	u8 num_tcs = 0, i, tc, up_mapped_tc, up_in_tc = 0;
+	u64 ref_prio_xoff[ICE_MAX_UP];
+	struct ice_vsi *vsi;
+	u32 up2tc;
 
-	dcbcfg = &pf->hw.port_info->local_dcbx_cfg;
+	vsi = ice_get_main_vsi(pf);
+	if (!vsi)
+		return false;
 
 	ice_for_each_traffic_class(i)
 		if (vsi->tc_cfg.ena_tc & BIT(i))
-			netdev_set_tc_queue(netdev,
-					    vsi->tc_cfg.tc_info[i].netdev_tc,
-					    vsi->tc_cfg.tc_info[i].qcount_tx,
-					    vsi->tc_cfg.tc_info[i].qoffset);
+			num_tcs++;
 
-	for (i = 0; i < ICE_MAX_USER_PRIORITY; i++) {
-		u8 ets_tc = dcbcfg->etscfg.prio_table[i];
+	/* first find out the TC to which the hung queue belongs to */
+	for (tc = 0; tc < num_tcs - 1; tc++)
+		if (ice_find_q_in_range(vsi->tc_cfg.tc_info[tc].qoffset,
+					vsi->tc_cfg.tc_info[tc + 1].qoffset,
+					txqueue))
+			break;
 
-		/* Get the mapped netdev TC# for the UP */
-		netdev_tc = vsi->tc_cfg.tc_info[ets_tc].netdev_tc;
-		netdev_set_prio_tc_map(netdev, i, netdev_tc);
+	/* Build a bit map of all UPs associated to the suspect hung queue TC,
+	 * so that we check for its counter increment.
+	 */
+	up2tc = rd32(&pf->hw, PRTDCB_TUP2TC);
+	for (i = 0; i < ICE_MAX_UP; i++) {
+		up_mapped_tc = (up2tc >> (i * 3)) & 0x7;
+		if (up_mapped_tc == tc)
+			up_in_tc |= BIT(i);
 	}
+
+	/* Now that we figured out that hung queue is PFC enabled, still the
+	 * Tx timeout can be legitimate. So to make sure Tx timeout is
+	 * absolutely caused by PFC storm, check if the counters are
+	 * incrementing.
+	 */
+	for (i = 0; i < ICE_MAX_UP; i++)
+		if (up_in_tc & BIT(i))
+			ref_prio_xoff[i] = pf->stats.priority_xoff_rx[i];
+
+	ice_update_dcb_stats(pf);
+
+	for (i = 0; i < ICE_MAX_UP; i++)
+		if (up_in_tc & BIT(i))
+			if (pf->stats.priority_xoff_rx[i] > ref_prio_xoff[i])
+				return true;
+
+	return false;
 }
 
 /**
- * ice_dcb_get_ena_tc - return bitmap of enabled TCs
- * @dcbcfg: DCB config to evaluate for enabled TCs
+ * ice_dcb_get_mode - gets the DCB mode
+ * @port_info: pointer to port info structure
+ * @host: if set it's HOST if not it's MANAGED
  */
-u8 ice_dcb_get_ena_tc(struct ice_dcbx_cfg *dcbcfg)
+static u8 ice_dcb_get_mode(struct ice_port_info *port_info, bool host)
 {
-	u8 i, num_tc, ena_tc = 1;
-
-	num_tc = ice_dcb_get_num_tc(dcbcfg);
+	u8 mode;
 
-	for (i = 0; i < num_tc; i++)
-		ena_tc |= BIT(i);
+	if (host)
+		mode = DCB_CAP_DCBX_HOST;
+	else
+		mode = DCB_CAP_DCBX_LLD_MANAGED;
 
-	return ena_tc;
+	if (port_info->qos_cfg.local_dcbx_cfg.dcbx_mode & ICE_DCBX_MODE_CEE)
+		return mode | DCB_CAP_DCBX_VER_CEE;
+	else
+		return mode | DCB_CAP_DCBX_VER_IEEE;
 }
 
