Merge branch 'qed-XDP-support'

Yuval Mintz says:

====================
qed*: Add XDP support

This patch series is intended to add XDP to the qede driver, although
it contains quite a bit of cleanups, refactorings and infrastructure
changes as well.

The content of this series can be roughly divided into:

 - Datapath improvements - mostly focused on having the datapath utilize
parameters which can be more tightly contained in cachelines.
Patches #1, #2, #8, #9 belong to this group.

 - Refactoring - done mostly in favour of XDP. Patches #3, #4, #5, #9.

 - Infrastructure changes - done in favour of XDP. Paches #6 and #7 belong
to this category [#7 being by far the biggest patch in the series].

 - Actual XDP support - last two patches [#10, #11].
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

14 files changed, 1879 insertions, 1291 deletions

diff --git a/drivers/net/ethernet/qlogic/qed/qed.h b/drivers/net/ethernet/qlogic/qed/qed.h
index 50b8a01ff512..244dd40ccac3 100644
--- a/drivers/net/ethernet/qlogic/qed/qed.h
+++ b/drivers/net/ethernet/qlogic/qed/qed.h
@@ -241,15 +241,6 @@ struct qed_hw_info {
 	enum qed_wol_support b_wol_support;
 };
 
-struct qed_hw_cid_data {
-	u32	cid;
-	bool	b_cid_allocated;
-
-	/* Additional identifiers */
-	u16	opaque_fid;
-	u8	vport_id;
-};
-
 /* maximun size of read/write commands (HW limit) */
 #define DMAE_MAX_RW_SIZE        0x2000
 
@@ -416,9 +407,6 @@ struct qed_hwfn {
 
 	struct qed_dcbx_info		*p_dcbx_info;
 
-	struct qed_hw_cid_data		*p_tx_cids;
-	struct qed_hw_cid_data		*p_rx_cids;
-
 	struct qed_dmae_info		dmae_info;
 
 	/* QM init */
diff --git a/drivers/net/ethernet/qlogic/qed/qed_dev.c b/drivers/net/ethernet/qlogic/qed/qed_dev.c
index 5be7b8a25425..00b9a67ba359 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_dev.c
+++ b/drivers/net/ethernet/qlogic/qed/qed_dev.c
@@ -137,15 +137,6 @@ void qed_resc_free(struct qed_dev *cdev)
 	for_each_hwfn(cdev, i) {
 		struct qed_hwfn *p_hwfn = &cdev->hwfns[i];
 
-		kfree(p_hwfn->p_tx_cids);
-		p_hwfn->p_tx_cids = NULL;
-		kfree(p_hwfn->p_rx_cids);
-		p_hwfn->p_rx_cids = NULL;
-	}
-
-	for_each_hwfn(cdev, i) {
-		struct qed_hwfn *p_hwfn = &cdev->hwfns[i];
-
 		qed_cxt_mngr_free(p_hwfn);
 		qed_qm_info_free(p_hwfn);
 		qed_spq_free(p_hwfn);
@@ -425,23 +416,6 @@ int qed_resc_alloc(struct qed_dev *cdev)
 	if (!cdev->fw_data)
 		return -ENOMEM;
 
-	/* Allocate Memory for the Queue->CID mapping */
-	for_each_hwfn(cdev, i) {
-		struct qed_hwfn *p_hwfn = &cdev->hwfns[i];
-		int tx_size = sizeof(struct qed_hw_cid_data) *
-				     RESC_NUM(p_hwfn, QED_L2_QUEUE);
-		int rx_size = sizeof(struct qed_hw_cid_data) *
-				     RESC_NUM(p_hwfn, QED_L2_QUEUE);
-
-		p_hwfn->p_tx_cids = kzalloc(tx_size, GFP_KERNEL);
-		if (!p_hwfn->p_tx_cids)
-			goto alloc_no_mem;
-
-		p_hwfn->p_rx_cids = kzalloc(rx_size, GFP_KERNEL);
-		if (!p_hwfn->p_rx_cids)
-			goto alloc_no_mem;
-	}
-
 	for_each_hwfn(cdev, i) {
 		struct qed_hwfn *p_hwfn = &cdev->hwfns[i];
 		u32 n_eqes, num_cons;
@@ -2283,12 +2257,12 @@ static void qed_chain_free_pbl(struct qed_dev *cdev, struct qed_chain *p_chain)
 {
 	void **pp_virt_addr_tbl = p_chain->pbl.pp_virt_addr_tbl;
 	u32 page_cnt = p_chain->page_cnt, i, pbl_size;
-	u8 *p_pbl_virt = p_chain->pbl.p_virt_table;
+	u8 *p_pbl_virt = p_chain->pbl_sp.p_virt_table;
 
 	if (!pp_virt_addr_tbl)
 		return;
 
-	if (!p_chain->pbl.p_virt_table)
+	if (!p_pbl_virt)
 		goto out;
 
 	for (i = 0; i < page_cnt; i++) {
@@ -2306,7 +2280,8 @@ static void qed_chain_free_pbl(struct qed_dev *cdev, struct qed_chain *p_chain)
 	pbl_size = page_cnt * QED_CHAIN_PBL_ENTRY_SIZE;
 	dma_free_coherent(&cdev->pdev->dev,
 			  pbl_size,
-			  p_chain->pbl.p_virt_table, p_chain->pbl.p_phys_table);
+			  p_chain->pbl_sp.p_virt_table,
+			  p_chain->pbl_sp.p_phys_table);
 out:
 	vfree(p_chain->pbl.pp_virt_addr_tbl);
 }
diff --git a/drivers/net/ethernet/qlogic/qed/qed_l2.c b/drivers/net/ethernet/qlogic/qed/qed_l2.c
index 900b253be317..6a3727c4c0c6 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_l2.c
+++ b/drivers/net/ethernet/qlogic/qed/qed_l2.c
@@ -23,6 +23,7 @@
 #include <linux/workqueue.h>
 #include <linux/bitops.h>
 #include <linux/bug.h>
+#include <linux/vmalloc.h>
 #include "qed.h"
 #include <linux/qed/qed_chain.h>
 #include "qed_cxt.h"
@@ -41,6 +42,124 @@
 #define QED_MAX_SGES_NUM 16
 #define CRC32_POLY 0x1edc6f41
 
+void qed_eth_queue_cid_release(struct qed_hwfn *p_hwfn,
+			       struct qed_queue_cid *p_cid)
+{
+	/* VFs' CIDs are 0-based in PF-view, and uninitialized on VF */
+	if (!p_cid->is_vf && IS_PF(p_hwfn->cdev))
+		qed_cxt_release_cid(p_hwfn, p_cid->cid);
+	vfree(p_cid);
+}
+
+/* The internal is only meant to be directly called by PFs initializeing CIDs
+ * for their VFs.
+ */
+struct qed_queue_cid *
+_qed_eth_queue_to_cid(struct qed_hwfn *p_hwfn,
+		      u16 opaque_fid,
+		      u32 cid,
+		      u8 vf_qid,
+		      struct qed_queue_start_common_params *p_params)
+{
+	bool b_is_same = (p_hwfn->hw_info.opaque_fid == opaque_fid);
+	struct qed_queue_cid *p_cid;
+	int rc;
+
+	p_cid = vmalloc(sizeof(*p_cid));
+	if (!p_cid)
+		return NULL;
+	memset(p_cid, 0, sizeof(*p_cid));
+
+	p_cid->opaque_fid = opaque_fid;
+	p_cid->cid = cid;
+	p_cid->vf_qid = vf_qid;
+	p_cid->rel = *p_params;
+
+	/* Don't try calculating the absolute indices for VFs */
+	if (IS_VF(p_hwfn->cdev)) {
+		p_cid->abs = p_cid->rel;
+		goto out;
+	}
+
+	/* Calculate the engine-absolute indices of the resources.
+	 * This would guarantee they're valid later on.
+	 * In some cases [SBs] we already have the right values.
+	 */
+	rc = qed_fw_vport(p_hwfn, p_cid->rel.vport_id, &p_cid->abs.vport_id);
+	if (rc)
+		goto fail;
+
+	rc = qed_fw_l2_queue(p_hwfn, p_cid->rel.queue_id, &p_cid->abs.queue_id);
+	if (rc)
+		goto fail;
+
+	/* In case of a PF configuring its VF's queues, the stats-id is already
+	 * absolute [since there's a single index that's suitable per-VF].
+	 */
+	if (b_is_same) {
+		rc = qed_fw_vport(p_hwfn, p_cid->rel.stats_id,
+				  &p_cid->abs.stats_id);
+		if (rc)
+			goto fail;
+	} else {
+		p_cid->abs.stats_id = p_cid->rel.stats_id;
+	}
+
+	/* SBs relevant information was already provided as absolute */
+	p_cid->abs.sb = p_cid->rel.sb;
+	p_cid->abs.sb_idx = p_cid->rel.sb_idx;
+
+	/* This is tricky - we're actually interested in whehter this is a PF
+	 * entry meant for the VF.
+	 */
+	if (!b_is_same)
+		p_cid->is_vf = true;
+out:
+	DP_VERBOSE(p_hwfn,
+		   QED_MSG_SP,
+		   "opaque_fid: %04x CID %08x vport %02x [%02x] qzone %04x [%04x] stats %02x [%02x] SB %04x PI %02x\n",
+		   p_cid->opaque_fid,
+		   p_cid->cid,
+		   p_cid->rel.vport_id,
+		   p_cid->abs.vport_id,
+		   p_cid->rel.queue_id,
+		   p_cid->abs.queue_id,
+		   p_cid->rel.stats_id,
+		   p_cid->abs.stats_id, p_cid->abs.sb, p_cid->abs.sb_idx);
+
+	return p_cid;
+
+fail:
+	vfree(p_cid);
+	return NULL;
+}
+
+static struct qed_queue_cid *qed_eth_queue_to_cid(struct qed_hwfn *p_hwfn,
+						  u16 opaque_fid, struct
+						  qed_queue_start_common_params
+						  *p_params)
+{
+	struct qed_queue_cid *p_cid;
+	u32 cid = 0;
+
+	/* Get a unique firmware CID for this queue, in case it's a PF.
+	 * VF's don't need a CID as the queue configuration will be done
+	 * by PF.
+	 */
+	if (IS_PF(p_hwfn->cdev)) {
+		if (qed_cxt_acquire_cid(p_hwfn, PROTOCOLID_ETH, &cid)) {
+			DP_NOTICE(p_hwfn, "Failed to acquire cid\n");
+			return NULL;
+		}
+	}
+
+	p_cid = _qed_eth_queue_to_cid(p_hwfn, opaque_fid, cid, 0, p_params);
+	if (!p_cid && IS_PF(p_hwfn->cdev))
+		qed_cxt_release_cid(p_hwfn, cid);
+
+	return p_cid;
+}
+
 int qed_sp_eth_vport_start(struct qed_hwfn *p_hwfn,
 			   struct qed_sp_vport_start_params *p_params)
 {
@@ -496,61 +615,26 @@ static int qed_filter_accept_cmd(struct qed_dev *cdev,
 	return 0;
 }
 
-static int qed_sp_release_queue_cid(
-	struct qed_hwfn *p_hwfn,
-	struct qed_hw_cid_data *p_cid_data)
-{
-	if (!p_cid_data->b_cid_allocated)
-		return 0;
-
-	qed_cxt_release_cid(p_hwfn, p_cid_data->cid);
-
-	p_cid_data->b_cid_allocated = false;
-
-	return 0;
-}
-
-int qed_sp_eth_rxq_start_ramrod(struct qed_hwfn *p_hwfn,
-				u16 opaque_fid,
-				u32 cid,
-				struct qed_queue_start_common_params *p_params,
-				u8 stats_id,
-				u16 bd_max_bytes,
-				dma_addr_t bd_chain_phys_addr,
-				dma_addr_t cqe_pbl_addr,
-				u16 cqe_pbl_size, bool b_use_zone_a_prod)
+int qed_eth_rxq_start_ramrod(struct qed_hwfn *p_hwfn,
+			     struct qed_queue_cid *p_cid,
+			     u16 bd_max_bytes,
+			     dma_addr_t bd_chain_phys_addr,
+			     dma_addr_t cqe_pbl_addr, u16 cqe_pbl_size)
 {
 	struct rx_queue_start_ramrod_data *p_ramrod = NULL;
 	struct qed_spq_entry *p_ent = NULL;
 	struct qed_sp_init_data init_data;
-	struct qed_hw_cid_data *p_rx_cid;
-	u16 abs_rx_q_id = 0;
-	u8 abs_vport_id = 0;
 	int rc = -EINVAL;
 
-	/* Store information for the stop */
-	p_rx_cid = &p_hwfn->p_rx_cids[p_params->queue_id];
-	p_rx_cid->cid = cid;
-	p_rx_cid->opaque_fid = opaque_fid;
-	p_rx_cid->vport_id = p_params->vport_id;
-
-	rc = qed_fw_vport(p_hwfn, p_params->vport_id, &abs_vport_id);
-	if (rc)
-		return rc;
-
-	rc = qed_fw_l2_queue(p_hwfn, p_params->queue_id, &abs_rx_q_id);
-	if (rc)
-		return rc;
-
 	DP_VERBOSE(p_hwfn, QED_MSG_SP,
-		   "opaque_fid=0x%x, cid=0x%x, rx_qid=0x%x, vport_id=0x%x, sb_id=0x%x\n",
-		   opaque_fid,
-		   cid, p_params->queue_id, p_params->vport_id, p_params->sb);
+		   "opaque_fid=0x%x, cid=0x%x, rx_qzone=0x%x, vport_id=0x%x, sb_id=0x%x\n",
+		   p_cid->opaque_fid, p_cid->cid,
+		   p_cid->abs.queue_id, p_cid->abs.vport_id, p_cid->abs.sb);
 
 	/* Get SPQ entry */
 	memset(&init_data, 0, sizeof(init_data));
-	init_data.cid = cid;
-	init_data.opaque_fid = opaque_fid;
+	init_data.cid = p_cid->cid;
+	init_data.opaque_fid = p_cid->opaque_fid;
 	init_data.comp_mode = QED_SPQ_MODE_EBLOCK;
 
 	rc = qed_sp_init_request(p_hwfn, &p_ent,
@@ -561,11 +645,11 @@ int qed_sp_eth_rxq_start_ramrod(struct qed_hwfn *p_hwfn,
 
 	p_ramrod = &p_ent->ramrod.rx_queue_start;
 
-	p_ramrod->sb_id = cpu_to_le16(p_params->sb);
-	p_ramrod->sb_index = p_params->sb_idx;
-	p_ramrod->vport_id = abs_vport_id;
-	p_ramrod->stats_counter_id = stats_id;
-	p_ramrod->rx_queue_id = cpu_to_le16(abs_rx_q_id);
+	p_ramrod->sb_id = cpu_to_le16(p_cid->abs.sb);
+	p_ramrod->sb_index = p_cid->abs.sb_idx;
+	p_ramrod->vport_id = p_cid->abs.vport_id;
+	p_ramrod->stats_counter_id = p_cid->abs.stats_id;
+	p_ramrod->rx_queue_id = cpu_to_le16(p_cid->abs.queue_id);
 	p_ramrod->complete_cqe_flg = 0;
 	p_ramrod->complete_event_flg = 1;
 
@@ -575,85 +659,85 @@ int qed_sp_eth_rxq_start_ramrod(struct qed_hwfn *p_hwfn,
 	p_ramrod->num_of_pbl_pages = cpu_to_le16(cqe_pbl_size);
 	DMA_REGPAIR_LE(p_ramrod->cqe_pbl_addr, cqe_pbl_addr);
 
-	if (p_params->vf_qid || b_use_zone_a_prod) {
-		p_ramrod->vf_rx_prod_index = p_params->vf_qid;
+	if (p_cid->is_vf) {
+		p_ramrod->vf_rx_prod_index = p_cid->vf_qid;
 		DP_VERBOSE(p_hwfn, QED_MSG_SP,
 			   "Queue%s is meant for VF rxq[%02x]\n",
-			   b_use_zone_a_prod ? " [legacy]" : "",
-			   p_params->vf_qid);
-		p_ramrod->vf_rx_prod_use_zone_a = b_use_zone_a_prod;
+			   !!p_cid->b_legacy_vf ? " [legacy]" : "",
+			   p_cid->vf_qid);
+		p_ramrod->vf_rx_prod_use_zone_a = !!p_cid->b_legacy_vf;
 	}
 
 	return qed_spq_post(p_hwfn, p_ent, NULL);
 }
 
 static int
-qed_sp_eth_rx_queue_start(struct qed_hwfn *p_hwfn,
-			  u16 opaque_fid,
-			  struct qed_queue_start_common_params *p_params,
+qed_eth_pf_rx_queue_start(struct qed_hwfn *p_hwfn,
+			  struct qed_queue_cid *p_cid,
 			  u16 bd_max_bytes,
 			  dma_addr_t bd_chain_phys_addr,
 			  dma_addr_t cqe_pbl_addr,
 			  u16 cqe_pbl_size, void __iomem **pp_prod)
 {
-	struct qed_hw_cid_data *p_rx_cid;
 	u32 init_prod_val = 0;
-	u16 abs_l2_queue = 0;
-	u8 abs_stats_id = 0;
-	int rc;
 