 /**
@@ -101,6 +137,67 @@ u8 ice_dcb_get_num_tc(struct ice_dcbx_cfg *dcbcfg)
 }
 
 /**
+ * ice_get_first_droptc - returns number of first droptc
+ * @vsi: used to find the first droptc
+ *
+ * This function returns the value of first_droptc.
+ * When DCB is enabled, first droptc information is derived from enabled_tc
+ * and PFC enabled bits. otherwise this function returns 0 as there is one
+ * TC without DCB (tc0)
+ */
+static u8 ice_get_first_droptc(struct ice_vsi *vsi)
+{
+	struct ice_dcbx_cfg *cfg = &vsi->port_info->qos_cfg.local_dcbx_cfg;
+	struct device *dev = ice_pf_to_dev(vsi->back);
+	u8 num_tc, ena_tc_map, pfc_ena_map;
+	u8 i;
+
+	num_tc = ice_dcb_get_num_tc(cfg);
+
+	/* get bitmap of enabled TCs */
+	ena_tc_map = ice_dcb_get_ena_tc(cfg);
+
+	/* get bitmap of PFC enabled TCs */
+	pfc_ena_map = cfg->pfc.pfcena;
+
+	/* get first TC that is not PFC enabled */
+	for (i = 0; i < num_tc; i++) {
+		if ((ena_tc_map & BIT(i)) && (!(pfc_ena_map & BIT(i)))) {
+			dev_dbg(dev, "first drop tc = %d\n", i);
+			return i;
+		}
+	}
+
+	dev_dbg(dev, "first drop tc = 0\n");
+	return 0;
+}
+
+/**
+ * ice_vsi_set_dcb_tc_cfg - Set VSI's TC based on DCB configuration
+ * @vsi: pointer to the VSI instance
+ */
+void ice_vsi_set_dcb_tc_cfg(struct ice_vsi *vsi)
+{
+	struct ice_dcbx_cfg *cfg = &vsi->port_info->qos_cfg.local_dcbx_cfg;
+
+	switch (vsi->type) {
+	case ICE_VSI_PF:
+		vsi->tc_cfg.ena_tc = ice_dcb_get_ena_tc(cfg);
+		vsi->tc_cfg.numtc = ice_dcb_get_num_tc(cfg);
+		break;
+	case ICE_VSI_CHNL:
+		vsi->tc_cfg.ena_tc = BIT(ice_get_first_droptc(vsi));
+		vsi->tc_cfg.numtc = 1;
+		break;
+	case ICE_VSI_CTRL:
+	case ICE_VSI_LB:
+	default:
+		vsi->tc_cfg.ena_tc = ICE_DFLT_TRAFFIC_CLASS;
+		vsi->tc_cfg.numtc = 1;
+	}
+}
+
+/**
  * ice_dcb_get_tc - Get the TC associated with the queue
  * @vsi: ptr to the VSI
  * @queue_index: queue number associated with VSI
@@ -116,17 +213,18 @@ u8 ice_dcb_get_tc(struct ice_vsi *vsi, int queue_index)
  */
 void ice_vsi_cfg_dcb_rings(struct ice_vsi *vsi)
 {
-	struct ice_ring *tx_ring, *rx_ring;
+	struct ice_tx_ring *tx_ring;
+	struct ice_rx_ring *rx_ring;
 	u16 qoffset, qcount;
 	int i, n;
 
 	if (!test_bit(ICE_FLAG_DCB_ENA, vsi->back->flags)) {
 		/* Reset the TC information */
-		for (i = 0; i < vsi->num_txq; i++) {
+		ice_for_each_txq(vsi, i) {
 			tx_ring = vsi->tx_rings[i];
 			tx_ring->dcb_tc = 0;
 		}
-		for (i = 0; i < vsi->num_rxq; i++) {
+		ice_for_each_rxq(vsi, i) {
 			rx_ring = vsi->rx_rings[i];
 			rx_ring->dcb_tc = 0;
 		}
@@ -139,16 +237,110 @@ void ice_vsi_cfg_dcb_rings(struct ice_vsi *vsi)
 
 		qoffset = vsi->tc_cfg.tc_info[n].qoffset;
 		qcount = vsi->tc_cfg.tc_info[n].qcount_tx;
-		for (i = qoffset; i < (qoffset + qcount); i++) {
-			tx_ring = vsi->tx_rings[i];
-			rx_ring = vsi->rx_rings[i];
-			tx_ring->dcb_tc = n;
-			rx_ring->dcb_tc = n;
+		for (i = qoffset; i < (qoffset + qcount); i++)
+			vsi->tx_rings[i]->dcb_tc = n;
+
+		qcount = vsi->tc_cfg.tc_info[n].qcount_rx;
+		for (i = qoffset; i < (qoffset + qcount); i++)
+			vsi->rx_rings[i]->dcb_tc = n;
+	}
+	/* applicable only if "all_enatc" is set, which will be set from
+	 * setup_tc method as part of configuring channels
+	 */
+	if (vsi->all_enatc) {
+		u8 first_droptc = ice_get_first_droptc(vsi);
+
+		/* When DCB is configured, TC for ADQ queues (which are really
+		 * PF queues) should be the first drop TC of the main VSI
+		 */
+		ice_for_each_chnl_tc(n) {
+			if (!(vsi->all_enatc & BIT(n)))
+				break;
+
+			qoffset = vsi->mqprio_qopt.qopt.offset[n];
+			qcount = vsi->mqprio_qopt.qopt.count[n];
+			for (i = qoffset; i < (qoffset + qcount); i++) {
+				vsi->tx_rings[i]->dcb_tc = first_droptc;
+				vsi->rx_rings[i]->dcb_tc = first_droptc;
+			}
 		}
 	}
 }
 
 /**
+ * ice_dcb_ena_dis_vsi - disable certain VSIs for DCB config/reconfig
+ * @pf: pointer to the PF instance
+ * @ena: true to enable VSIs, false to disable
+ * @locked: true if caller holds RTNL lock, false otherwise
+ *
+ * Before a new DCB configuration can be applied, VSIs of type PF, SWITCHDEV
+ * and CHNL need to be brought down. Following completion of DCB configuration
+ * the VSIs that were downed need to be brought up again. This helper function
+ * does both.
+ */
+static void ice_dcb_ena_dis_vsi(struct ice_pf *pf, bool ena, bool locked)
+{
+	int i;
+
+	ice_for_each_vsi(pf, i) {
+		struct ice_vsi *vsi = pf->vsi[i];
+
+		if (!vsi)
+			continue;
+
+		switch (vsi->type) {
+		case ICE_VSI_CHNL:
+		case ICE_VSI_SWITCHDEV_CTRL:
+		case ICE_VSI_PF:
+			if (ena)
+				ice_ena_vsi(vsi, locked);
+			else
+				ice_dis_vsi(vsi, locked);
+			break;
+		default:
+			continue;
+		}
+	}
+}
+
+/**
+ * ice_dcb_bwchk - check if ETS bandwidth input parameters are correct
+ * @pf: pointer to the PF struct
+ * @dcbcfg: pointer to DCB config structure
+ */
+int ice_dcb_bwchk(struct ice_pf *pf, struct ice_dcbx_cfg *dcbcfg)
+{
+	struct ice_dcb_ets_cfg *etscfg = &dcbcfg->etscfg;
+	u8 num_tc, total_bw = 0;
+	int i;
+
+	/* returns number of contigous TCs and 1 TC for non-contigous TCs,
+	 * since at least 1 TC has to be configured
+	 */
+	num_tc = ice_dcb_get_num_tc(dcbcfg);
+
+	/* no bandwidth checks required if there's only one TC, so assign
+	 * all bandwidth to TC0 and return
+	 */
+	if (num_tc == 1) {
+		etscfg->tcbwtable[0] = ICE_TC_MAX_BW;
+		return 0;
+	}
+
+	for (i = 0; i < num_tc; i++)
+		total_bw += etscfg->tcbwtable[i];
+
+	if (!total_bw) {
+		etscfg->tcbwtable[0] = ICE_TC_MAX_BW;
+	} else if (total_bw != ICE_TC_MAX_BW) {
+		dev_err(ice_pf_to_dev(pf), "Invalid config, total bandwidth must equal 100\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+/**
  * ice_pf_dcb_cfg - Apply new DCB configuration
  * @pf: pointer to the PF struct
  * @new_cfg: DCBX config to apply
@@ -160,12 +352,13 @@ int ice_pf_dcb_cfg(struct ice_pf *pf, struct ice_dcbx_cfg *new_cfg, bool locked)
 	struct ice_dcbx_cfg *old_cfg, *curr_cfg;
 	struct device *dev = ice_pf_to_dev(pf);
 	int ret = ICE_DCB_NO_HW_CHG;
+	struct iidc_event *event;
 	struct ice_vsi *pf_vsi;
 
-	curr_cfg = &pf->hw.port_info->local_dcbx_cfg;
+	curr_cfg = &pf->hw.port_info->qos_cfg.local_dcbx_cfg;
 
 	/* FW does not care if change happened */
-	if (!pf->hw.port_info->is_sw_lldp)
+	if (!pf->hw.port_info->qos_cfg.is_sw_lldp)
 		ret = ICE_DCB_HW_CHG_RST;
 
 	/* Enable DCB tagging only when more than one TC */
@@ -182,6 +375,9 @@ int ice_pf_dcb_cfg(struct ice_pf *pf, struct ice_dcbx_cfg *new_cfg, bool locked)
 		return ret;
 	}
 