-	if (IS_VF(p_hwfn->cdev)) {
-		return qed_vf_pf_rxq_start(p_hwfn,
-					   p_params->queue_id,
-					   p_params->sb,
-					   (u8)p_params->sb_idx,
-					   bd_max_bytes,
-					   bd_chain_phys_addr,
-					   cqe_pbl_addr, cqe_pbl_size, pp_prod);
-	}
-
-	rc = qed_fw_l2_queue(p_hwfn, p_params->queue_id, &abs_l2_queue);
-	if (rc)
-		return rc;
-
-	rc = qed_fw_vport(p_hwfn, p_params->vport_id, &abs_stats_id);
-	if (rc)
-		return rc;
-
-	*pp_prod = (u8 __iomem *)p_hwfn->regview +
-				 GTT_BAR0_MAP_REG_MSDM_RAM +
-				 MSTORM_ETH_PF_PRODS_OFFSET(abs_l2_queue);
+	*pp_prod = p_hwfn->regview +
+		   GTT_BAR0_MAP_REG_MSDM_RAM +
+		    MSTORM_ETH_PF_PRODS_OFFSET(p_cid->abs.queue_id);
 
 	/* Init the rcq, rx bd and rx sge (if valid) producers to 0 */
 	__internal_ram_wr(p_hwfn, *pp_prod, sizeof(u32),
 			  (u32 *)(&init_prod_val));
 
+	return qed_eth_rxq_start_ramrod(p_hwfn, p_cid,
+					bd_max_bytes,
+					bd_chain_phys_addr,
+					cqe_pbl_addr, cqe_pbl_size);
+}
+
+static int
+qed_eth_rx_queue_start(struct qed_hwfn *p_hwfn,
+		       u16 opaque_fid,
+		       struct qed_queue_start_common_params *p_params,
+		       u16 bd_max_bytes,
+		       dma_addr_t bd_chain_phys_addr,
+		       dma_addr_t cqe_pbl_addr,
+		       u16 cqe_pbl_size,
+		       struct qed_rxq_start_ret_params *p_ret_params)
+{
+	struct qed_queue_cid *p_cid;
+	int rc;
+
 	/* Allocate a CID for the queue */
-	p_rx_cid = &p_hwfn->p_rx_cids[p_params->queue_id];
-	rc = qed_cxt_acquire_cid(p_hwfn, PROTOCOLID_ETH, &p_rx_cid->cid);
-	if (rc) {
-		DP_NOTICE(p_hwfn, "Failed to acquire cid\n");
-		return rc;
-	}
-	p_rx_cid->b_cid_allocated = true;
+	p_cid = qed_eth_queue_to_cid(p_hwfn, opaque_fid, p_params);
+	if (!p_cid)
+		return -ENOMEM;
 
-	rc = qed_sp_eth_rxq_start_ramrod(p_hwfn,
-					 opaque_fid,
-					 p_rx_cid->cid,
-					 p_params,
-					 abs_stats_id,
+	if (IS_PF(p_hwfn->cdev)) {
+		rc = qed_eth_pf_rx_queue_start(p_hwfn, p_cid,
+					       bd_max_bytes,
+					       bd_chain_phys_addr,
+					       cqe_pbl_addr, cqe_pbl_size,
+					       &p_ret_params->p_prod);
+	} else {
+		rc = qed_vf_pf_rxq_start(p_hwfn, p_cid,
 					 bd_max_bytes,
 					 bd_chain_phys_addr,
-					 cqe_pbl_addr, cqe_pbl_size, false);
+					 cqe_pbl_addr,
+					 cqe_pbl_size, &p_ret_params->p_prod);
+	}
 
+	/* Provide the caller with a reference to as handler */
 	if (rc)
-		qed_sp_release_queue_cid(p_hwfn, p_rx_cid);
+		qed_eth_queue_cid_release(p_hwfn, p_cid);
+	else
+		p_ret_params->p_handle = (void *)p_cid;
 
 	return rc;
 }
 
 int qed_sp_eth_rx_queues_update(struct qed_hwfn *p_hwfn,
-				u16 rx_queue_id,
+				void **pp_rxq_handles,
 				u8 num_rxqs,
 				u8 complete_cqe_flg,
 				u8 complete_event_flg,
@@ -663,8 +747,7 @@ int qed_sp_eth_rx_queues_update(struct qed_hwfn *p_hwfn,
 	struct rx_queue_update_ramrod_data *p_ramrod = NULL;
 	struct qed_spq_entry *p_ent = NULL;
 	struct qed_sp_init_data init_data;
-	struct qed_hw_cid_data *p_rx_cid;
-	u16 qid, abs_rx_q_id = 0;
+	struct qed_queue_cid *p_cid;
 	int rc = -EINVAL;
 	u8 i;
 
@@ -673,12 +756,11 @@ int qed_sp_eth_rx_queues_update(struct qed_hwfn *p_hwfn,
 	init_data.p_comp_data = p_comp_data;
 
 	for (i = 0; i < num_rxqs; i++) {
-		qid = rx_queue_id + i;
-		p_rx_cid = &p_hwfn->p_rx_cids[qid];
+		p_cid = ((struct qed_queue_cid **)pp_rxq_handles)[i];
 
 		/* Get SPQ entry */
-		init_data.cid = p_rx_cid->cid;
-		init_data.opaque_fid = p_rx_cid->opaque_fid;
+		init_data.cid = p_cid->cid;
+		init_data.opaque_fid = p_cid->opaque_fid;
 
 		rc = qed_sp_init_request(p_hwfn, &p_ent,
 					 ETH_RAMROD_RX_QUEUE_UPDATE,
@@ -687,10 +769,9 @@ int qed_sp_eth_rx_queues_update(struct qed_hwfn *p_hwfn,
 			return rc;
 
 		p_ramrod = &p_ent->ramrod.rx_queue_update;
+		p_ramrod->vport_id = p_cid->abs.vport_id;
 
-		qed_fw_vport(p_hwfn, p_rx_cid->vport_id, &p_ramrod->vport_id);
-		qed_fw_l2_queue(p_hwfn, qid, &abs_rx_q_id);
-		p_ramrod->rx_queue_id = cpu_to_le16(abs_rx_q_id);
+		p_ramrod->rx_queue_id = cpu_to_le16(p_cid->abs.queue_id);
 		p_ramrod->complete_cqe_flg = complete_cqe_flg;
 		p_ramrod->complete_event_flg = complete_event_flg;
 
@@ -702,24 +783,19 @@ int qed_sp_eth_rx_queues_update(struct qed_hwfn *p_hwfn,
 	return rc;
 }
 
-int qed_sp_eth_rx_queue_stop(struct qed_hwfn *p_hwfn,
-			     u16 rx_queue_id,
-			     bool eq_completion_only, bool cqe_completion)
+static int
+qed_eth_pf_rx_queue_stop(struct qed_hwfn *p_hwfn,
+			 struct qed_queue_cid *p_cid,
+			 bool b_eq_completion_only, bool b_cqe_completion)
 {
-	struct qed_hw_cid_data *p_rx_cid = &p_hwfn->p_rx_cids[rx_queue_id];
 	struct rx_queue_stop_ramrod_data *p_ramrod = NULL;
 	struct qed_spq_entry *p_ent = NULL;
 	struct qed_sp_init_data init_data;
-	u16 abs_rx_q_id = 0;
-	int rc = -EINVAL;
-
-	if (IS_VF(p_hwfn->cdev))
-		return qed_vf_pf_rxq_stop(p_hwfn, rx_queue_id, cqe_completion);
+	int rc;
 
-	/* Get SPQ entry */
 	memset(&init_data, 0, sizeof(init_data));
-	init_data.cid = p_rx_cid->cid;
-	init_data.opaque_fid = p_rx_cid->opaque_fid;
+	init_data.cid = p_cid->cid;
+	init_data.opaque_fid = p_cid->opaque_fid;
 	init_data.comp_mode = QED_SPQ_MODE_EBLOCK;
 
 	rc = qed_sp_init_request(p_hwfn, &p_ent,
@@ -729,62 +805,53 @@ int qed_sp_eth_rx_queue_stop(struct qed_hwfn *p_hwfn,
 		return rc;
 
 	p_ramrod = &p_ent->ramrod.rx_queue_stop;
-
-	qed_fw_vport(p_hwfn, p_rx_cid->vport_id, &p_ramrod->vport_id);
-	qed_fw_l2_queue(p_hwfn, rx_queue_id, &abs_rx_q_id);
-	p_ramrod->rx_queue_id = cpu_to_le16(abs_rx_q_id);
+	p_ramrod->vport_id = p_cid->abs.vport_id;
+	p_ramrod->rx_queue_id = cpu_to_le16(p_cid->abs.queue_id);
 
 	/* Cleaning the queue requires the completion to arrive there.
 	 * In addition, VFs require the answer to come as eqe to PF.
 	 */
-	p_ramrod->complete_cqe_flg =
-		(!!(p_rx_cid->opaque_fid == p_hwfn->hw_info.opaque_fid) &&
-		 !eq_completion_only) || cqe_completion;
-	p_ramrod->complete_event_flg =
-		!(p_rx_cid->opaque_fid == p_hwfn->hw_info.opaque_fid) ||
-		eq_completion_only;
+	p_ramrod->complete_cqe_flg = (!p_cid->is_vf &&
+				      !b_eq_completion_only) ||
+				     b_cqe_completion;
+	p_ramrod->complete_event_flg = p_cid->is_vf || b_eq_completion_only;
 
-	rc = qed_spq_post(p_hwfn, p_ent, NULL);
-	if (rc)
-		return rc;
+	return qed_spq_post(p_hwfn, p_ent, NULL);
+}
+
+int qed_eth_rx_queue_stop(struct qed_hwfn *p_hwfn,
+			  void *p_rxq,
+			  bool eq_completion_only, bool cqe_completion)
+{
+	struct qed_queue_cid *p_cid = (struct qed_queue_cid *)p_rxq;
+	int rc = -EINVAL;
 
-	return qed_sp_release_queue_cid(p_hwfn, p_rx_cid);
+	if (IS_PF(p_hwfn->cdev))
+		rc = qed_eth_pf_rx_queue_stop(p_hwfn, p_cid,
+					      eq_completion_only,
+					      cqe_completion);
+	else
+		rc = qed_vf_pf_rxq_stop(p_hwfn, p_cid, cqe_completion);
+
+	if (!rc)
+		qed_eth_queue_cid_release(p_hwfn, p_cid);
+	return rc;
 }
 
-int qed_sp_eth_txq_start_ramrod(struct qed_hwfn  *p_hwfn,
-				u16  opaque_fid,
-				u32  cid,
-				struct qed_queue_start_common_params *p_params,
-				u8  stats_id,
-				dma_addr_t pbl_addr,
-				u16 pbl_size,
-				union qed_qm_pq_params *p_pq_params)
+int
+qed_eth_txq_start_ramrod(struct qed_hwfn *p_hwfn,
+			 struct qed_queue_cid *p_cid,
+			 dma_addr_t pbl_addr, u16 pbl_size, u16 pq_id)
 {
 	struct tx_queue_start_ramrod_data *p_ramrod = NULL;
 	struct qed_spq_entry *p_ent = NULL;
 	struct qed_sp_init_data init_data;
-	struct qed_hw_cid_data *p_tx_cid;
-	u16 pq_id, abs_tx_q_id = 0;
 	int rc = -EINVAL;
-	u8 abs_vport_id;
-
-	/* Store information for the stop */
-	p_tx_cid = &p_hwfn->p_tx_cids[p_params->queue_id];
-	p_tx_cid->cid		= cid;
-	p_tx_cid->opaque_fid	= opaque_fid;
-
-	rc = qed_fw_vport(p_hwfn, p_params->vport_id, &abs_vport_id);
-	if (rc)
-		return rc;
-
-	rc = qed_fw_l2_queue(p_hwfn, p_params->queue_id, &abs_tx_q_id);
-	if (rc)
-		return rc;
 
 	/* Get SPQ entry */
 	memset(&init_data, 0, sizeof(init_data));
-	init_data.cid = cid;
-	init_data.opaque_fid = opaque_fid;
+	init_data.cid = p_cid->cid;
+	init_data.opaque_fid = p_cid->opaque_fid;
 	init_data.comp_mode = QED_SPQ_MODE_EBLOCK;
 
 	rc = qed_sp_init_request(p_hwfn, &p_ent,
@@ -794,96 +861,92 @@ int qed_sp_eth_txq_start_ramrod(struct qed_hwfn  *p_hwfn,
 		return rc;
 
 	p_ramrod = &p_ent->ramrod.tx_queue_start;
-	p_ramrod->vport_id = abs_vport_id;
+	p_ramrod->vport_id = p_cid->abs.vport_id;
 
-	p_ramrod->sb_id = cpu_to_le16(p_params->sb);
-	p_ramrod->sb_index = p_params->sb_idx;
-	p_ramrod->stats_counter_id = stats_id;
+	p_ramrod->sb_id = cpu_to_le16(p_cid->abs.sb);
+	p_ramrod->sb_index = p_cid->abs.sb_idx;
+	p_ramrod->stats_counter_id = p_cid->abs.stats_id;
 
-	p_ramrod->queue_zone_id = cpu_to_le16(abs_tx_q_id);
+	p_ramrod->queue_zone_id = cpu_to_le16(p_cid->abs.queue_id);
+	p_ramrod->same_as_last_id = cpu_to_le16(p_cid->abs.queue_id);
 
 	p_ramrod->pbl_size = cpu_to_le16(pbl_size);
 	DMA_REGPAIR_LE(p_ramrod->pbl_base_addr, pbl_addr);
 
-	pq_id = qed_get_qm_pq(p_hwfn, PROTOCOLID_ETH, p_pq_params);
 	p_ramrod->qm_pq_id = cpu_to_le16(pq_id);
 
 	return qed_spq_post(p_hwfn, p_ent, NULL);
 }
 
 static int
-qed_sp_eth_tx_queue_start(struct qed_hwfn *p_hwfn,
-			  u16 opaque_fid,
-			  struct qed_queue_start_common_params *p_params,
+qed_eth_pf_tx_queue_start(struct qed_hwfn *p_hwfn,
+			  struct qed_queue_cid *p_cid,
+			  u8 tc,
 			  dma_addr_t pbl_addr,
 			  u16 pbl_size, void __iomem **pp_doorbell)
 {
-	struct qed_hw_cid_data *p_tx_cid;
 	union qed_qm_pq_params pq_params;
-	u8 abs_stats_id = 0;
 	int rc;
 
-	if (IS_VF(p_hwfn->cdev)) {
-		return qed_vf_pf_txq_start(p_hwfn,
-					   p_params->queue_id,
-					   p_params->sb,
-					   p_params->sb_idx,
-					   pbl_addr, pbl_size, pp_doorbell);
-	}
+	memset(&pq_params, 0, sizeof(pq_params));
 
-	rc = qed_fw_vport(p_hwfn, p_params->vport_id, &abs_stats_id);
+	rc = qed_eth_txq_start_ramrod(p_hwfn, p_cid,
+				      pbl_addr, pbl_size,
+				      qed_get_qm_pq(p_hwfn, PROTOCOLID_ETH,
+						    &pq_params));
 	if (rc)
 		return rc;
 
-	p_tx_cid = &p_hwfn->p_tx_cids[p_params->queue_id];
-	memset(p_tx_cid, 0, sizeof(*p_tx_cid));
-	memset(&pq_params, 0, sizeof(pq_params));
+	/* Provide the caller with the necessary return values */
+	*pp_doorbell = p_hwfn->doorbells +
+		       qed_db_addr(p_cid->cid, DQ_DEMS_LEGACY);
 
-	/* Allocate a CID for the queue */
-	rc = qed_cxt_acquire_cid(p_hwfn, PROTOCOLID_ETH, &p_tx_cid->cid);
-	if (rc) {
-		DP_NOTICE(p_hwfn, "Failed to acquire cid\n");
-		return rc;
-	}
-	p_tx_cid->b_cid_allocated = true;
+	return 0;
+}
 
-	DP_VERBOSE(p_hwfn, QED_MSG_SP,
-		   "opaque_fid=0x%x, cid=0x%x, tx_qid=0x%x, vport_id=0x%x, sb_id=0x%x\n",
-		   opaque_fid, p_tx_cid->cid,
-		   p_params->queue_id, p_params->vport_id, p_params->sb);
-
-	rc = qed_sp_eth_txq_start_ramrod(p_hwfn,
-					 opaque_fid,
-					 p_tx_cid->cid,
-					 p_params,
-					 abs_stats_id,
-					 pbl_addr,
-					 pbl_size,
-					 &pq_params);
-
-	*pp_doorbell = (u8 __iomem *)p_hwfn->doorbells +
-				     qed_db_addr(p_tx_cid->cid, DQ_DEMS_LEGACY);
+static int
+qed_eth_tx_queue_start(struct qed_hwfn *p_hwfn,
+		       u16 opaque_fid,
+		       struct qed_queue_start_common_params *p_params,
+		       u8 tc,
+		       dma_addr_t pbl_addr,
+		       u16 pbl_size,
+		       struct qed_txq_start_ret_params *p_ret_params)
+{
+	struct qed_queue_cid *p_cid;
+	int rc;
+
+	p_cid = qed_eth_queue_to_cid(p_hwfn, opaque_fid, p_params);
+	if (!p_cid)
+		return -EINVAL;
+
+	if (IS_PF(p_hwfn->cdev))
+		rc = qed_eth_pf_tx_queue_start(p_hwfn, p_cid, tc,
+					       pbl_addr, pbl_size,
+					       &p_ret_params->p_doorbell);
+	else
+		rc = qed_vf_pf_txq_start(p_hwfn, p_cid,
+					 pbl_addr, pbl_size,
+					 &p_ret_params->p_doorbell);
 
 	if (rc)
-		qed_sp_release_queue_cid(p_hwfn, p_tx_cid);
+		qed_eth_queue_cid_release(p_hwfn, p_cid);
+	else
+		p_ret_params->p_handle = (void *)p_cid;
 
 	return rc;
 }
 
-int qed_sp_eth_tx_queue_stop(struct qed_hwfn *p_hwfn, u16 tx_queue_id)
+static int
+qed_eth_pf_tx_queue_stop(struct qed_hwfn *p_hwfn, struct qed_queue_cid *p_cid)
 {
-	struct qed_hw_cid_data *p_tx_cid = &p_hwfn->p_tx_cids[tx_queue_id];
 	struct qed_spq_entry *p_ent = NULL;
 	struct qed_sp_init_data init_data;
-	int rc = -EINVAL;
-
-	if (IS_VF(p_hwfn->cdev))
-		return qed_vf_pf_txq_stop(p_hwfn, tx_queue_id);
+	int rc;
 