+	if (ice_dcb_bwchk(pf, new_cfg))
+		return -EINVAL;
+
 	/* Store old config in case FW config fails */
 	old_cfg = kmemdup(curr_cfg, sizeof(*old_cfg), GFP_KERNEL);
 	if (!old_cfg)
@@ -195,12 +391,25 @@ int ice_pf_dcb_cfg(struct ice_pf *pf, struct ice_dcbx_cfg *new_cfg, bool locked)
 		goto free_cfg;
 	}
 
+	/* Notify AUX drivers about impending change to TCs */
+	event = kzalloc(sizeof(*event), GFP_KERNEL);
+	if (!event) {
+		ret = -ENOMEM;
+		goto free_cfg;
+	}
+
+	set_bit(IIDC_EVENT_BEFORE_TC_CHANGE, event->type);
+	ice_send_event_to_aux(pf, event);
+	kfree(event);
+
 	/* avoid race conditions by holding the lock while disabling and
 	 * re-enabling the VSI
 	 */
 	if (!locked)
 		rtnl_lock();
-	ice_dis_vsi(pf_vsi, true);
+
+	/* disable VSIs affected by DCB changes */
+	ice_dcb_ena_dis_vsi(pf, false, true);
 
 	memcpy(curr_cfg, new_cfg, sizeof(*curr_cfg));
 	memcpy(&curr_cfg->etsrec, &curr_cfg->etscfg, sizeof(curr_cfg->etsrec));
@@ -209,7 +418,7 @@ int ice_pf_dcb_cfg(struct ice_pf *pf, struct ice_dcbx_cfg *new_cfg, bool locked)
 	/* Only send new config to HW if we are in SW LLDP mode. Otherwise,
 	 * the new config came from the HW in the first place.
 	 */
-	if (pf->hw.port_info->is_sw_lldp) {
+	if (pf->hw.port_info->qos_cfg.is_sw_lldp) {
 		ret = ice_set_dcb_cfg(pf->hw.port_info);
 		if (ret) {
 			dev_err(dev, "Set DCB Config failed\n");
@@ -228,7 +437,8 @@ int ice_pf_dcb_cfg(struct ice_pf *pf, struct ice_dcbx_cfg *new_cfg, bool locked)
 	ice_pf_dcb_recfg(pf);
 
 out:
-	ice_ena_vsi(pf_vsi, true);
+	/* enable previously downed VSIs */
+	ice_dcb_ena_dis_vsi(pf, true, true);
 	if (!locked)
 		rtnl_unlock();
 free_cfg:
@@ -242,7 +452,7 @@ free_cfg:
  */
 static void ice_cfg_etsrec_defaults(struct ice_port_info *pi)
 {
-	struct ice_dcbx_cfg *dcbcfg = &pi->local_dcbx_cfg;
+	struct ice_dcbx_cfg *dcbcfg = &pi->qos_cfg.local_dcbx_cfg;
 	u8 i;
 
 	/* Ensure ETS recommended DCB configuration is not already set */
@@ -318,7 +528,7 @@ void ice_dcb_rebuild(struct ice_pf *pf)
 	struct ice_aqc_port_ets_elem buf = { 0 };
 	struct device *dev = ice_pf_to_dev(pf);
 	struct ice_dcbx_cfg *err_cfg;
-	enum ice_status ret;
+	int ret;
 
 	ret = ice_query_port_ets(pf->hw.port_info, &buf, sizeof(buf), NULL);
 	if (ret) {
@@ -326,13 +536,9 @@ void ice_dcb_rebuild(struct ice_pf *pf)
 		goto dcb_error;
 	}
 
-	/* If DCB was not enabled previously, we are done */
-	if (!test_bit(ICE_FLAG_DCB_ENA, pf->flags))
-		return;
-
 	mutex_lock(&pf->tc_mutex);
 
-	if (!pf->hw.port_info->is_sw_lldp)
+	if (!pf->hw.port_info->qos_cfg.is_sw_lldp)
 		ice_cfg_etsrec_defaults(pf->hw.port_info);
 
 	ret = ice_set_dcb_cfg(pf->hw.port_info);
@@ -341,15 +547,15 @@ void ice_dcb_rebuild(struct ice_pf *pf)
 		goto dcb_error;
 	}
 
-	if (!pf->hw.port_info->is_sw_lldp) {
+	if (!pf->hw.port_info->qos_cfg.is_sw_lldp) {
 		ret = ice_cfg_lldp_mib_change(&pf->hw, true);
-		if (ret && !pf->hw.port_info->is_sw_lldp) {
+		if (ret && !pf->hw.port_info->qos_cfg.is_sw_lldp) {
 			dev_err(dev, "Failed to register for MIB changes\n");
 			goto dcb_error;
 		}
 	}
 
-	dev_info(dev, "DCB restored after reset\n");
+	dev_info(dev, "DCB info restored\n");
 	ret = ice_query_port_ets(pf->hw.port_info, &buf, sizeof(buf), NULL);
 	if (ret) {
 		dev_err(dev, "Query Port ETS failed\n");
@@ -396,11 +602,12 @@ static int ice_dcb_init_cfg(struct ice_pf *pf, bool locked)
 	int ret = 0;
 
 	pi = pf->hw.port_info;
-	newcfg = kmemdup(&pi->local_dcbx_cfg, sizeof(*newcfg), GFP_KERNEL);
+	newcfg = kmemdup(&pi->qos_cfg.local_dcbx_cfg, sizeof(*newcfg),
+			 GFP_KERNEL);
 	if (!newcfg)
 		return -ENOMEM;
 
-	memset(&pi->local_dcbx_cfg, 0, sizeof(*newcfg));
+	memset(&pi->qos_cfg.local_dcbx_cfg, 0, sizeof(*newcfg));
 
 	dev_info(ice_pf_to_dev(pf), "Configuring initial DCB values\n");
 	if (ice_pf_dcb_cfg(pf, newcfg, locked))
@@ -417,7 +624,7 @@ static int ice_dcb_init_cfg(struct ice_pf *pf, bool locked)
  * @ets_willing: configure ETS willing
  * @locked: was this function called with RTNL held
  */
-static int ice_dcb_sw_dflt_cfg(struct ice_pf *pf, bool ets_willing, bool locked)
+int ice_dcb_sw_dflt_cfg(struct ice_pf *pf, bool ets_willing, bool locked)
 {
 	struct ice_aqc_port_ets_elem buf = { 0 };
 	struct ice_dcbx_cfg *dcbcfg;
@@ -431,7 +638,7 @@ static int ice_dcb_sw_dflt_cfg(struct ice_pf *pf, bool ets_willing, bool locked)
 	if (!dcbcfg)
 		return -ENOMEM;
 
-	memset(&pi->local_dcbx_cfg, 0, sizeof(*dcbcfg));
+	memset(&pi->qos_cfg.local_dcbx_cfg, 0, sizeof(*dcbcfg));
 
 	dcbcfg->etscfg.willing = ets_willing ? 1 : 0;
 	dcbcfg->etscfg.maxtcs = hw->func_caps.common_cap.maxtc;
@@ -448,7 +655,7 @@ static int ice_dcb_sw_dflt_cfg(struct ice_pf *pf, bool ets_willing, bool locked)
 	dcbcfg->numapps = 1;
 	dcbcfg->app[0].selector = ICE_APP_SEL_ETHTYPE;
 	dcbcfg->app[0].priority = 3;
-	dcbcfg->app[0].prot_id = ICE_APP_PROT_ID_FCOE;
+	dcbcfg->app[0].prot_id = ETH_P_FCOE;
 
 	ret = ice_pf_dcb_cfg(pf, dcbcfg, locked);
 	kfree(dcbcfg);
@@ -466,16 +673,21 @@ static int ice_dcb_sw_dflt_cfg(struct ice_pf *pf, bool ets_willing, bool locked)
  */
 static bool ice_dcb_tc_contig(u8 *prio_table)
 {
-	u8 max_tc = 0;
+	bool found_empty = false;
+	u8 used_tc = 0;
 	int i;
 
-	for (i = 0; i < CEE_DCBX_MAX_PRIO; i++) {
-		u8 cur_tc = prio_table[i];
+	/* Create a bitmap of used TCs */
+	for (i = 0; i < CEE_DCBX_MAX_PRIO; i++)
+		used_tc |= BIT(prio_table[i]);
 
-		if (cur_tc > max_tc)
-			return false;
-		else if (cur_tc == max_tc)
-			max_tc++;
+	for (i = 0; i < CEE_DCBX_MAX_PRIO; i++) {
+		if (used_tc & BIT(i)) {
+			if (found_empty)
+				return false;
+		} else {
+			found_empty = true;
+		}
 	}
 
 	return true;
@@ -489,7 +701,7 @@ static bool ice_dcb_tc_contig(u8 *prio_table)
  */
 static int ice_dcb_noncontig_cfg(struct ice_pf *pf)
 {
-	struct ice_dcbx_cfg *dcbcfg = &pf->hw.port_info->local_dcbx_cfg;
+	struct ice_dcbx_cfg *dcbcfg = &pf->hw.port_info->qos_cfg.local_dcbx_cfg;
 	struct device *dev = ice_pf_to_dev(pf);
 	int ret;
 