-	/* Get SPQ entry */
 	memset(&init_data, 0, sizeof(init_data));
-	init_data.cid = p_tx_cid->cid;
-	init_data.opaque_fid = p_tx_cid->opaque_fid;
+	init_data.cid = p_cid->cid;
+	init_data.opaque_fid = p_cid->opaque_fid;
 	init_data.comp_mode = QED_SPQ_MODE_EBLOCK;
 
 	rc = qed_sp_init_request(p_hwfn, &p_ent,
@@ -892,11 +955,22 @@ int qed_sp_eth_tx_queue_stop(struct qed_hwfn *p_hwfn, u16 tx_queue_id)
 	if (rc)
 		return rc;
 
-	rc = qed_spq_post(p_hwfn, p_ent, NULL);
-	if (rc)
-		return rc;
+	return qed_spq_post(p_hwfn, p_ent, NULL);
+}
+
+int qed_eth_tx_queue_stop(struct qed_hwfn *p_hwfn, void *p_handle)
+{
+	struct qed_queue_cid *p_cid = (struct qed_queue_cid *)p_handle;
+	int rc;
+
+	if (IS_PF(p_hwfn->cdev))
+		rc = qed_eth_pf_tx_queue_stop(p_hwfn, p_cid);
+	else
+		rc = qed_vf_pf_txq_stop(p_hwfn, p_cid);
 
-	return qed_sp_release_queue_cid(p_hwfn, p_tx_cid);
+	if (!rc)
+		qed_eth_queue_cid_release(p_hwfn, p_cid);
+	return rc;
 }
 
 static enum eth_filter_action qed_filter_action(enum qed_filter_opcode opcode)
@@ -1880,58 +1954,53 @@ static int qed_update_vport(struct qed_dev *cdev,
 }
 
 static int qed_start_rxq(struct qed_dev *cdev,
-			 struct qed_queue_start_common_params *params,
+			 u8 rss_num,
+			 struct qed_queue_start_common_params *p_params,
 			 u16 bd_max_bytes,
 			 dma_addr_t bd_chain_phys_addr,
 			 dma_addr_t cqe_pbl_addr,
 			 u16 cqe_pbl_size,
-			 void __iomem **pp_prod)
+			 struct qed_rxq_start_ret_params *ret_params)
 {
 	struct qed_hwfn *p_hwfn;
 	int rc, hwfn_index;
 
-	hwfn_index = params->rss_id % cdev->num_hwfns;
+	hwfn_index = rss_num % cdev->num_hwfns;
 	p_hwfn = &cdev->hwfns[hwfn_index];
 
-	/* Fix queue ID in 100g mode */
-	params->queue_id /= cdev->num_hwfns;
-
-	rc = qed_sp_eth_rx_queue_start(p_hwfn,
-				       p_hwfn->hw_info.opaque_fid,
-				       params,
-				       bd_max_bytes,
-				       bd_chain_phys_addr,
-				       cqe_pbl_addr,
-				       cqe_pbl_size,
-				       pp_prod);
+	p_params->queue_id = p_params->queue_id / cdev->num_hwfns;
+	p_params->stats_id = p_params->vport_id;
 
+	rc = qed_eth_rx_queue_start(p_hwfn,
+				    p_hwfn->hw_info.opaque_fid,
+				    p_params,
+				    bd_max_bytes,
+				    bd_chain_phys_addr,
+				    cqe_pbl_addr, cqe_pbl_size, ret_params);
 	if (rc) {
-		DP_ERR(cdev, "Failed to start RXQ#%d\n", params->queue_id);
+		DP_ERR(cdev, "Failed to start RXQ#%d\n", p_params->queue_id);
 		return rc;
 	}
 
 	DP_VERBOSE(cdev, (QED_MSG_SPQ | NETIF_MSG_IFUP),
-		   "Started RX-Q %d [rss %d] on V-PORT %d and SB %d\n",
-		   params->queue_id, params->rss_id, params->vport_id,
-		   params->sb);
+		   "Started RX-Q %d [rss_num %d] on V-PORT %d and SB %d\n",
+		   p_params->queue_id, rss_num, p_params->vport_id,
+		   p_params->sb);
 
 	return 0;
 }
 
-static int qed_stop_rxq(struct qed_dev *cdev,
-			struct qed_stop_rxq_params *params)
+static int qed_stop_rxq(struct qed_dev *cdev, u8 rss_id, void *handle)
 {
 	int rc, hwfn_index;
 	struct qed_hwfn *p_hwfn;
 
-	hwfn_index	= params->rss_id % cdev->num_hwfns;
-	p_hwfn		= &cdev->hwfns[hwfn_index];
+	hwfn_index = rss_id % cdev->num_hwfns;
+	p_hwfn = &cdev->hwfns[hwfn_index];
 
-	rc = qed_sp_eth_rx_queue_stop(p_hwfn,
-				      params->rx_queue_id / cdev->num_hwfns,
-				      params->eq_completion_only, false);
+	rc = qed_eth_rx_queue_stop(p_hwfn, handle, false, false);
 	if (rc) {
-		DP_ERR(cdev, "Failed to stop RXQ#%d\n", params->rx_queue_id);
+		DP_ERR(cdev, "Failed to stop RXQ#%02x\n", rss_id);
 		return rc;
 	}
 
@@ -1939,26 +2008,24 @@ static int qed_stop_rxq(struct qed_dev *cdev,
 }
 
 static int qed_start_txq(struct qed_dev *cdev,
+			 u8 rss_num,
 			 struct qed_queue_start_common_params *p_params,
 			 dma_addr_t pbl_addr,
 			 u16 pbl_size,
-			 void __iomem **pp_doorbell)
+			 struct qed_txq_start_ret_params *ret_params)
 {
 	struct qed_hwfn *p_hwfn;
 	int rc, hwfn_index;
 
-	hwfn_index	= p_params->rss_id % cdev->num_hwfns;
-	p_hwfn		= &cdev->hwfns[hwfn_index];
-
-	/* Fix queue ID in 100g mode */
-	p_params->queue_id /= cdev->num_hwfns;
+	hwfn_index = rss_num % cdev->num_hwfns;
+	p_hwfn = &cdev->hwfns[hwfn_index];
+	p_params->queue_id = p_params->queue_id / cdev->num_hwfns;
+	p_params->stats_id = p_params->vport_id;
 
-	rc = qed_sp_eth_tx_queue_start(p_hwfn,
-				       p_hwfn->hw_info.opaque_fid,
-				       p_params,
-				       pbl_addr,
-				       pbl_size,
-				       pp_doorbell);
+	rc = qed_eth_tx_queue_start(p_hwfn,
+				    p_hwfn->hw_info.opaque_fid,
+				    p_params, 0,
+				    pbl_addr, pbl_size, ret_params);
 
 	if (rc) {
 		DP_ERR(cdev, "Failed to start TXQ#%d\n", p_params->queue_id);
@@ -1966,8 +2033,8 @@ static int qed_start_txq(struct qed_dev *cdev,
 	}
 
 	DP_VERBOSE(cdev, (QED_MSG_SPQ | NETIF_MSG_IFUP),
-		   "Started TX-Q %d [rss %d] on V-PORT %d and SB %d\n",
-		   p_params->queue_id, p_params->rss_id, p_params->vport_id,
+		   "Started TX-Q %d [rss_num %d] on V-PORT %d and SB %d\n",
+		   p_params->queue_id, rss_num, p_params->vport_id,
 		   p_params->sb);
 
 	return 0;
@@ -1981,19 +2048,17 @@ static int qed_fastpath_stop(struct qed_dev *cdev)
 	return 0;
 }
 
-static int qed_stop_txq(struct qed_dev *cdev,
-			struct qed_stop_txq_params *params)
+static int qed_stop_txq(struct qed_dev *cdev, u8 rss_id, void *handle)
 {
 	struct qed_hwfn *p_hwfn;
 	int rc, hwfn_index;
 
-	hwfn_index	= params->rss_id % cdev->num_hwfns;
-	p_hwfn		= &cdev->hwfns[hwfn_index];
+	hwfn_index = rss_id % cdev->num_hwfns;
+	p_hwfn = &cdev->hwfns[hwfn_index];
 
-	rc = qed_sp_eth_tx_queue_stop(p_hwfn,
-				      params->tx_queue_id / cdev->num_hwfns);
+	rc = qed_eth_tx_queue_stop(p_hwfn, handle);
 	if (rc) {
-		DP_ERR(cdev, "Failed to stop TXQ#%d\n", params->tx_queue_id);
+		DP_ERR(cdev, "Failed to stop TXQ#%02x\n", rss_id);
 		return rc;
 	}
 
diff --git a/drivers/net/ethernet/qlogic/qed/qed_l2.h b/drivers/net/ethernet/qlogic/qed/qed_l2.h
index e495d62fcc03..48c9bfc28140 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_l2.h
+++ b/drivers/net/ethernet/qlogic/qed/qed_l2.h
@@ -78,11 +78,34 @@ struct qed_filter_mcast {
 	unsigned char mac[QED_MAX_MC_ADDRS][ETH_ALEN];
 };
 
-int qed_sp_eth_rx_queue_stop(struct qed_hwfn *p_hwfn,
-			     u16 rx_queue_id,
-			     bool eq_completion_only, bool cqe_completion);
+/**
+ * @brief qed_eth_rx_queue_stop - This ramrod closes an Rx queue
+ *
+ * @param p_hwfn
+ * @param p_rxq			Handler of queue to close
+ * @param eq_completion_only	If True completion will be on
+ *				EQe, if False completion will be
+ *				on EQe if p_hwfn opaque
+ *				different from the RXQ opaque
+ *				otherwise on CQe.
+ * @param cqe_completion	If True completion will be
+ *				receive on CQe.
+ * @return int
+ */
+int
+qed_eth_rx_queue_stop(struct qed_hwfn *p_hwfn,
+		      void *p_rxq,
+		      bool eq_completion_only, bool cqe_completion);
 
-int qed_sp_eth_tx_queue_stop(struct qed_hwfn *p_hwfn, u16 tx_queue_id);
+/**
+ * @brief qed_eth_tx_queue_stop - closes a Tx queue
+ *
+ * @param p_hwfn
+ * @param p_txq - handle to Tx queue needed to be closed
+ *
+ * @return int
+ */
+int qed_eth_tx_queue_stop(struct qed_hwfn *p_hwfn, void *p_txq);
 
 enum qed_tpa_mode {
 	QED_TPA_MODE_NONE,
@@ -196,19 +219,19 @@ int qed_sp_eth_filter_ucast(struct qed_hwfn *p_hwfn,
  * @note At the moment - only used by non-linux VFs.
  *
  * @param p_hwfn
- * @param rx_queue_id		RX Queue ID
- * @param num_rxqs		Allow to update multiple rx
- *				queues, from rx_queue_id to
- *				(rx_queue_id + num_rxqs)
+ * @param pp_rxq_handlers	An array of queue handlers to be updated.
+ * @param num_rxqs              number of queues to update.
  * @param complete_cqe_flg	Post completion to the CQE Ring if set
  * @param complete_event_flg	Post completion to the Event Ring if set
+ * @param comp_mode
+ * @param p_comp_data
  *
  * @return int
  */
 
 int
 qed_sp_eth_rx_queues_update(struct qed_hwfn *p_hwfn,
-			    u16 rx_queue_id,
+			    void **pp_rxq_handlers,
 			    u8 num_rxqs,
 			    u8 complete_cqe_flg,
 			    u8 complete_event_flg,
@@ -217,27 +240,79 @@ qed_sp_eth_rx_queues_update(struct qed_hwfn *p_hwfn,
 
 void qed_get_vport_stats(struct qed_dev *cdev, struct qed_eth_stats *stats);
 
-int qed_sp_eth_vport_start(struct qed_hwfn *p_hwfn,
-			   struct qed_sp_vport_start_params *p_params);
+void qed_reset_vport_stats(struct qed_dev *cdev);
+
+struct qed_queue_cid {
+	/* 'Relative' is a relative term ;-). Usually the indices [not counting
+	 * SBs] would be PF-relative, but there are some cases where that isn't
+	 * the case - specifically for a PF configuring its VF indices it's
+	 * possible some fields [E.g., stats-id] in 'rel' would already be abs.
+	 */
+	struct qed_queue_start_common_params rel;
+	struct qed_queue_start_common_params abs;
+	u32 cid;
+	u16 opaque_fid;
+
+	/* VFs queues are mapped differently, so we need to know the
+	 * relative queue associated with them [0-based].
+	 * Notice this is relevant on the *PF* queue-cid of its VF's queues,
+	 * and not on the VF itself.
+	 */
+	bool is_vf;
+	u8 vf_qid;
+
+	/* Legacy VFs might have Rx producer located elsewhere */
+	bool b_legacy_vf;
+};
 
-int qed_sp_eth_rxq_start_ramrod(struct qed_hwfn *p_hwfn,
-				u16 opaque_fid,
-				u32 cid,
-				struct qed_queue_start_common_params *params,
-				u8 stats_id,
-				u16 bd_max_bytes,
-				dma_addr_t bd_chain_phys_addr,
-				dma_addr_t cqe_pbl_addr,
-				u16 cqe_pbl_size, bool b_use_zone_a_prod);
-
-int qed_sp_eth_txq_start_ramrod(struct qed_hwfn  *p_hwfn,
-				u16  opaque_fid,
-				u32  cid,
-				struct qed_queue_start_common_params *p_params,
-				u8  stats_id,
-				dma_addr_t pbl_addr,
-				u16 pbl_size,
-				union qed_qm_pq_params *p_pq_params);
+void qed_eth_queue_cid_release(struct qed_hwfn *p_hwfn,
+			       struct qed_queue_cid *p_cid);
+
+struct qed_queue_cid *_qed_eth_queue_to_cid(struct qed_hwfn *p_hwfn,
+					    u16 opaque_fid,
+					    u32 cid,
+					    u8 vf_qid,
+					    struct qed_queue_start_common_params
+					    *p_params);
+
+int
+qed_sp_eth_vport_start(struct qed_hwfn *p_hwfn,
+		       struct qed_sp_vport_start_params *p_params);
+
+/**
+ * @brief - Starts an Rx queue, when queue_cid is already prepared
+ *
+ * @param p_hwfn
+ * @param p_cid
+ * @param bd_max_bytes
+ * @param bd_chain_phys_addr
+ * @param cqe_pbl_addr
+ * @param cqe_pbl_size
+ *
+ * @return int
+ */
+int
+qed_eth_rxq_start_ramrod(struct qed_hwfn *p_hwfn,
+			 struct qed_queue_cid *p_cid,
+			 u16 bd_max_bytes,
+			 dma_addr_t bd_chain_phys_addr,
+			 dma_addr_t cqe_pbl_addr, u16 cqe_pbl_size);
+
+/**
+ * @brief - Starts a Tx queue, where queue_cid is already prepared
+ *
+ * @param p_hwfn
+ * @param p_cid
+ * @param pbl_addr
+ * @param pbl_size
+ * @param p_pq_params - parameters for choosing the PQ for this Tx queue
+ *
+ * @return int
+ */
+int
+qed_eth_txq_start_ramrod(struct qed_hwfn *p_hwfn,
+			 struct qed_queue_cid *p_cid,
+			 dma_addr_t pbl_addr, u16 pbl_size, u16 pq_id);
 
 u8 qed_mcast_bin_from_mac(u8 *mac);
 
diff --git a/drivers/net/ethernet/qlogic/qed/qed_sp_commands.c b/drivers/net/ethernet/qlogic/qed/qed_sp_commands.c
index 2888eb0628f8..d0a58282f2a8 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_sp_commands.c
+++ b/drivers/net/ethernet/qlogic/qed/qed_sp_commands.c
@@ -347,11 +347,11 @@ int qed_sp_pf_start(struct qed_hwfn *p_hwfn,
 