@@ -519,7 +731,8 @@ static int ice_dcb_noncontig_cfg(struct ice_pf *pf)
  */
 void ice_pf_dcb_recfg(struct ice_pf *pf)
 {
-	struct ice_dcbx_cfg *dcbcfg = &pf->hw.port_info->local_dcbx_cfg;
+	struct ice_dcbx_cfg *dcbcfg = &pf->hw.port_info->qos_cfg.local_dcbx_cfg;
+	struct iidc_event *event;
 	u8 tc_map = 0;
 	int v, ret;
 
@@ -540,6 +753,8 @@ void ice_pf_dcb_recfg(struct ice_pf *pf)
 				tc_map = ICE_DFLT_TRAFFIC_CLASS;
 				ice_dcb_noncontig_cfg(pf);
 			}
+		} else if (vsi->type == ICE_VSI_CHNL) {
+			tc_map = BIT(ice_get_first_droptc(vsi));
 		} else {
 			tc_map = ICE_DFLT_TRAFFIC_CLASS;
 		}
@@ -550,11 +765,25 @@ void ice_pf_dcb_recfg(struct ice_pf *pf)
 				vsi->idx);
 			continue;
 		}
+		/* no need to proceed with remaining cfg if it is CHNL
+		 * or switchdev VSI
+		 */
+		if (vsi->type == ICE_VSI_CHNL ||
+		    vsi->type == ICE_VSI_SWITCHDEV_CTRL)
+			continue;
 
 		ice_vsi_map_rings_to_vectors(vsi);
 		if (vsi->type == ICE_VSI_PF)
 			ice_dcbnl_set_all(vsi);
 	}
+	/* Notify the AUX drivers that TC change is finished */
+	event = kzalloc(sizeof(*event), GFP_KERNEL);
+	if (!event)
+		return;
+
+	set_bit(IIDC_EVENT_AFTER_TC_CHANGE, event->type);
+	ice_send_event_to_aux(pf, event);
+	kfree(event);
 }
 
 /**
@@ -572,7 +801,7 @@ int ice_init_pf_dcb(struct ice_pf *pf, bool locked)
 	port_info = hw->port_info;
 
 	err = ice_init_dcb(hw, false);
-	if (err && !port_info->is_sw_lldp) {
+	if (err && !port_info->qos_cfg.is_sw_lldp) {
 		dev_err(dev, "Error initializing DCB %d\n", err);
 		goto dcb_init_err;
 	}
@@ -585,6 +814,11 @@ int ice_init_pf_dcb(struct ice_pf *pf, bool locked)
 		/* FW LLDP is disabled, activate SW DCBX/LLDP mode */
 		dev_info(dev, "FW LLDP is disabled, DCBx/LLDP in SW mode.\n");
 		clear_bit(ICE_FLAG_FW_LLDP_AGENT, pf->flags);
+		err = ice_aq_set_pfc_mode(&pf->hw, ICE_AQC_PFC_VLAN_BASED_PFC,
+					  NULL);
+		if (err)
+			dev_info(dev, "Failed to set VLAN PFC mode\n");
+
 		err = ice_dcb_sw_dflt_cfg(pf, true, locked);
 		if (err) {
 			dev_err(dev, "Failed to set local DCB config %d\n",
@@ -605,14 +839,14 @@ int ice_init_pf_dcb(struct ice_pf *pf, bool locked)
 
 		ice_cfg_sw_lldp(pf_vsi, false, true);
 
-		pf->dcbx_cap = DCB_CAP_DCBX_HOST | DCB_CAP_DCBX_VER_IEEE;
+		pf->dcbx_cap = ice_dcb_get_mode(port_info, true);
 		return 0;
 	}
 
 	set_bit(ICE_FLAG_FW_LLDP_AGENT, pf->flags);
 
-	/* DCBX in FW and LLDP enabled in FW */
-	pf->dcbx_cap = DCB_CAP_DCBX_LLD_MANAGED | DCB_CAP_DCBX_VER_IEEE;
+	/* DCBX/LLDP enabled in FW, set DCBNL mode advertisement */
+	pf->dcbx_cap = ice_dcb_get_mode(port_info, false);
 
 	err = ice_dcb_init_cfg(pf, locked);
 	if (err)
@@ -668,39 +902,35 @@ void ice_update_dcb_stats(struct ice_pf *pf)
  * ice_tx_prepare_vlan_flags_dcb - prepare VLAN tagging for DCB
  * @tx_ring: ring to send buffer on
  * @first: pointer to struct ice_tx_buf
+ *
+ * This should not be called if the outer VLAN is software offloaded as the VLAN
+ * tag will already be configured with the correct ID and priority bits
  */
-int
-ice_tx_prepare_vlan_flags_dcb(struct ice_ring *tx_ring,
+void
+ice_tx_prepare_vlan_flags_dcb(struct ice_tx_ring *tx_ring,
 			      struct ice_tx_buf *first)
 {
 	struct sk_buff *skb = first->skb;
 
 	if (!test_bit(ICE_FLAG_DCB_ENA, tx_ring->vsi->back->flags))
-		return 0;
+		return;
 
 	/* Insert 802.1p priority into VLAN header */
-	if ((first->tx_flags & (ICE_TX_FLAGS_HW_VLAN | ICE_TX_FLAGS_SW_VLAN)) ||
+	if ((first->tx_flags & ICE_TX_FLAGS_HW_VLAN ||
+	     first->tx_flags & ICE_TX_FLAGS_HW_OUTER_SINGLE_VLAN) ||
 	    skb->priority != TC_PRIO_CONTROL) {
 		first->tx_flags &= ~ICE_TX_FLAGS_VLAN_PR_M;
 		/* Mask the lower 3 bits to set the 802.1p priority */
 		first->tx_flags |= (skb->priority & 0x7) <<
 				   ICE_TX_FLAGS_VLAN_PR_S;
-		if (first->tx_flags & ICE_TX_FLAGS_SW_VLAN) {
-			struct vlan_ethhdr *vhdr;
-			int rc;
-
-			rc = skb_cow_head(skb, 0);
-			if (rc < 0)
-				return rc;
-			vhdr = (struct vlan_ethhdr *)skb->data;
-			vhdr->h_vlan_TCI = htons(first->tx_flags >>
-						 ICE_TX_FLAGS_VLAN_S);
-		} else {
+		/* if this is not already set it means a VLAN 0 + priority needs
+		 * to be offloaded
+		 */
+		if (tx_ring->flags & ICE_TX_FLAGS_RING_VLAN_L2TAG2)
+			first->tx_flags |= ICE_TX_FLAGS_HW_OUTER_SINGLE_VLAN;
+		else
 			first->tx_flags |= ICE_TX_FLAGS_HW_VLAN;
-		}
 	}
-
-	return 0;
 }
 
 /**
@@ -718,8 +948,7 @@ ice_dcb_process_lldp_set_mib_change(struct ice_pf *pf,
 	struct ice_dcbx_cfg tmp_dcbx_cfg;
 	bool need_reconfig = false;
 	struct ice_port_info *pi;
-	struct ice_vsi *pf_vsi;
-	u8 type;
+	u8 mib_type;
 	int ret;
 
 	/* Not DCB capable or capability disabled */
@@ -734,20 +963,20 @@ ice_dcb_process_lldp_set_mib_change(struct ice_pf *pf,
 	pi = pf->hw.port_info;
 	mib = (struct ice_aqc_lldp_get_mib *)&event->desc.params.raw;
 	/* Ignore if event is not for Nearest Bridge */
-	type = ((mib->type >> ICE_AQ_LLDP_BRID_TYPE_S) &
-		ICE_AQ_LLDP_BRID_TYPE_M);
-	dev_dbg(dev, "LLDP event MIB bridge type 0x%x\n", type);
-	if (type != ICE_AQ_LLDP_BRID_TYPE_NEAREST_BRID)
+	mib_type = ((mib->type >> ICE_AQ_LLDP_BRID_TYPE_S) &
+		    ICE_AQ_LLDP_BRID_TYPE_M);
+	dev_dbg(dev, "LLDP event MIB bridge type 0x%x\n", mib_type);
+	if (mib_type != ICE_AQ_LLDP_BRID_TYPE_NEAREST_BRID)
 		return;
 