 	/* Place EQ address in RAMROD */
 	DMA_REGPAIR_LE(p_ramrod->event_ring_pbl_addr,
-		       p_hwfn->p_eq->chain.pbl.p_phys_table);
+		       p_hwfn->p_eq->chain.pbl_sp.p_phys_table);
 	page_cnt = (u8)qed_chain_get_page_cnt(&p_hwfn->p_eq->chain);
 	p_ramrod->event_ring_num_pages = page_cnt;
 	DMA_REGPAIR_LE(p_ramrod->consolid_q_pbl_addr,
-		       p_hwfn->p_consq->chain.pbl.p_phys_table);
+		       p_hwfn->p_consq->chain.pbl_sp.p_phys_table);
 
 	qed_tunn_set_pf_start_params(p_hwfn, p_tunn, &p_ramrod->tunnel_config);
 
diff --git a/drivers/net/ethernet/qlogic/qed/qed_sriov.c b/drivers/net/ethernet/qlogic/qed/qed_sriov.c
index f3f742a4e59a..85b09dd1787a 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_sriov.c
+++ b/drivers/net/ethernet/qlogic/qed/qed_sriov.c
@@ -808,37 +808,70 @@ static void qed_iov_free_vf_igu_sbs(struct qed_hwfn *p_hwfn,
 
 static int qed_iov_init_hw_for_vf(struct qed_hwfn *p_hwfn,
 				  struct qed_ptt *p_ptt,
-				  u16 rel_vf_id, u16 num_rx_queues)
+				  struct qed_iov_vf_init_params *p_params)
 {
 	u8 num_of_vf_avaiable_chains = 0;
 	struct qed_vf_info *vf = NULL;
+	u16 qid, num_irqs;
 	int rc = 0;
 	u32 cids;
 	u8 i;
 
-	vf = qed_iov_get_vf_info(p_hwfn, rel_vf_id, false);
+	vf = qed_iov_get_vf_info(p_hwfn, p_params->rel_vf_id, false);
 	if (!vf) {
 		DP_ERR(p_hwfn, "qed_iov_init_hw_for_vf : vf is NULL\n");
 		return -EINVAL;
 	}
 
 	if (vf->b_init) {
-		DP_NOTICE(p_hwfn, "VF[%d] is already active.\n", rel_vf_id);
+		DP_NOTICE(p_hwfn, "VF[%d] is already active.\n",
+			  p_params->rel_vf_id);
 		return -EINVAL;
 	}
 
+	/* Perform sanity checking on the requested queue_id */
+	for (i = 0; i < p_params->num_queues; i++) {
+		u16 min_vf_qzone = FEAT_NUM(p_hwfn, QED_PF_L2_QUE);
+		u16 max_vf_qzone = min_vf_qzone +
+		    FEAT_NUM(p_hwfn, QED_VF_L2_QUE) - 1;
+
+		qid = p_params->req_rx_queue[i];
+		if (qid < min_vf_qzone || qid > max_vf_qzone) {
+			DP_NOTICE(p_hwfn,
+				  "Can't enable Rx qid [%04x] for VF[%d]: qids [0x%04x,...,0x%04x] available\n",
+				  qid,
+				  p_params->rel_vf_id,
+				  min_vf_qzone, max_vf_qzone);
+			return -EINVAL;
+		}
+
+		qid = p_params->req_tx_queue[i];
+		if (qid > max_vf_qzone) {
+			DP_NOTICE(p_hwfn,
+				  "Can't enable Tx qid [%04x] for VF[%d]: max qid 0x%04x\n",
+				  qid, p_params->rel_vf_id, max_vf_qzone);
+			return -EINVAL;
+		}
+
+		/* If client *really* wants, Tx qid can be shared with PF */
+		if (qid < min_vf_qzone)
+			DP_VERBOSE(p_hwfn,
+				   QED_MSG_IOV,
+				   "VF[%d] is using PF qid [0x%04x] for Txq[0x%02x]\n",
+				   p_params->rel_vf_id, qid, i);
+	}
+
 	/* Limit number of queues according to number of CIDs */
 	qed_cxt_get_proto_cid_count(p_hwfn, PROTOCOLID_ETH, &cids);
 	DP_VERBOSE(p_hwfn,
 		   QED_MSG_IOV,
 		   "VF[%d] - requesting to initialize for 0x%04x queues [0x%04x CIDs available]\n",
-		   vf->relative_vf_id, num_rx_queues, (u16) cids);
-	num_rx_queues = min_t(u16, num_rx_queues, ((u16) cids));
+		   vf->relative_vf_id, p_params->num_queues, (u16)cids);
+	num_irqs = min_t(u16, p_params->num_queues, ((u16)cids));
 
 	num_of_vf_avaiable_chains = qed_iov_alloc_vf_igu_sbs(p_hwfn,
 							     p_ptt,
-							     vf,
-							     num_rx_queues);
+							     vf, num_irqs);
 	if (!num_of_vf_avaiable_chains) {
 		DP_ERR(p_hwfn, "no available igu sbs\n");
 		return -ENOMEM;
@@ -849,25 +882,22 @@ static int qed_iov_init_hw_for_vf(struct qed_hwfn *p_hwfn,
 	vf->num_txqs = num_of_vf_avaiable_chains;
 
 	for (i = 0; i < vf->num_rxqs; i++) {
-		u16 queue_id = qed_int_queue_id_from_sb_id(p_hwfn,
-							   vf->igu_sbs[i]);
+		struct qed_vf_q_info *p_queue = &vf->vf_queues[i];
 
-		if (queue_id > RESC_NUM(p_hwfn, QED_L2_QUEUE)) {
-			DP_NOTICE(p_hwfn,
-				  "VF[%d] will require utilizing of out-of-bounds queues - %04x\n",
-				  vf->relative_vf_id, queue_id);
-			return -EINVAL;
-		}
+		p_queue->fw_rx_qid = p_params->req_rx_queue[i];
+		p_queue->fw_tx_qid = p_params->req_tx_queue[i];
 
 		/* CIDs are per-VF, so no problem having them 0-based. */
-		vf->vf_queues[i].fw_rx_qid = queue_id;
-		vf->vf_queues[i].fw_tx_qid = queue_id;
-		vf->vf_queues[i].fw_cid = i;
+		p_queue->fw_cid = i;
 
 		DP_VERBOSE(p_hwfn, QED_MSG_IOV,
-			   "VF[%d] - [%d] SB %04x, Tx/Rx queue %04x CID %04x\n",
-			   vf->relative_vf_id, i, vf->igu_sbs[i], queue_id, i);
+			   "VF[%d] - Q[%d] SB %04x, qid [Rx %04x Tx %04x]  CID %04x\n",
+			   vf->relative_vf_id,
+			   i, vf->igu_sbs[i],
+			   p_queue->fw_rx_qid,
+			   p_queue->fw_tx_qid, p_queue->fw_cid);
 	}
+
 	rc = qed_iov_enable_vf_access(p_hwfn, p_ptt, vf);
 	if (!rc) {
 		vf->b_init = true;
@@ -1187,8 +1217,19 @@ static void qed_iov_vf_cleanup(struct qed_hwfn *p_hwfn,
 
 	p_vf->num_active_rxqs = 0;
 
-	for (i = 0; i < QED_MAX_VF_CHAINS_PER_PF; i++)
-		p_vf->vf_queues[i].rxq_active = 0;
+	for (i = 0; i < QED_MAX_VF_CHAINS_PER_PF; i++) {
+		struct qed_vf_q_info *p_queue = &p_vf->vf_queues[i];
+
+		if (p_queue->p_rx_cid) {
+			qed_eth_queue_cid_release(p_hwfn, p_queue->p_rx_cid);
+			p_queue->p_rx_cid = NULL;
+		}
+
+		if (p_queue->p_tx_cid) {
+			qed_eth_queue_cid_release(p_hwfn, p_queue->p_tx_cid);
+			p_queue->p_tx_cid = NULL;
+		}
+	}
 
 	memset(&p_vf->shadow_config, 0, sizeof(p_vf->shadow_config));
 	memset(&p_vf->acquire, 0, sizeof(p_vf->acquire));
@@ -1594,21 +1635,21 @@ static int qed_iov_configure_vport_forced(struct qed_hwfn *p_hwfn,
 
 		/* Update all the Rx queues */
 		for (i = 0; i < QED_MAX_VF_CHAINS_PER_PF; i++) {
-			u16 qid;
+			struct qed_queue_cid *p_cid;
 
-			if (!p_vf->vf_queues[i].rxq_active)
+			p_cid = p_vf->vf_queues[i].p_rx_cid;
+			if (!p_cid)
 				continue;
 
-			qid = p_vf->vf_queues[i].fw_rx_qid;
-
-			rc = qed_sp_eth_rx_queues_update(p_hwfn, qid,
+			rc = qed_sp_eth_rx_queues_update(p_hwfn,
+							 (void **)&p_cid,
 							 1, 0, 1,
 							 QED_SPQ_MODE_EBLOCK,
 							 NULL);
 			if (rc) {
 				DP_NOTICE(p_hwfn,
 					  "Failed to send Rx update fo queue[0x%04x]\n",
-					  qid);
+					  p_cid->rel.queue_id);
 				return rc;
 			}
 		}
@@ -1782,23 +1823,34 @@ static void qed_iov_vf_mbx_start_rxq(struct qed_hwfn *p_hwfn,
 	struct qed_queue_start_common_params params;
 	struct qed_iov_vf_mbx *mbx = &vf->vf_mbx;
 	u8 status = PFVF_STATUS_NO_RESOURCE;
+	struct qed_vf_q_info *p_queue;
 	struct vfpf_start_rxq_tlv *req;
 	bool b_legacy_vf = false;
 	int rc;
 
-	memset(&params, 0, sizeof(params));
 	req = &mbx->req_virt->start_rxq;
 
 	if (!qed_iov_validate_rxq(p_hwfn, vf, req->rx_qid) ||
 	    !qed_iov_validate_sb(p_hwfn, vf, req->hw_sb))
 		goto out;
 
-	params.queue_id =  vf->vf_queues[req->rx_qid].fw_rx_qid;
-	params.vf_qid = req->rx_qid;
+	/* Acquire a new queue-cid */
+	p_queue = &vf->vf_queues[req->rx_qid];
+
+	memset(&params, 0, sizeof(params));
+	params.queue_id = p_queue->fw_rx_qid;
 	params.vport_id = vf->vport_id;
+	params.stats_id = vf->abs_vf_id + 0x10;
 	params.sb = req->hw_sb;
 	params.sb_idx = req->sb_index;
 
+	p_queue->p_rx_cid = _qed_eth_queue_to_cid(p_hwfn,
+						  vf->opaque_fid,
+						  p_queue->fw_cid,
+						  req->rx_qid, &params);
+	if (!p_queue->p_rx_cid)
+		goto out;
+
 	/* Legacy VFs have their Producers in a different location, which they
 	 * calculate on their own and clean the producer prior to this.
 	 */
@@ -1811,21 +1863,19 @@ static void qed_iov_vf_mbx_start_rxq(struct qed_hwfn *p_hwfn,
 		       MSTORM_ETH_VF_PRODS_OFFSET(vf->abs_vf_id, req->rx_qid),
 		       0);
 	}
+	p_queue->p_rx_cid->b_legacy_vf = b_legacy_vf;
 
-	rc = qed_sp_eth_rxq_start_ramrod(p_hwfn, vf->opaque_fid,
-					 vf->vf_queues[req->rx_qid].fw_cid,
-					 &params,
-					 vf->abs_vf_id + 0x10,
-					 req->bd_max_bytes,
-					 req->rxq_addr,
-					 req->cqe_pbl_addr, req->cqe_pbl_size,
-					 b_legacy_vf);
-
+	rc = qed_eth_rxq_start_ramrod(p_hwfn,
+				      p_queue->p_rx_cid,
+				      req->bd_max_bytes,
+				      req->rxq_addr,
+				      req->cqe_pbl_addr, req->cqe_pbl_size);
 	if (rc) {
 		status = PFVF_STATUS_FAILURE;
+		qed_eth_queue_cid_release(p_hwfn, p_queue->p_rx_cid);
+		p_queue->p_rx_cid = NULL;
 	} else {
 		status = PFVF_STATUS_SUCCESS;
-		vf->vf_queues[req->rx_qid].rxq_active = true;
 		vf->num_active_rxqs++;
 	}
 
@@ -1882,7 +1932,9 @@ static void qed_iov_vf_mbx_start_txq(struct qed_hwfn *p_hwfn,
 	u8 status = PFVF_STATUS_NO_RESOURCE;
 	union qed_qm_pq_params pq_params;
 	struct vfpf_start_txq_tlv *req;
+	struct qed_vf_q_info *p_queue;
 	int rc;
+	u16 pq;
 
 	/* Prepare the parameters which would choose the right PQ */
 	memset(&pq_params, 0, sizeof(pq_params));
@@ -1896,24 +1948,31 @@ static void qed_iov_vf_mbx_start_txq(struct qed_hwfn *p_hwfn,
 	    !qed_iov_validate_sb(p_hwfn, vf, req->hw_sb))
 		goto out;
 
-	params.queue_id =  vf->vf_queues[req->tx_qid].fw_tx_qid;
+	/* Acquire a new queue-cid */
+	p_queue = &vf->vf_queues[req->tx_qid];
+
+	params.queue_id = p_queue->fw_tx_qid;
 	params.vport_id = vf->vport_id;
+	params.stats_id = vf->abs_vf_id + 0x10;
 	params.sb = req->hw_sb;
 	params.sb_idx = req->sb_index;
 
-	rc = qed_sp_eth_txq_start_ramrod(p_hwfn,
-					 vf->opaque_fid,
-					 vf->vf_queues[req->tx_qid].fw_cid,
-					 &params,
-					 vf->abs_vf_id + 0x10,
-					 req->pbl_addr,
-					 req->pbl_size, &pq_params);
+	p_queue->p_tx_cid = _qed_eth_queue_to_cid(p_hwfn,
+						  vf->opaque_fid,
+						  p_queue->fw_cid,
+						  req->tx_qid, &params);
+	if (!p_queue->p_tx_cid)
+		goto out;
 
+	pq = qed_get_qm_pq(p_hwfn, PROTOCOLID_ETH, &pq_params);
+	rc = qed_eth_txq_start_ramrod(p_hwfn, p_queue->p_tx_cid,
+				      req->pbl_addr, req->pbl_size, pq);
 	if (rc) {
 		status = PFVF_STATUS_FAILURE;
+		qed_eth_queue_cid_release(p_hwfn, p_queue->p_tx_cid);
+		p_queue->p_tx_cid = NULL;
 	} else {
 		status = PFVF_STATUS_SUCCESS;
-		vf->vf_queues[req->tx_qid].txq_active = true;
 	}
 
 out:
@@ -1924,6 +1983,7 @@ static int qed_iov_vf_stop_rxqs(struct qed_hwfn *p_hwfn,
 				struct qed_vf_info *vf,
 				u16 rxq_id, u8 num_rxqs, bool cqe_completion)
 {
+	struct qed_vf_q_info *p_queue;
 	int rc = 0;
 	int qid;
 
@@ -1931,16 +1991,18 @@ static int qed_iov_vf_stop_rxqs(struct qed_hwfn *p_hwfn,
 		return -EINVAL;
 
 	for (qid = rxq_id; qid < rxq_id + num_rxqs; qid++) {
-		if (vf->vf_queues[qid].rxq_active) {
-			rc = qed_sp_eth_rx_queue_stop(p_hwfn,
-						      vf->vf_queues[qid].
-						      fw_rx_qid, false,
-						      cqe_completion);
+		p_queue = &vf->vf_queues[qid];
 
-			if (rc)
-				return rc;
-		}
-		vf->vf_queues[qid].rxq_active = false;
+		if (!p_queue->p_rx_cid)
+			continue;
+
+		rc = qed_eth_rx_queue_stop(p_hwfn,
+					   p_queue->p_rx_cid,
+					   false, cqe_completion);
+		if (rc)
+			return rc;
+
+		vf->vf_queues[qid].p_rx_cid = NULL;
 		vf->num_active_rxqs--;
 	}
 
@@ -1951,22 +2013,24 @@ static int qed_iov_vf_stop_txqs(struct qed_hwfn *p_hwfn,
 				struct qed_vf_info *vf, u16 txq_id, u8 num_txqs)
 {
 	int rc = 0;
+	struct qed_vf_q_info *p_queue;
 	int qid;
 
 	if (txq_id + num_txqs > ARRAY_SIZE(vf->vf_queues))
 		return -EINVAL;
 
 	for (qid = txq_id; qid < txq_id + num_txqs; qid++) {
-		if (vf->vf_queues[qid].txq_active) {
-			rc = qed_sp_eth_tx_queue_stop(p_hwfn,
-						      vf->vf_queues[qid].
-						      fw_tx_qid);
+		p_queue = &vf->vf_queues[qid];
+		if (!p_queue->p_tx_cid)
+			continue;
 
-			if (rc)
-				return rc;
-		}
-		vf->vf_queues[qid].txq_active = false;
+		rc = qed_eth_tx_queue_stop(p_hwfn, p_queue->p_tx_cid);
+		if (rc)
+			return rc;
+
+		p_queue->p_tx_cid = NULL;
 	}
+
 	return rc;
 }
 
@@ -2021,10 +2085,11 @@ static void qed_iov_vf_mbx_update_rxqs(struct qed_hwfn *p_hwfn,
 				       struct qed_ptt *p_ptt,
 				       struct qed_vf_info *vf)
 {
+	struct qed_queue_cid *handlers[QED_MAX_VF_CHAINS_PER_PF];
 	u16 length = sizeof(struct pfvf_def_resp_tlv);
 	struct qed_iov_vf_mbx *mbx = &vf->vf_mbx;
 	struct vfpf_update_rxq_tlv *req;
-	u8 status = PFVF_STATUS_SUCCESS;
+	u8 status = PFVF_STATUS_FAILURE;
 	u8 complete_event_flg;
 	u8 complete_cqe_flg;
 	u16 qid;
@@ -2035,29 +2100,36 @@ static void qed_iov_vf_mbx_update_rxqs(struct qed_hwfn *p_hwfn,
 	complete_cqe_flg = !!(req->flags & VFPF_RXQ_UPD_COMPLETE_CQE_FLAG);
 	complete_event_flg = !!(req->flags & VFPF_RXQ_UPD_COMPLETE_EVENT_FLAG);
 
+	/* Validate inputs */
+	if (req->num_rxqs + req->rx_qid > QED_MAX_VF_CHAINS_PER_PF ||
+	    !qed_iov_validate_rxq(p_hwfn, vf, req->rx_qid)) {
+		DP_INFO(p_hwfn, "VF[%d]: Incorrect Rxqs [%04x, %02x]\n",
+			vf->relative_vf_id, req->rx_qid, req->num_rxqs);
+		goto out;
+	}
+
 	for (i = 0; i < req->num_rxqs; i++) {
 		qid = req->rx_qid + i;
-
-		if (!vf->vf_queues[qid].rxq_active) {
-			DP_NOTICE(p_hwfn, "VF rx_qid = %d isn`t active!\n",
-				  qid);
-			status = PFVF_STATUS_FAILURE;
-			break;
+		if (!vf->vf_queues[qid].p_rx_cid) {
+			DP_INFO(p_hwfn,
+				"VF[%d] rx_qid = %d isn`t active!\n",
+				vf->relative_vf_id, qid);
+			goto out;
 		}
 
-		rc = qed_sp_eth_rx_queues_update(p_hwfn,
-						 vf->vf_queues[qid].fw_rx_qid,
-						 1,
-						 complete_cqe_flg,
-						 complete_event_flg,
-						 QED_SPQ_MODE_EBLOCK, NULL);
-
-		if (rc) {
-			status = PFVF_STATUS_FAILURE;
-			break;
-		}
+		handlers[i] = vf->vf_queues[qid].p_rx_cid;
 	}
 
+	rc = qed_sp_eth_rx_queues_update(p_hwfn, (void **)&handlers,
+					 req->num_rxqs,
+					 complete_cqe_flg,
+					 complete_event_flg,
+					 QED_SPQ_MODE_EBLOCK, NULL);
+	if (rc)
+		goto out;
+
+	status = PFVF_STATUS_SUCCESS;
+out:
 	qed_iov_prepare_resp(p_hwfn, p_ptt, vf, CHANNEL_TLV_UPDATE_RXQ,
 			     length, status);
 }
@@ -2268,7 +2340,7 @@ qed_iov_vp_update_rss_param(struct qed_hwfn *p_hwfn,
 			DP_NOTICE(p_hwfn,
 				  "rss_ind_table[%d] = %d, rxq is out of range\n",
 				  i, q_idx);
-		else if (!vf->vf_queues[q_idx].rxq_active)
+		else if (!vf->vf_queues[q_idx].p_rx_cid)
 			DP_NOTICE(p_hwfn,
 				  "rss_ind_table[%d] = %d, rxq is not active\n",
 				  i, q_idx);
@@ -3468,8 +3540,28 @@ int qed_sriov_disable(struct qed_dev *cdev, bool pci_enabled)
 	return 0;
 }
 
+static void qed_sriov_enable_qid_config(struct qed_hwfn *hwfn,
+					u16 vfid,
+					struct qed_iov_vf_init_params *params)
+{
+	u16 base, i;
+
+	/* Since we have an equal resource distribution per-VF, and we assume
+	 * PF has acquired the QED_PF_L2_QUE first queues, we start setting
+	 * sequentially from there.
+	 */
+	base = FEAT_NUM(hwfn, QED_PF_L2_QUE) + vfid * params->num_queues;
+
+	params->rel_vf_id = vfid;
+	for (i = 0; i < params->num_queues; i++) {
+		params->req_rx_queue[i] = base + i;
+		params->req_tx_queue[i] = base + i;
+	}
+}
+
 static int qed_sriov_enable(struct qed_dev *cdev, int num)
 {
+	struct qed_iov_vf_init_params params;
 	int i, j, rc;
 
 	if (num >= RESC_NUM(&cdev->hwfns[0], QED_VPORT)) {
@@ -3478,15 +3570,17 @@ static int qed_sriov_enable(struct qed_dev *cdev, int num)
 		return -EINVAL;
 	}
 
+	memset(&params, 0, sizeof(params));
+
 	/* Initialize HW for VF access */
 	for_each_hwfn(cdev, j) {
 		struct qed_hwfn *hwfn = &cdev->hwfns[j];
 		struct qed_ptt *ptt = qed_ptt_acquire(hwfn);
-		int num_queues;
 
 		/* Make sure not to use more than 16 queues per VF */
-		num_queues = min_t(int,
-				   FEAT_NUM(hwfn, QED_VF_L2_QUE) / num, 16);
+		params.num_queues = min_t(int,
+					  FEAT_NUM(hwfn, QED_VF_L2_QUE) / num,
+					  16);
 
 		if (!ptt) {
 			DP_ERR(hwfn, "Failed to acquire ptt\n");
@@ -3498,7 +3592,8 @@ static int qed_sriov_enable(struct qed_dev *cdev, int num)
 			if (!qed_iov_is_valid_vfid(hwfn, i, false, true))
 				continue;
 
-			rc = qed_iov_init_hw_for_vf(hwfn, ptt, i, num_queues);
+			qed_sriov_enable_qid_config(hwfn, i, &params);
+			rc = qed_iov_init_hw_for_vf(hwfn, ptt, &params);
 			if (rc) {
 				DP_ERR(cdev, "Failed to enable VF[%d]\n", i);
 				qed_ptt_release(hwfn, ptt);
diff --git a/drivers/net/ethernet/qlogic/qed/qed_sriov.h b/drivers/net/ethernet/qlogic/qed/qed_sriov.h
index 3cf515b1b427..509c02b4772e 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_sriov.h
+++ b/drivers/net/ethernet/qlogic/qed/qed_sriov.h
@@ -58,6 +58,23 @@ struct qed_public_vf_info {
 	int tx_rate;
 };
 
+struct qed_iov_vf_init_params {
+	u16 rel_vf_id;
+
+	/* Number of requested Queues; Currently, don't support different
+	 * number of Rx/Tx queues.
+	 */
+
+	u16 num_queues;
+
+	/* Allow the client to choose which qzones to use for Rx/Tx,
+	 * and which queue_base to use for Tx queues on a per-queue basis.
+	 * Notice values should be relative to the PF resources.
+	 */
+	u16 req_rx_queue[QED_MAX_VF_CHAINS_PER_PF];
+	u16 req_tx_queue[QED_MAX_VF_CHAINS_PER_PF];
+};
+
 /* This struct is part of qed_dev and contains data relevant to all hwfns;
  * Initialized only if SR-IOV cpabability is exposed in PCIe config space.
  */
@@ -99,10 +116,10 @@ struct qed_iov_vf_mbx {
 
 struct qed_vf_q_info {
 	u16 fw_rx_qid;
+	struct qed_queue_cid *p_rx_cid;
 	u16 fw_tx_qid;
+	struct qed_queue_cid *p_tx_cid;
 	u8 fw_cid;
-	u8 rxq_active;
-	u8 txq_active;
 };
 
 enum vf_state {
diff --git a/drivers/net/ethernet/qlogic/qed/qed_vf.c b/drivers/net/ethernet/qlogic/qed/qed_vf.c
index 3c0633642f4c..60b31a8ede73 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_vf.c
+++ b/drivers/net/ethernet/qlogic/qed/qed_vf.c
@@ -388,18 +388,18 @@ free_p_iov:
 #define MSTORM_QZONE_START(dev)   (TSTORM_QZONE_START +	\
 				   (TSTORM_QZONE_SIZE * NUM_OF_L2_QUEUES(dev)))
 
-int qed_vf_pf_rxq_start(struct qed_hwfn *p_hwfn,
-			u8 rx_qid,
-			u16 sb,
-			u8 sb_index,
-			u16 bd_max_bytes,
-			dma_addr_t bd_chain_phys_addr,
-			dma_addr_t cqe_pbl_addr,
-			u16 cqe_pbl_size, void __iomem **pp_prod)
+int
+qed_vf_pf_rxq_start(struct qed_hwfn *p_hwfn,
+		    struct qed_queue_cid *p_cid,
+		    u16 bd_max_bytes,
+		    dma_addr_t bd_chain_phys_addr,
+		    dma_addr_t cqe_pbl_addr,
+		    u16 cqe_pbl_size, void __iomem **pp_prod)
 {
 	struct qed_vf_iov *p_iov = p_hwfn->vf_iov_info;
 	struct pfvf_start_queue_resp_tlv *resp;
 	struct vfpf_start_rxq_tlv *req;
+	u8 rx_qid = p_cid->rel.queue_id;
 	int rc;
 
 	/* clear mailbox and prep first tlv */
@@ -409,21 +409,22 @@ int qed_vf_pf_rxq_start(struct qed_hwfn *p_hwfn,
 	req->cqe_pbl_addr = cqe_pbl_addr;
 	req->cqe_pbl_size = cqe_pbl_size;
 	req->rxq_addr = bd_chain_phys_addr;
-	req->hw_sb = sb;
-	req->sb_index = sb_index;
+	req->hw_sb = p_cid->rel.sb;
+	req->sb_index = p_cid->rel.sb_idx;
 	req->bd_max_bytes = bd_max_bytes;
 	req->stat_id = -1;
 
 	/* If PF is legacy, we'll need to calculate producers ourselves
 	 * as well as clean them.
 	 */
-	if (pp_prod && p_iov->b_pre_fp_hsi) {
+	if (p_iov->b_pre_fp_hsi) {
 		u8 hw_qid = p_iov->acquire_resp.resc.hw_qid[rx_qid];
 		u32 init_prod_val = 0;
 
-		*pp_prod = (u8 __iomem *)p_hwfn->regview +
-					 MSTORM_QZONE_START(p_hwfn->cdev) +
-					 hw_qid * MSTORM_QZONE_SIZE;
+		*pp_prod = (u8 __iomem *)
+		    p_hwfn->regview +
+		    MSTORM_QZONE_START(p_hwfn->cdev) +
+		    hw_qid * MSTORM_QZONE_SIZE;
 
 		/* Init the rcq, rx bd and rx sge (if valid) producers to 0 */
 		__internal_ram_wr(p_hwfn, *pp_prod, sizeof(u32),
@@ -444,7 +445,7 @@ int qed_vf_pf_rxq_start(struct qed_hwfn *p_hwfn,
 	}
 
 	/* Learn the address of the producer from the response */
-	if (pp_prod && !p_iov->b_pre_fp_hsi) {
+	if (!p_iov->b_pre_fp_hsi) {
 		u32 init_prod_val = 0;
 
 		*pp_prod = (u8 __iomem *)p_hwfn->regview + resp->offset;
@@ -462,7 +463,8 @@ exit:
 	return rc;
 }
 
-int qed_vf_pf_rxq_stop(struct qed_hwfn *p_hwfn, u16 rx_qid, bool cqe_completion)
+int qed_vf_pf_rxq_stop(struct qed_hwfn *p_hwfn,
+		       struct qed_queue_cid *p_cid, bool cqe_completion)
 {
 	struct qed_vf_iov *p_iov = p_hwfn->vf_iov_info;
 	struct vfpf_stop_rxqs_tlv *req;
@@ -472,7 +474,7 @@ int qed_vf_pf_rxq_stop(struct qed_hwfn *p_hwfn, u16 rx_qid, bool cqe_completion)
 	/* clear mailbox and prep first tlv */
 	req = qed_vf_pf_prep(p_hwfn, CHANNEL_TLV_STOP_RXQS, sizeof(*req));
 
-	req->rx_qid = rx_qid;
+	req->rx_qid = p_cid->rel.queue_id;
 	req->num_rxqs = 1;
 	req->cqe_completion = cqe_completion;
 
@@ -496,28 +498,28 @@ exit:
 	return rc;
 }
 
-int qed_vf_pf_txq_start(struct qed_hwfn *p_hwfn,
-			u16 tx_queue_id,
-			u16 sb,
-			u8 sb_index,
-			dma_addr_t pbl_addr,
-			u16 pbl_size, void __iomem **pp_doorbell)
+int
+qed_vf_pf_txq_start(struct qed_hwfn *p_hwfn,
+		    struct qed_queue_cid *p_cid,
+		    dma_addr_t pbl_addr,
+		    u16 pbl_size, void __iomem **pp_doorbell)
 {
 	struct qed_vf_iov *p_iov = p_hwfn->vf_iov_info;
 	struct pfvf_start_queue_resp_tlv *resp;
 	struct vfpf_start_txq_tlv *req;
+	u16 qid = p_cid->rel.queue_id;
 	int rc;
 
 	/* clear mailbox and prep first tlv */
 	req = qed_vf_pf_prep(p_hwfn, CHANNEL_TLV_START_TXQ, sizeof(*req));
 
-	req->tx_qid = tx_queue_id;
+	req->tx_qid = qid;
 
 	/* Tx */
 	req->pbl_addr = pbl_addr;
 	req->pbl_size = pbl_size;
-	req->hw_sb = sb;
-	req->sb_index = sb_index;
+	req->hw_sb = p_cid->rel.sb;
+	req->sb_index = p_cid->rel.sb_idx;
 
 	/* add list termination tlv */
 	qed_add_tlv(p_hwfn, &p_iov->offset,
@@ -533,33 +535,29 @@ int qed_vf_pf_txq_start(struct qed_hwfn *p_hwfn,
 		goto exit;
 	}
 
-	if (pp_doorbell) {
-		/* Modern PFs provide the actual offsets, while legacy
-		 * provided only the queue id.
-		 */
-		if (!p_iov->b_pre_fp_hsi) {
-			*pp_doorbell = (u8 __iomem *)p_hwfn->doorbells +
-						     resp->offset;
-		} else {
-			u8 cid = p_iov->acquire_resp.resc.cid[tx_queue_id];
-			u32 db_addr;
-
-			db_addr = qed_db_addr_vf(cid, DQ_DEMS_LEGACY);
-			*pp_doorbell = (u8 __iomem *)p_hwfn->doorbells +
-						     db_addr;
-		}
+	/* Modern PFs provide the actual offsets, while legacy
+	 * provided only the queue id.
+	 */
+	if (!p_iov->b_pre_fp_hsi) {
+		*pp_doorbell = (u8 __iomem *)p_hwfn->doorbells + resp->offset;
+	} else {
+		u8 cid = p_iov->acquire_resp.resc.cid[qid];
 
-		DP_VERBOSE(p_hwfn, QED_MSG_IOV,
-			   "Txq[0x%02x]: doorbell at %p [offset 0x%08x]\n",
-			   tx_queue_id, *pp_doorbell, resp->offset);
+		*pp_doorbell = (u8 __iomem *)p_hwfn->doorbells +
+					     qed_db_addr_vf(cid,
+							    DQ_DEMS_LEGACY);
 	}
+
+	DP_VERBOSE(p_hwfn, QED_MSG_IOV,
+		   "Txq[0x%02x]: doorbell at %p [offset 0x%08x]\n",
+		   qid, *pp_doorbell, resp->offset);
 exit:
 	qed_vf_pf_req_end(p_hwfn, rc);
 
 	return rc;
 }
 
-int qed_vf_pf_txq_stop(struct qed_hwfn *p_hwfn, u16 tx_qid)
+int qed_vf_pf_txq_stop(struct qed_hwfn *p_hwfn, struct qed_queue_cid *p_cid)
 {
 	struct qed_vf_iov *p_iov = p_hwfn->vf_iov_info;
 	struct vfpf_stop_txqs_tlv *req;
@@ -569,7 +567,7 @@ int qed_vf_pf_txq_stop(struct qed_hwfn *p_hwfn, u16 tx_qid)
 	/* clear mailbox and prep first tlv */
 	req = qed_vf_pf_prep(p_hwfn, CHANNEL_TLV_STOP_TXQS, sizeof(*req));
 
-	req->tx_qid = tx_qid;
+	req->tx_qid = p_cid->rel.queue_id;
 	req->num_txqs = 1;
 
 	/* add list termination tlv */
diff --git a/drivers/net/ethernet/qlogic/qed/qed_vf.h b/drivers/net/ethernet/qlogic/qed/qed_vf.h
index 325c250d4ee5..11eb3854e6f2 100644
--- a/drivers/net/ethernet/qlogic/qed/qed_vf.h
+++ b/drivers/net/ethernet/qlogic/qed/qed_vf.h
@@ -666,10 +666,7 @@ int qed_vf_hw_prepare(struct qed_hwfn *p_hwfn);
 /**
  * @brief VF - start the RX Queue by sending a message to the PF
  * @param p_hwfn
- * @param cid                   - zero based within the VF
- * @param rx_queue_id           - zero based within the VF
- * @param sb                    - VF status block for this queue
- * @param sb_index              - Index within the status block
+ * @param p_cid			- Only relative fields are relevant
  * @param bd_max_bytes          - maximum number of bytes per bd
  * @param bd_chain_phys_addr    - physical address of bd chain
  * @param cqe_pbl_addr          - physical address of pbl
@@ -680,9 +677,7 @@ int qed_vf_hw_prepare(struct qed_hwfn *p_hwfn);
  * @return int
  */
 int qed_vf_pf_rxq_start(struct qed_hwfn *p_hwfn,
-			u8 rx_queue_id,
-			u16 sb,
-			u8 sb_index,
+			struct qed_queue_cid *p_cid,
 			u16 bd_max_bytes,
 			dma_addr_t bd_chain_phys_addr,
 			dma_addr_t cqe_pbl_addr,
@@ -702,24 +697,23 @@ int qed_vf_pf_rxq_start(struct qed_hwfn *p_hwfn,
  *
  * @return int
  */
-int qed_vf_pf_txq_start(struct qed_hwfn *p_hwfn,
-			u16 tx_queue_id,
-			u16 sb,
-			u8 sb_index,
-			dma_addr_t pbl_addr,
-			u16 pbl_size, void __iomem **pp_doorbell);
+int
+qed_vf_pf_txq_start(struct qed_hwfn *p_hwfn,
+		    struct qed_queue_cid *p_cid,
+		    dma_addr_t pbl_addr,
+		    u16 pbl_size, void __iomem **pp_doorbell);
 
 /**
  * @brief VF - stop the RX queue by sending a message to the PF
  *
  * @param p_hwfn
- * @param rx_qid
+ * @param p_cid
  * @param cqe_completion
  *
  * @return int
  */
 int qed_vf_pf_rxq_stop(struct qed_hwfn *p_hwfn,
-		       u16 rx_qid, bool cqe_completion);
+		       struct qed_queue_cid *p_cid, bool cqe_completion);
 
 /**
  * @brief VF - stop the TX queue by sending a message to the PF
@@ -729,7 +723,7 @@ int qed_vf_pf_rxq_stop(struct qed_hwfn *p_hwfn,
  *
  * @return int
  */
-int qed_vf_pf_txq_stop(struct qed_hwfn *p_hwfn, u16 tx_qid);
+int qed_vf_pf_txq_stop(struct qed_hwfn *p_hwfn, struct qed_queue_cid *p_cid);
 
 /**
  * @brief VF - send a vport update command
@@ -902,9 +896,7 @@ static inline int qed_vf_hw_prepare(struct qed_hwfn *p_hwfn)
 }
 
 static inline int qed_vf_pf_rxq_start(struct qed_hwfn *p_hwfn,
-				      u8 rx_queue_id,
-				      u16 sb,
-				      u8 sb_index,
+				      struct qed_queue_cid *p_cid,
 				      u16 bd_max_bytes,
 				      dma_addr_t bd_chain_phys_adr,
 				      dma_addr_t cqe_pbl_addr,
@@ -914,9 +906,7 @@ static inline int qed_vf_pf_rxq_start(struct qed_hwfn *p_hwfn,
 }
 
 static inline int qed_vf_pf_txq_start(struct qed_hwfn *p_hwfn,
-				      u16 tx_queue_id,
-				      u16 sb,
-				      u8 sb_index,
+				      struct qed_queue_cid *p_cid,
 				      dma_addr_t pbl_addr,
 				      u16 pbl_size, void __iomem **pp_doorbell)
 {
@@ -924,12 +914,14 @@ static inline int qed_vf_pf_txq_start(struct qed_hwfn *p_hwfn,
 }
 
 static inline int qed_vf_pf_rxq_stop(struct qed_hwfn *p_hwfn,
-				     u16 rx_qid, bool cqe_completion)
+				     struct qed_queue_cid *p_cid,
+				     bool cqe_completion)
 {
 	return -EINVAL;
 }
 
-static inline int qed_vf_pf_txq_stop(struct qed_hwfn *p_hwfn, u16 tx_qid)
+static inline int qed_vf_pf_txq_stop(struct qed_hwfn *p_hwfn,
+				     struct qed_queue_cid *p_cid)
 {
 	return -EINVAL;
 }
diff --git a/drivers/net/ethernet/qlogic/qede/qede.h b/drivers/net/ethernet/qlogic/qede/qede.h
index 0cba21bf9d5f..c79dc78746fc 100644
--- a/drivers/net/ethernet/qlogic/qede/qede.h
+++ b/drivers/net/ethernet/qlogic/qede/qede.h
@@ -16,6 +16,7 @@
 #include <linux/bitmap.h>
 #include <linux/kernel.h>
 #include <linux/mutex.h>
+#include <linux/bpf.h>
 #include <linux/io.h>
 #include <linux/qed/common_hsi.h>
 #include <linux/qed/eth_common.h>
@@ -127,10 +128,9 @@ struct qede_dev {
 
 	const struct qed_eth_ops	*ops;
 
-	struct qed_dev_eth_info	dev_info;
+	struct qed_dev_eth_info dev_info;
 #define QEDE_MAX_RSS_CNT(edev)	((edev)->dev_info.num_queues)
-#define QEDE_MAX_TSS_CNT(edev)	((edev)->dev_info.num_queues * \
-				 (edev)->dev_info.num_tc)
+#define QEDE_MAX_TSS_CNT(edev)	((edev)->dev_info.num_queues)
 
 	struct qede_fastpath		*fp_array;
 	u8				req_num_tx;
@@ -139,17 +139,9 @@ struct qede_dev {
 	u8				fp_num_rx;
 	u16				req_queues;
 	u16				num_queues;
-	u8				num_tc;
 #define QEDE_QUEUE_CNT(edev)	((edev)->num_queues)
 #define QEDE_RSS_COUNT(edev)	((edev)->num_queues - (edev)->fp_num_tx)
-#define QEDE_TSS_COUNT(edev)	(((edev)->num_queues - (edev)->fp_num_rx) * \
-				 (edev)->num_tc)
-#define QEDE_TX_IDX(edev, txqidx)	((edev)->fp_num_rx + (txqidx) % \
-					 QEDE_TSS_COUNT(edev))
-#define QEDE_TC_IDX(edev, txqidx)	((txqidx) / QEDE_TSS_COUNT(edev))
-#define QEDE_TX_QUEUE(edev, txqidx)	\
-	(&(edev)->fp_array[QEDE_TX_IDX((edev), (txqidx))].txqs[QEDE_TC_IDX(\
-							(edev), (txqidx))])
+#define QEDE_TSS_COUNT(edev)	((edev)->num_queues - (edev)->fp_num_rx)
 
 	struct qed_int_info		int_info;
 	unsigned char			primary_mac[ETH_ALEN];
@@ -196,6 +188,8 @@ struct qede_dev {
 	bool wol_enabled;
 
 	struct qede_rdma_dev		rdma_info;
+
+	struct bpf_prog *xdp_prog;
 };
 
 enum QEDE_STATE {
@@ -225,39 +219,67 @@ enum qede_agg_state {
 };
 
 struct qede_agg_info {
-	struct sw_rx_data replace_buf;
-	dma_addr_t replace_buf_mapping;
-	struct sw_rx_data start_buf;
-	dma_addr_t start_buf_mapping;
-	struct eth_fast_path_rx_tpa_start_cqe start_cqe;
-	enum qede_agg_state agg_state;
+	/* rx_buf is a data buffer that can be placed / consumed from rx bd
+	 * chain. It has two purposes: We will preallocate the data buffer
+	 * for each aggregation when we open the interface and will place this
+	 * buffer on the rx-bd-ring when we receive TPA_START. We don't want
+	 * to be in a state where allocation fails, as we can't reuse the
+	 * consumer buffer in the rx-chain since FW may still be writing to it
+	 * (since header needs to be modified for TPA).
+	 * The second purpose is to keep a pointer to the bd buffer during
+	 * aggregation.
+	 */
+	struct sw_rx_data buffer;
+	dma_addr_t buffer_mapping;
+
 	struct sk_buff *skb;
-	int frag_id;
+
+	/* We need some structs from the start cookie until termination */
 	u16 vlan_tag;
+	u16 start_cqe_bd_len;
+	u8 start_cqe_placement_offset;
+
+	u8 state;
+	u8 frag_id;
+
+	u8 tunnel_type;
 };
 
 struct qede_rx_queue {
-	__le16			*hw_cons_ptr;
-	struct sw_rx_data	*sw_rx_ring;
-	u16			sw_rx_cons;
-	u16			sw_rx_prod;
-	struct qed_chain	rx_bd_ring;
-	struct qed_chain	rx_comp_ring;
-	void __iomem		*hw_rxq_prod_addr;
+	__le16 *hw_cons_ptr;
+	void __iomem *hw_rxq_prod_addr;
+
+	/* Required for the allocation of replacement buffers */
+	struct device *dev;
+
+	struct bpf_prog *xdp_prog;
+
+	u16 sw_rx_cons;
+	u16 sw_rx_prod;
+
+	u16 num_rx_buffers; /* Slowpath */
+	u8 data_direction;
+	u8 rxq_id;
+
+	u32 rx_buf_size;
+	u32 rx_buf_seg_size;
+
+	u64 rcv_pkts;
+
+	struct sw_rx_data *sw_rx_ring;
+	struct qed_chain rx_bd_ring;
+	struct qed_chain rx_comp_ring ____cacheline_aligned;
 
 	/* GRO */
-	struct qede_agg_info	tpa_info[ETH_TPA_MAX_AGGS_NUM];
+	struct qede_agg_info tpa_info[ETH_TPA_MAX_AGGS_NUM];
 
-	int			rx_buf_size;
-	unsigned int		rx_buf_seg_size;
+	u64 rx_hw_errors;
+	u64 rx_alloc_errors;
+	u64 rx_ip_frags;
 
-	u16			num_rx_buffers;
-	u16			rxq_id;
+	u64 xdp_no_pass;
 
-	u64			rcv_pkts;
-	u64			rx_hw_errors;
-	u64			rx_alloc_errors;
-	u64			rx_ip_frags;
+	void *handle;
 };
 
 union db_prod {
@@ -273,20 +295,39 @@ struct sw_tx_bd {
 };
 
 struct qede_tx_queue {
-	int			index; /* Queue index */
-	__le16			*hw_cons_ptr;
-	struct sw_tx_bd		*sw_tx_ring;
-	u16			sw_tx_cons;
-	u16			sw_tx_prod;
-	struct qed_chain	tx_pbl;
-	void __iomem		*doorbell_addr;
-	union db_prod		tx_db;
-
-	u16			num_tx_buffers;
-	u64			xmit_pkts;
-	u64			stopped_cnt;
-
-	bool			is_legacy;
+	u8 is_xdp;
+	bool is_legacy;
+	u16 sw_tx_cons;
+	u16 sw_tx_prod;
+	u16 num_tx_buffers; /* Slowpath only */
+
+	u64 xmit_pkts;
+	u64 stopped_cnt;
+
+	__le16 *hw_cons_ptr;
+
+	/* Needed for the mapping of packets */
+	struct device *dev;
+
+	void __iomem *doorbell_addr;
+	union db_prod tx_db;
+	int index; /* Slowpath only */
+#define QEDE_TXQ_XDP_TO_IDX(edev, txq)	((txq)->index - \
+					 QEDE_MAX_TSS_CNT(edev))
+#define QEDE_TXQ_IDX_TO_XDP(edev, idx)	((idx) + QEDE_MAX_TSS_CNT(edev))
+
+	/* Regular Tx requires skb + metadata for release purpose,
+	 * while XDP requires only the pages themselves.
+	 */
+	union {
+		struct sw_tx_bd *skbs;
+		struct page **pages;
+	} sw_tx_ring;
+
+	struct qed_chain tx_pbl;
+
+	/* Slowpath; Should be kept in end [unless missing padding] */
+	void *handle;
 };
 
 #define BD_UNMAP_ADDR(bd)		HILO_U64(le32_to_cpu((bd)->addr.hi), \
@@ -303,13 +344,16 @@ struct qede_fastpath {
 	struct qede_dev	*edev;
 #define QEDE_FASTPATH_TX	BIT(0)
 #define QEDE_FASTPATH_RX	BIT(1)
+#define QEDE_FASTPATH_XDP	BIT(2)
 #define QEDE_FASTPATH_COMBINED	(QEDE_FASTPATH_TX | QEDE_FASTPATH_RX)
 	u8			type;
 	u8			id;
+	u8			xdp_xmit;
 	struct napi_struct	napi;
 	struct qed_sb_info	*sb_info;
 	struct qede_rx_queue	*rxq;
-	struct qede_tx_queue	*txqs;
+	struct qede_tx_queue	*txq;
+	struct qede_tx_queue	*xdp_tx;
 
 #define VEC_NAME_SIZE	(sizeof(((struct net_device *)0)->name) + 8)
 	char	name[VEC_NAME_SIZE];
@@ -332,8 +376,13 @@ struct qede_fastpath {
 #define QEDE_SP_VXLAN_PORT_CONFIG	2
 #define QEDE_SP_GENEVE_PORT_CONFIG	3
 
-union qede_reload_args {
-	u16 mtu;
+struct qede_reload_args {
+	void (*func)(struct qede_dev *edev, struct qede_reload_args *args);
+	union {
+		netdev_features_t features;
+		struct bpf_prog *new_prog;
+		u16 mtu;
+	} u;
 };
 
 #ifdef CONFIG_DCB
@@ -342,15 +391,14 @@ void qede_set_dcbnl_ops(struct net_device *ndev);
 void qede_config_debug(uint debug, u32 *p_dp_module, u8 *p_dp_level);
 void qede_set_ethtool_ops(struct net_device *netdev);
 void qede_reload(struct qede_dev *edev,
-		 void (*func)(struct qede_dev *edev,
-			      union qede_reload_args *args),
-		 union qede_reload_args *args);
+		 struct qede_reload_args *args, bool is_locked);
 int qede_change_mtu(struct net_device *dev, int new_mtu);
 void qede_fill_by_demand_stats(struct qede_dev *edev);
+void __qede_lock(struct qede_dev *edev);
+void __qede_unlock(struct qede_dev *edev);
 bool qede_has_rx_work(struct qede_rx_queue *rxq);
 int qede_txq_has_work(struct qede_tx_queue *txq);
-void qede_recycle_rx_bd_ring(struct qede_rx_queue *rxq, struct qede_dev *edev,
-			     u8 count);
+void qede_recycle_rx_bd_ring(struct qede_rx_queue *rxq, u8 count);
 void qede_update_rx_prod(struct qede_dev *edev, struct qede_rx_queue *rxq);
 
 #define RX_RING_SIZE_POW	13
diff --git a/drivers/net/ethernet/qlogic/qede/qede_ethtool.c b/drivers/net/ethernet/qlogic/qede/qede_ethtool.c
index 8a3debef39ee..1c48f445c93b 100644
--- a/drivers/net/ethernet/qlogic/qede/qede_ethtool.c
+++ b/drivers/net/ethernet/qlogic/qede/qede_ethtool.c
@@ -16,13 +16,6 @@
 #include <linux/capability.h>
 #include "qede.h"
 
-#define QEDE_STAT_OFFSET(stat_name) (offsetof(struct qede_stats, stat_name))
-#define QEDE_STAT_STRING(stat_name) (#stat_name)
-#define _QEDE_STAT(stat_name, pf_only) \
-	 {QEDE_STAT_OFFSET(stat_name), QEDE_STAT_STRING(stat_name), pf_only}
-#define QEDE_PF_STAT(stat_name)		_QEDE_STAT(stat_name, true)
-#define QEDE_STAT(stat_name)		_QEDE_STAT(stat_name, false)
-
 #define QEDE_RQSTAT_OFFSET(stat_name) \
 	 (offsetof(struct qede_rx_queue, stat_name))
 #define QEDE_RQSTAT_STRING(stat_name) (#stat_name)
@@ -39,12 +32,10 @@ static const struct {
 	QEDE_RQSTAT(rx_hw_errors),
 	QEDE_RQSTAT(rx_alloc_errors),
 	QEDE_RQSTAT(rx_ip_frags),
+	QEDE_RQSTAT(xdp_no_pass),
 };
 
 #define QEDE_NUM_RQSTATS ARRAY_SIZE(qede_rqstats_arr)
-#define QEDE_RQSTATS_DATA(dev, sindex, rqindex) \
-	(*((u64 *)(((char *)(dev->fp_array[(rqindex)].rxq)) +\
-		    qede_rqstats_arr[(sindex)].offset)))
 #define QEDE_TQSTAT_OFFSET(stat_name) \
 	(offsetof(struct qede_tx_queue, stat_name))
 #define QEDE_TQSTAT_STRING(stat_name) (#stat_name)
@@ -59,10 +50,12 @@ static const struct {
 	QEDE_TQSTAT(stopped_cnt),
 };
 
-#define QEDE_TQSTATS_DATA(dev, sindex, tssid, tcid) \
-	(*((u64 *)(((void *)(&dev->fp_array[tssid].txqs[tcid])) +\
-		   qede_tqstats_arr[(sindex)].offset)))
-
+#define QEDE_STAT_OFFSET(stat_name) (offsetof(struct qede_stats, stat_name))
+#define QEDE_STAT_STRING(stat_name) (#stat_name)
+#define _QEDE_STAT(stat_name, pf_only) \
+	 {QEDE_STAT_OFFSET(stat_name), QEDE_STAT_STRING(stat_name), pf_only}
+#define QEDE_PF_STAT(stat_name)	_QEDE_STAT(stat_name, true)
+#define QEDE_STAT(stat_name)	_QEDE_STAT(stat_name, false)
 static const struct {
 	u64 offset;
 	char string[ETH_GSTRING_LEN];
@@ -136,10 +129,6 @@ static const struct {
 	QEDE_STAT(coalesced_bytes),
 };
 
-#define QEDE_STATS_DATA(dev, index) \
-	(*((u64 *)(((char *)(dev)) + offsetof(struct qede_dev, stats) \
-			+ qede_stats_arr[(index)].offset)))
-
 #define QEDE_NUM_STATS	ARRAY_SIZE(qede_stats_arr)
 
 enum {
@@ -170,39 +159,60 @@ static const char qede_tests_str_arr[QEDE_ETHTOOL_TEST_MAX][ETH_GSTRING_LEN] = {
 	"Nvram (online)\t\t",
 };
 
+static void qede_get_strings_stats_txq(struct qede_dev *edev,
+				       struct qede_tx_queue *txq, u8 **buf)
+{
+	int i;
+
+	for (i = 0; i < QEDE_NUM_TQSTATS; i++) {
+		if (txq->is_xdp)
+			sprintf(*buf, "%d [XDP]: %s",
+				QEDE_TXQ_XDP_TO_IDX(edev, txq),
+				qede_tqstats_arr[i].string);
+		else
+			sprintf(*buf, "%d: %s", txq->index,
+				qede_tqstats_arr[i].string);
+		*buf += ETH_GSTRING_LEN;
+	}
+}
+
+static void qede_get_strings_stats_rxq(struct qede_dev *edev,
+				       struct qede_rx_queue *rxq, u8 **buf)
+{
+	int i;
+
+	for (i = 0; i < QEDE_NUM_RQSTATS; i++) {
+		sprintf(*buf, "%d: %s", rxq->rxq_id,
+			qede_rqstats_arr[i].string);
+		*buf += ETH_GSTRING_LEN;
+	}
+}
+
 static void qede_get_strings_stats(struct qede_dev *edev, u8 *buf)
 {
-	int i, j, k;
+	struct qede_fastpath *fp;
+	int i;
 
-	for (i = 0, k = 0; i < QEDE_QUEUE_CNT(edev); i++) {
-		int tc;
+	/* Account for queue statistics */
+	for (i = 0; i < QEDE_QUEUE_CNT(edev); i++) {
+		fp = &edev->fp_array[i];
 
-		if (edev->fp_array[i].type & QEDE_FASTPATH_RX) {
-			for (j = 0; j < QEDE_NUM_RQSTATS; j++)
-				sprintf(buf + (k + j) * ETH_GSTRING_LEN,
-					"%d:   %s", i,
-					qede_rqstats_arr[j].string);
-			k += QEDE_NUM_RQSTATS;
-		}
+		if (fp->type & QEDE_FASTPATH_RX)
+			qede_get_strings_stats_rxq(edev, fp->rxq, &buf);
 
-		if (edev->fp_array[i].type & QEDE_FASTPATH_TX) {
-			for (tc = 0; tc < edev->num_tc; tc++) {
-				for (j = 0; j < QEDE_NUM_TQSTATS; j++)
-					sprintf(buf + (k + j) *
-						ETH_GSTRING_LEN,
-						"%d.%d: %s", i, tc,
-						qede_tqstats_arr[j].string);
-				k += QEDE_NUM_TQSTATS;
-			}
-		}
+		if (fp->type & QEDE_FASTPATH_XDP)
+			qede_get_strings_stats_txq(edev, fp->xdp_tx, &buf);
+
+		if (fp->type & QEDE_FASTPATH_TX)
+			qede_get_strings_stats_txq(edev, fp->txq, &buf);
 	}
 
-	for (i = 0, j = 0; i < QEDE_NUM_STATS; i++) {
+	/* Account for non-queue statistics */
+	for (i = 0; i < QEDE_NUM_STATS; i++) {
 		if (IS_VF(edev) && qede_stats_arr[i].pf_only)
 			continue;
-		strcpy(buf + (k + j) * ETH_GSTRING_LEN,
-		       qede_stats_arr[i].string);
-		j++;
+		strcpy(buf, qede_stats_arr[i].string);
+		buf += ETH_GSTRING_LEN;
 	}
 }
 
@@ -228,42 +238,61 @@ static void qede_get_strings(struct net_device *dev, u32 stringset, u8 *buf)
 	}
 }
 
+static void qede_get_ethtool_stats_txq(struct qede_tx_queue *txq, u64 **buf)
+{
+	int i;
+
+	for (i = 0; i < QEDE_NUM_TQSTATS; i++) {
+		**buf = *((u64 *)(((void *)txq) + qede_tqstats_arr[i].offset));
+		(*buf)++;
+	}
+}
+
+static void qede_get_ethtool_stats_rxq(struct qede_rx_queue *rxq, u64 **buf)
+{
+	int i;
+
+	for (i = 0; i < QEDE_NUM_RQSTATS; i++) {
+		**buf = *((u64 *)(((void *)rxq) + qede_rqstats_arr[i].offset));
+		(*buf)++;
+	}
+}
+
 static void qede_get_ethtool_stats(struct net_device *dev,
 				   struct ethtool_stats *stats, u64 *buf)
 {
 	struct qede_dev *edev = netdev_priv(dev);
-	int sidx, cnt = 0;
-	int qid;
+	struct qede_fastpath *fp;
+	int i;
 
 	qede_fill_by_demand_stats(edev);
 
-	mutex_lock(&edev->qede_lock);
+	/* Need to protect the access to the fastpath array */
+	__qede_lock(edev);
 
-	for (qid = 0; qid < QEDE_QUEUE_CNT(edev); qid++) {
-		int tc;
+	for (i = 0; i < QEDE_QUEUE_CNT(edev); i++) {
+		fp = &edev->fp_array[i];
 
-		if (edev->fp_array[qid].type & QEDE_FASTPATH_RX) {
-			for (sidx = 0; sidx < QEDE_NUM_RQSTATS; sidx++)
-				buf[cnt++] = QEDE_RQSTATS_DATA(edev, sidx, qid);
-		}
+		if (fp->type & QEDE_FASTPATH_RX)
+			qede_get_ethtool_stats_rxq(fp->rxq, &buf);
 
-		if (edev->fp_array[qid].type & QEDE_FASTPATH_TX) {
-			for (tc = 0; tc < edev->num_tc; tc++) {
-				for (sidx = 0; sidx < QEDE_NUM_TQSTATS; sidx++)
-					buf[cnt++] = QEDE_TQSTATS_DATA(edev,
-								       sidx,
-								       qid, tc);
-			}
-		}
+		if (fp->type & QEDE_FASTPATH_XDP)
+			qede_get_ethtool_stats_txq(fp->xdp_tx, &buf);
+
+		if (fp->type & QEDE_FASTPATH_TX)
+			qede_get_ethtool_stats_txq(fp->txq, &buf);
 	}
 
-	for (sidx = 0; sidx < QEDE_NUM_STATS; sidx++) {
-		if (IS_VF(edev) && qede_stats_arr[sidx].pf_only)
+	for (i = 0; i < QEDE_NUM_STATS; i++) {
+		if (IS_VF(edev) && qede_stats_arr[i].pf_only)
 			continue;
-		buf[cnt++] = QEDE_STATS_DATA(edev, sidx);
+		*buf = *((u64 *)(((void *)&edev->stats) +
+				 qede_stats_arr[i].offset));
+
+		buf++;
 	}
 
-	mutex_unlock(&edev->qede_lock);
+	__qede_unlock(edev);
 }
 
 static int qede_get_sset_count(struct net_device *dev, int stringset)
@@ -280,8 +309,18 @@ static int qede_get_sset_count(struct net_device *dev, int stringset)
 				if (qede_stats_arr[i].pf_only)
 					num_stats--;
 		}
-		return num_stats + QEDE_RSS_COUNT(edev) * QEDE_NUM_RQSTATS +
-		       QEDE_TSS_COUNT(edev) * QEDE_NUM_TQSTATS * edev->num_tc;
+
+		/* Account for the Regular Tx statistics */
+		num_stats += QEDE_TSS_COUNT(edev) * QEDE_NUM_TQSTATS;
+
+		/* Account for the Regular Rx statistics */
+		num_stats += QEDE_RSS_COUNT(edev) * QEDE_NUM_RQSTATS;
+
+		/* Account for XDP statistics [if needed] */
+		if (edev->xdp_prog)
+			num_stats += QEDE_RSS_COUNT(edev) * QEDE_NUM_TQSTATS;
+		return num_stats;
+
 	case ETH_SS_PRIV_FLAGS:
 		return QEDE_PRI_FLAG_LEN;
 	case ETH_SS_TEST:
@@ -352,6 +391,8 @@ static int qede_get_link_ksettings(struct net_device *dev,
 	struct qede_dev *edev = netdev_priv(dev);
 	struct qed_link_output current_link;
 
+	__qede_lock(edev);
+
 	memset(&current_link, 0, sizeof(current_link));
 	edev->ops->common->get_link(edev->cdev, &current_link);
 
@@ -371,6 +412,9 @@ static int qede_get_link_ksettings(struct net_device *dev,
 		base->speed = SPEED_UNKNOWN;
 		base->duplex = DUPLEX_UNKNOWN;
 	}
+
+	__qede_unlock(edev);
+
 	base->port = current_link.port;
 	base->autoneg = (current_link.autoneg) ? AUTONEG_ENABLE :
 			AUTONEG_DISABLE;
@@ -679,8 +723,7 @@ static int qede_set_ringparam(struct net_device *dev,
 	edev->q_num_rx_buffers = ering->rx_pending;
 	edev->q_num_tx_buffers = ering->tx_pending;
 
-	if (netif_running(edev->ndev))
-		qede_reload(edev, NULL, NULL);
+	qede_reload(edev, NULL, false);
 
 	return 0;
 }
@@ -765,29 +808,27 @@ static int qede_get_regs_len(struct net_device *ndev)
 		return -EINVAL;
 }
 
-static void qede_update_mtu(struct qede_dev *edev, union qede_reload_args *args)
+static void qede_update_mtu(struct qede_dev *edev,
+			    struct qede_reload_args *args)
 {
-	edev->ndev->mtu = args->mtu;
+	edev->ndev->mtu = args->u.mtu;
 }
 
 /* Netdevice NDOs */
 int qede_change_mtu(struct net_device *ndev, int new_mtu)
 {
 	struct qede_dev *edev = netdev_priv(ndev);
-	union qede_reload_args args;
+	struct qede_reload_args args;
 
 	DP_VERBOSE(edev, (NETIF_MSG_IFUP | NETIF_MSG_IFDOWN),
 		   "Configuring MTU size of %d\n", new_mtu);
 
-	/* Set the mtu field and re-start the interface if needed*/
-	args.mtu = new_mtu;
-
-	if (netif_running(edev->ndev))
-		qede_reload(edev, &qede_update_mtu, &args);
-
-	qede_update_mtu(edev, &args);
+	/* Set the mtu field and re-start the interface if needed */
+	args.u.mtu = new_mtu;
+	args.func = &qede_update_mtu;
+	qede_reload(edev, &args, false);
 
-	edev->ops->common->update_mtu(edev->cdev, args.mtu);
+	edev->ops->common->update_mtu(edev->cdev, new_mtu);
 
 	return 0;
 }
@@ -871,8 +912,7 @@ static int qede_set_channels(struct net_device *dev,
 		       sizeof(edev->rss_params.rss_ind_table));
 	}
 
-	if (netif_running(dev))
-		qede_reload(edev, NULL, NULL);
+	qede_reload(edev, NULL, false);
 
 	return 0;
 }
@@ -1178,7 +1218,7 @@ static int qede_selftest_transmit_traffic(struct qede_dev *edev,
 
 	for_each_queue(i) {
 		if (edev->fp_array[i].type & QEDE_FASTPATH_TX) {
-			txq = edev->fp_array[i].txqs;
+			txq = edev->fp_array[i].txq;
 			break;
 		}
 	}
@@ -1190,7 +1230,7 @@ static int qede_selftest_transmit_traffic(struct qede_dev *edev,
 
 	/* Fill the entry in the SW ring and the BDs in the FW ring */
 	idx = txq->sw_tx_prod & NUM_TX_BDS_MAX;
-	txq->sw_tx_ring[idx].skb = skb;
+	txq->sw_tx_ring.skbs[idx].skb = skb;
 	first_bd = qed_chain_produce(&txq->tx_pbl);
 	memset(first_bd, 0, sizeof(*first_bd));
 	val = 1 << ETH_TX_1ST_BD_FLAGS_START_BD_SHIFT;
@@ -1244,7 +1284,7 @@ static int qede_selftest_transmit_traffic(struct qede_dev *edev,
 	dma_unmap_single(&edev->pdev->dev, BD_UNMAP_ADDR(first_bd),
 			 BD_UNMAP_LEN(first_bd), DMA_TO_DEVICE);
 	txq->sw_tx_cons++;
-	txq->sw_tx_ring[idx].skb = NULL;
+	txq->sw_tx_ring.skbs[idx].skb = NULL;
 
 	return 0;
 }
@@ -1312,13 +1352,13 @@ static int qede_selftest_receive_traffic(struct qede_dev *edev)
 					break;
 				}
 
-			qede_recycle_rx_bd_ring(rxq, edev, 1);
+			qede_recycle_rx_bd_ring(rxq, 1);
 			qed_chain_recycle_consumed(&rxq->rx_comp_ring);
 			break;
 		}
 
 		DP_INFO(edev, "Not the transmitted packet\n");
-		qede_recycle_rx_bd_ring(rxq, edev, 1);
+		qede_recycle_rx_bd_ring(rxq, 1);
 		qed_chain_recycle_consumed(&rxq->rx_comp_ring);
 	}
 
diff --git a/drivers/net/ethernet/qlogic/qede/qede_main.c b/drivers/net/ethernet/qlogic/qede/qede_main.c
index b84a2c4ef083..172ff6da92ad 100644
--- a/drivers/net/ethernet/qlogic/qede/qede_main.c
+++ b/drivers/net/ethernet/qlogic/qede/qede_main.c
@@ -94,12 +94,26 @@ static int qede_probe(struct pci_dev *pdev, const struct pci_device_id *id);
 
 #define TX_TIMEOUT		(5 * HZ)
 
+/* Utilize last protocol index for XDP */
+#define XDP_PI	11
+
 static void qede_remove(struct pci_dev *pdev);
 static void qede_shutdown(struct pci_dev *pdev);
-static int qede_alloc_rx_buffer(struct qede_dev *edev,
-				struct qede_rx_queue *rxq);
 static void qede_link_update(void *dev, struct qed_link_output *link);
 
+/* The qede lock is used to protect driver state change and driver flows that
+ * are not reentrant.
+ */
+void __qede_lock(struct qede_dev *edev)
+{
+	mutex_lock(&edev->qede_lock);
+}
+
+void __qede_unlock(struct qede_dev *edev)
+{
+	mutex_unlock(&edev->qede_lock);
+}
+
 #ifdef CONFIG_QED_SRIOV
 static int qede_set_vf_vlan(struct net_device *ndev, int vf, u16 vlan, u8 qos,
 			    __be16 vlan_proto)
@@ -290,12 +304,12 @@ static int qede_free_tx_pkt(struct qede_dev *edev,
 			    struct qede_tx_queue *txq, int *len)
 {
 	u16 idx = txq->sw_tx_cons & NUM_TX_BDS_MAX;
-	struct sk_buff *skb = txq->sw_tx_ring[idx].skb;
+	struct sk_buff *skb = txq->sw_tx_ring.skbs[idx].skb;
 	struct eth_tx_1st_bd *first_bd;
 	struct eth_tx_bd *tx_data_bd;
 	int bds_consumed = 0;
 	int nbds;
-	bool data_split = txq->sw_tx_ring[idx].flags & QEDE_TSO_SPLIT_BD;
+	bool data_split = txq->sw_tx_ring.skbs[idx].flags & QEDE_TSO_SPLIT_BD;
 	int i, split_bd_len = 0;
 
 	if (unlikely(!skb)) {
@@ -335,20 +349,19 @@ static int qede_free_tx_pkt(struct qede_dev *edev,
 
 	/* Free skb */
 	dev_kfree_skb_any(skb);
-	txq->sw_tx_ring[idx].skb = NULL;
-	txq->sw_tx_ring[idx].flags = 0;
+	txq->sw_tx_ring.skbs[idx].skb = NULL;
+	txq->sw_tx_ring.skbs[idx].flags = 0;
 
 	return 0;
 }
 
 /* Unmap the data and free skb when mapping failed during start_xmit */
-static void qede_free_failed_tx_pkt(struct qede_dev *edev,
-				    struct qede_tx_queue *txq,
+static void qede_free_failed_tx_pkt(struct qede_tx_queue *txq,
 				    struct eth_tx_1st_bd *first_bd,
 				    int nbd, bool data_split)
 {
 	u16 idx = txq->sw_tx_prod & NUM_TX_BDS_MAX;
-	struct sk_buff *skb = txq->sw_tx_ring[idx].skb;
+	struct sk_buff *skb = txq->sw_tx_ring.skbs[idx].skb;
 	struct eth_tx_bd *tx_data_bd;
 	int i, split_bd_len = 0;
 
@@ -365,7 +378,7 @@ static void qede_free_failed_tx_pkt(struct qede_dev *edev,
 		nbd--;
 	}
 
-	dma_unmap_single(&edev->pdev->dev, BD_UNMAP_ADDR(first_bd),
+	dma_unmap_single(txq->dev, BD_UNMAP_ADDR(first_bd),
 			 BD_UNMAP_LEN(first_bd) + split_bd_len, DMA_TO_DEVICE);
 
 	/* Unmap the data of the skb frags */
@@ -373,7 +386,7 @@ static void qede_free_failed_tx_pkt(struct qede_dev *edev,
 		tx_data_bd = (struct eth_tx_bd *)
 			qed_chain_produce(&txq->tx_pbl);
 		if (tx_data_bd->nbytes)
-			dma_unmap_page(&edev->pdev->dev,
+			dma_unmap_page(txq->dev,
 				       BD_UNMAP_ADDR(tx_data_bd),
 				       BD_UNMAP_LEN(tx_data_bd), DMA_TO_DEVICE);
 	}
@@ -384,12 +397,11 @@ static void qede_free_failed_tx_pkt(struct qede_dev *edev,
 
 	/* Free skb */
 	dev_kfree_skb_any(skb);
-	txq->sw_tx_ring[idx].skb = NULL;
-	txq->sw_tx_ring[idx].flags = 0;
+	txq->sw_tx_ring.skbs[idx].skb = NULL;
+	txq->sw_tx_ring.skbs[idx].flags = 0;
 }
 
-static u32 qede_xmit_type(struct qede_dev *edev,
-			  struct sk_buff *skb, int *ipv6_ext)
+static u32 qede_xmit_type(struct sk_buff *skb, int *ipv6_ext)
 {
 	u32 rc = XMIT_L4_CSUM;
 	__be16 l3_proto;
@@ -456,18 +468,16 @@ static void qede_set_params_for_ipv6_ext(struct sk_buff *skb,
 	second_bd->data.bitfields2 = cpu_to_le16(bd2_bits2);
 }
 
-static int map_frag_to_bd(struct qede_dev *edev,
+static int map_frag_to_bd(struct qede_tx_queue *txq,
 			  skb_frag_t *frag, struct eth_tx_bd *bd)
 {
 	dma_addr_t mapping;
 
 	/* Map skb non-linear frag data for DMA */
-	mapping = skb_frag_dma_map(&edev->pdev->dev, frag, 0,
+	mapping = skb_frag_dma_map(txq->dev, frag, 0,
 				   skb_frag_size(frag), DMA_TO_DEVICE);
-	if (unlikely(dma_mapping_error(&edev->pdev->dev, mapping))) {
-		DP_NOTICE(edev, "Unable to map frag - dropping packet\n");
+	if (unlikely(dma_mapping_error(txq->dev, mapping)))
 		return -ENOMEM;
-	}
 
 	/* Setup the data pointer of the frag data */
 	BD_SET_UNMAP_ADDR_LEN(bd, mapping, skb_frag_size(frag));
@@ -487,8 +497,7 @@ static u16 qede_get_skb_hlen(struct sk_buff *skb, bool is_encap_pkt)
 
 /* +2 for 1st BD for headers and 2nd BD for headlen (if required) */
 #if ((MAX_SKB_FRAGS + 2) > ETH_TX_MAX_BDS_PER_NON_LSO_PACKET)
-static bool qede_pkt_req_lin(struct qede_dev *edev, struct sk_buff *skb,
-			     u8 xmit_type)
+static bool qede_pkt_req_lin(struct sk_buff *skb, u8 xmit_type)
 {
 	int allowed_frags = ETH_TX_MAX_BDS_PER_NON_LSO_PACKET - 1;
 
@@ -524,6 +533,47 @@ static inline void qede_update_tx_producer(struct qede_tx_queue *txq)
 	mmiowb();
 }
 
+static int qede_xdp_xmit(struct qede_dev *edev, struct qede_fastpath *fp,
+			 struct sw_rx_data *metadata, u16 padding, u16 length)
+{
+	struct qede_tx_queue *txq = fp->xdp_tx;
+	u16 idx = txq->sw_tx_prod & NUM_TX_BDS_MAX;
+	struct eth_tx_1st_bd *first_bd;
+
+	if (!qed_chain_get_elem_left(&txq->tx_pbl)) {
+		txq->stopped_cnt++;
+		return -ENOMEM;
+	}
+
+	first_bd = (struct eth_tx_1st_bd *)qed_chain_produce(&txq->tx_pbl);
+
+	memset(first_bd, 0, sizeof(*first_bd));
+	first_bd->data.bd_flags.bitfields =
+	    BIT(ETH_TX_1ST_BD_FLAGS_START_BD_SHIFT);
+	first_bd->data.bitfields |=
+	    (length & ETH_TX_DATA_1ST_BD_PKT_LEN_MASK) <<
+	    ETH_TX_DATA_1ST_BD_PKT_LEN_SHIFT;
+	first_bd->data.nbds = 1;
+
+	/* We can safely ignore the offset, as it's 0 for XDP */
+	BD_SET_UNMAP_ADDR_LEN(first_bd, metadata->mapping + padding, length);
+
+	/* Synchronize the buffer back to device, as program [probably]
+	 * has changed it.
+	 */
+	dma_sync_single_for_device(&edev->pdev->dev,
+				   metadata->mapping + padding,
+				   length, PCI_DMA_TODEVICE);
+
+	txq->sw_tx_ring.pages[idx] = metadata->data;
+	txq->sw_tx_prod++;
+
+	/* Mark the fastpath for future XDP doorbell */
+	fp->xdp_xmit = 1;
+
+	return 0;
+}
+
 /* Main transmit function */
 static netdev_tx_t qede_start_xmit(struct sk_buff *skb,
 				   struct net_device *ndev)
@@ -547,15 +597,15 @@ static netdev_tx_t qede_start_xmit(struct sk_buff *skb,
 	/* Get tx-queue context and netdev index */
 	txq_index = skb_get_queue_mapping(skb);
 	WARN_ON(txq_index >= QEDE_TSS_COUNT(edev));
-	txq = QEDE_TX_QUEUE(edev, txq_index);
+	txq = edev->fp_array[edev->fp_num_rx + txq_index].txq;
 	netdev_txq = netdev_get_tx_queue(ndev, txq_index);
 
 	WARN_ON(qed_chain_get_elem_left(&txq->tx_pbl) < (MAX_SKB_FRAGS + 1));
 
-	xmit_type = qede_xmit_type(edev, skb, &ipv6_ext);
+	xmit_type = qede_xmit_type(skb, &ipv6_ext);
 
 #if ((MAX_SKB_FRAGS + 2) > ETH_TX_MAX_BDS_PER_NON_LSO_PACKET)
-	if (qede_pkt_req_lin(edev, skb, xmit_type)) {
+	if (qede_pkt_req_lin(skb, xmit_type)) {
 		if (skb_linearize(skb)) {
 			DP_NOTICE(edev,
 				  "SKB linearization failed - silently dropping this SKB\n");
@@ -567,7 +617,7 @@ static netdev_tx_t qede_start_xmit(struct sk_buff *skb,
 
 	/* Fill the entry in the SW ring and the BDs in the FW ring */
 	idx = txq->sw_tx_prod & NUM_TX_BDS_MAX;
-	txq->sw_tx_ring[idx].skb = skb;
+	txq->sw_tx_ring.skbs[idx].skb = skb;
 	first_bd = (struct eth_tx_1st_bd *)
 		   qed_chain_produce(&txq->tx_pbl);
 	memset(first_bd, 0, sizeof(*first_bd));
@@ -575,11 +625,11 @@ static netdev_tx_t qede_start_xmit(struct sk_buff *skb,
 		1 << ETH_TX_1ST_BD_FLAGS_START_BD_SHIFT;
 
 	/* Map skb linear data for DMA and set in the first BD */
-	mapping = dma_map_single(&edev->pdev->dev, skb->data,
+	mapping = dma_map_single(txq->dev, skb->data,
 				 skb_headlen(skb), DMA_TO_DEVICE);
-	if (unlikely(dma_mapping_error(&edev->pdev->dev, mapping))) {
+	if (unlikely(dma_mapping_error(txq->dev, mapping))) {
 		DP_NOTICE(edev, "SKB mapping failed\n");
-		qede_free_failed_tx_pkt(edev, txq, first_bd, 0, false);
+		qede_free_failed_tx_pkt(txq, first_bd, 0, false);
 		qede_update_tx_producer(txq);
 		return NETDEV_TX_OK;
 	}
@@ -687,7 +737,7 @@ static netdev_tx_t qede_start_xmit(struct sk_buff *skb,
 			/* this marks the BD as one that has no
 			 * individual mapping
 			 */
-			txq->sw_tx_ring[idx].flags |= QEDE_TSO_SPLIT_BD;
+			txq->sw_tx_ring.skbs[idx].flags |= QEDE_TSO_SPLIT_BD;
 
 			first_bd->nbytes = cpu_to_le16(hlen);
 
@@ -703,12 +753,11 @@ static netdev_tx_t qede_start_xmit(struct sk_buff *skb,
 	/* Handle fragmented skb */
 	/* special handle for frags inside 2nd and 3rd bds.. */
 	while (tx_data_bd && frag_idx < skb_shinfo(skb)->nr_frags) {
-		rc = map_frag_to_bd(edev,
+		rc = map_frag_to_bd(txq,
 				    &skb_shinfo(skb)->frags[frag_idx],
 				    tx_data_bd);
 		if (rc) {
-			qede_free_failed_tx_pkt(edev, txq, first_bd, nbd,
-						data_split);
+			qede_free_failed_tx_pkt(txq, first_bd, nbd, data_split);
 			qede_update_tx_producer(txq);
 			return NETDEV_TX_OK;
 		}
@@ -728,12 +777,11 @@ static netdev_tx_t qede_start_xmit(struct sk_buff *skb,
 
 		memset(tx_data_bd, 0, sizeof(*tx_data_bd));
 
-		rc = map_frag_to_bd(edev,
+		rc = map_frag_to_bd(txq,
 				    &skb_shinfo(skb)->frags[frag_idx],
 				    tx_data_bd);
 		if (rc) {
-			qede_free_failed_tx_pkt(edev, txq, first_bd, nbd,
-						data_split);
+			qede_free_failed_tx_pkt(txq, first_bd, nbd, data_split);
 			qede_update_tx_producer(txq);
 			return NETDEV_TX_OK;
 		}
@@ -798,6 +846,27 @@ int qede_txq_has_work(struct qede_tx_queue *txq)
 	return hw_bd_cons != qed_chain_get_cons_idx(&txq->tx_pbl);
 }
 
+static void qede_xdp_tx_int(struct qede_dev *edev, struct qede_tx_queue *txq)
+{
+	struct eth_tx_1st_bd *bd;
+	u16 hw_bd_cons;
+
+	hw_bd_cons = le16_to_cpu(*txq->hw_cons_ptr);
+	barrier();
+
+	while (hw_bd_cons != qed_chain_get_cons_idx(&txq->tx_pbl)) {
+		bd = (struct eth_tx_1st_bd *)qed_chain_consume(&txq->tx_pbl);
+
+		dma_unmap_single(&edev->pdev->dev, BD_UNMAP_ADDR(bd),
+				 PAGE_SIZE, DMA_BIDIRECTIONAL);
+		__free_page(txq->sw_tx_ring.pages[txq->sw_tx_cons &
+						  NUM_TX_BDS_MAX]);
+
+		txq->sw_tx_cons++;
+		txq->xmit_pkts++;
+	}
+}
+
 static int qede_tx_int(struct qede_dev *edev, struct qede_tx_queue *txq)
 {
 	struct netdev_queue *netdev_txq;
@@ -881,16 +950,6 @@ bool qede_has_rx_work(struct qede_rx_queue *rxq)
 	return hw_comp_cons != sw_comp_cons;
 }
 
-static bool qede_has_tx_work(struct qede_fastpath *fp)
-{
-	u8 tc;
-
-	for (tc = 0; tc < fp->edev->num_tc; tc++)
-		if (qede_txq_has_work(&fp->txqs[tc]))
-			return true;
-	return false;
-}
-
 static inline void qede_rx_bd_ring_consume(struct qede_rx_queue *rxq)
 {
 	qed_chain_consume(&rxq->rx_bd_ring);
@@ -900,8 +959,7 @@ static inline void qede_rx_bd_ring_consume(struct qede_rx_queue *rxq)
 /* This function reuses the buffer(from an offset) from
  * consumer index to producer index in the bd ring
  */
-static inline void qede_reuse_page(struct qede_dev *edev,
-				   struct qede_rx_queue *rxq,
+static inline void qede_reuse_page(struct qede_rx_queue *rxq,
 				   struct sw_rx_data *curr_cons)
 {
 	struct eth_rx_bd *rx_bd_prod = qed_chain_produce(&rxq->rx_bd_ring);
@@ -923,27 +981,62 @@ static inline void qede_reuse_page(struct qede_dev *edev,
 /* In case of allocation failures reuse buffers
  * from consumer index to produce buffers for firmware
  */
-void qede_recycle_rx_bd_ring(struct qede_rx_queue *rxq,
-			     struct qede_dev *edev, u8 count)
+void qede_recycle_rx_bd_ring(struct qede_rx_queue *rxq, u8 count)
 {
 	struct sw_rx_data *curr_cons;
 
 	for (; count > 0; count--) {
 		curr_cons = &rxq->sw_rx_ring[rxq->sw_rx_cons & NUM_RX_BDS_MAX];
-		qede_reuse_page(edev, rxq, curr_cons);
+		qede_reuse_page(rxq, curr_cons);
 		qede_rx_bd_ring_consume(rxq);
 	}
 }
 
-static inline int qede_realloc_rx_buffer(struct qede_dev *edev,
-					 struct qede_rx_queue *rxq,
+static int qede_alloc_rx_buffer(struct qede_rx_queue *rxq)
+{
+	struct sw_rx_data *sw_rx_data;
+	struct eth_rx_bd *rx_bd;
+	dma_addr_t mapping;
+	struct page *data;
+
+	data = alloc_pages(GFP_ATOMIC, 0);
+	if (unlikely(!data))
+		return -ENOMEM;
+
+	/* Map the entire page as it would be used
+	 * for multiple RX buffer segment size mapping.
+	 */
+	mapping = dma_map_page(rxq->dev, data, 0,
+			       PAGE_SIZE, rxq->data_direction);
+	if (unlikely(dma_mapping_error(rxq->dev, mapping))) {
+		__free_page(data);
+		return -ENOMEM;
+	}
+
+	sw_rx_data = &rxq->sw_rx_ring[rxq->sw_rx_prod & NUM_RX_BDS_MAX];
+	sw_rx_data->page_offset = 0;
+	sw_rx_data->data = data;
+	sw_rx_data->mapping = mapping;
+
+	/* Advance PROD and get BD pointer */
+	rx_bd = (struct eth_rx_bd *)qed_chain_produce(&rxq->rx_bd_ring);
+	WARN_ON(!rx_bd);
+	rx_bd->addr.hi = cpu_to_le32(upper_32_bits(mapping));
+	rx_bd->addr.lo = cpu_to_le32(lower_32_bits(mapping));
+
+	rxq->sw_rx_prod++;
+
+	return 0;
+}
+
+static inline int qede_realloc_rx_buffer(struct qede_rx_queue *rxq,
 					 struct sw_rx_data *curr_cons)
 {
 	/* Move to the next segment in the page */
 	curr_cons->page_offset += rxq->rx_buf_seg_size;
 
 	if (curr_cons->page_offset == PAGE_SIZE) {
-		if (unlikely(qede_alloc_rx_buffer(edev, rxq))) {
+		if (unlikely(qede_alloc_rx_buffer(rxq))) {
 			/* Since we failed to allocate new buffer
 			 * current buffer can be used again.
 			 */
@@ -952,15 +1045,15 @@ static inline int qede_realloc_rx_buffer(struct qede_dev *edev,
 			return -ENOMEM;
 		}
 
-		dma_unmap_page(&edev->pdev->dev, curr_cons->mapping,
-			       PAGE_SIZE, DMA_FROM_DEVICE);
+		dma_unmap_page(rxq->dev, curr_cons->mapping,
+			       PAGE_SIZE, rxq->data_direction);
 	} else {
 		/* Increment refcount of the page as we don't want
 		 * network stack to take the ownership of the page
 		 * which can be recycled multiple times by the driver.
 		 */
 		page_ref_inc(curr_cons->data);
-		qede_reuse_page(edev, rxq, curr_cons);
+		qede_reuse_page(rxq, curr_cons);
 	}
 
 	return 0;
@@ -994,22 +1087,20 @@ void qede_update_rx_prod(struct qede_dev *edev, struct qede_rx_queue *rxq)
 	mmiowb();
 }
 
-static u32 qede_get_rxhash(struct qede_dev *edev,
-			   u8 bitfields,
-			   __le32 rss_hash, enum pkt_hash_types *rxhash_type)