 	/* Check MIB Type and return if event for Remote MIB update */
-	type = mib->type & ICE_AQ_LLDP_MIB_TYPE_M;
-	dev_dbg(dev, "LLDP event mib type %s\n", type ? "remote" : "local");
-	if (type == ICE_AQ_LLDP_MIB_REMOTE) {
+	mib_type = mib->type & ICE_AQ_LLDP_MIB_TYPE_M;
+	dev_dbg(dev, "LLDP event mib type %s\n", mib_type ? "remote" : "local");
+	if (mib_type == ICE_AQ_LLDP_MIB_REMOTE) {
 		/* Update the remote cached instance and return */
 		ret = ice_aq_get_dcb_cfg(pi->hw, ICE_AQ_LLDP_MIB_REMOTE,
 					 ICE_AQ_LLDP_BRID_TYPE_NEAREST_BRID,
-					 &pi->remote_dcbx_cfg);
+					 &pi->qos_cfg.remote_dcbx_cfg);
 		if (ret) {
 			dev_err(dev, "Failed to get remote DCB config\n");
 			return;
@@ -757,10 +986,11 @@ ice_dcb_process_lldp_set_mib_change(struct ice_pf *pf,
 	mutex_lock(&pf->tc_mutex);
 
 	/* store the old configuration */
-	tmp_dcbx_cfg = pf->hw.port_info->local_dcbx_cfg;
+	tmp_dcbx_cfg = pf->hw.port_info->qos_cfg.local_dcbx_cfg;
 
 	/* Reset the old DCBX configuration data */
-	memset(&pi->local_dcbx_cfg, 0, sizeof(pi->local_dcbx_cfg));
+	memset(&pi->qos_cfg.local_dcbx_cfg, 0,
+	       sizeof(pi->qos_cfg.local_dcbx_cfg));
 
 	/* Get updated DCBX data from firmware */
 	ret = ice_get_dcb_cfg(pf->hw.port_info);
@@ -770,19 +1000,22 @@ ice_dcb_process_lldp_set_mib_change(struct ice_pf *pf,
 	}
 
 	/* No change detected in DCBX configs */
-	if (!memcmp(&tmp_dcbx_cfg, &pi->local_dcbx_cfg, sizeof(tmp_dcbx_cfg))) {
+	if (!memcmp(&tmp_dcbx_cfg, &pi->qos_cfg.local_dcbx_cfg,
+		    sizeof(tmp_dcbx_cfg))) {
 		dev_dbg(dev, "No change detected in DCBX configuration.\n");
 		goto out;
 	}
 
+	pf->dcbx_cap = ice_dcb_get_mode(pi, false);
+
 	need_reconfig = ice_dcb_need_recfg(pf, &tmp_dcbx_cfg,
-					   &pi->local_dcbx_cfg);
-	ice_dcbnl_flush_apps(pf, &tmp_dcbx_cfg, &pi->local_dcbx_cfg);
+					   &pi->qos_cfg.local_dcbx_cfg);
+	ice_dcbnl_flush_apps(pf, &tmp_dcbx_cfg, &pi->qos_cfg.local_dcbx_cfg);
 	if (!need_reconfig)
 		goto out;
 
 	/* Enable DCB tagging only when more than one TC */
-	if (ice_dcb_get_num_tc(&pi->local_dcbx_cfg) > 1) {
+	if (ice_dcb_get_num_tc(&pi->qos_cfg.local_dcbx_cfg) > 1) {
 		dev_dbg(dev, "DCB tagging enabled (num TC > 1)\n");
 		set_bit(ICE_FLAG_DCB_ENA, pf->flags);
 	} else {
@@ -790,14 +1023,9 @@ ice_dcb_process_lldp_set_mib_change(struct ice_pf *pf,
 		clear_bit(ICE_FLAG_DCB_ENA, pf->flags);
 	}
 
-	pf_vsi = ice_get_main_vsi(pf);
-	if (!pf_vsi) {
-		dev_dbg(dev, "PF VSI doesn't exist\n");
-		goto out;
-	}
-
 	rtnl_lock();
-	ice_dis_vsi(pf_vsi, true);
+	/* disable VSIs affected by DCB changes */
+	ice_dcb_ena_dis_vsi(pf, false, true);
 
 	ret = ice_query_port_ets(pf->hw.port_info, &buf, sizeof(buf), NULL);
 	if (ret) {
@@ -808,7 +1036,8 @@ ice_dcb_process_lldp_set_mib_change(struct ice_pf *pf,
 	/* changes in configuration update VSI */
 	ice_pf_dcb_recfg(pf);
 
-	ice_ena_vsi(pf_vsi, true);
+	/* enable previously downed VSIs */
+	ice_dcb_ena_dis_vsi(pf, true, true);
 unlock_rtnl:
 	rtnl_unlock();
 out: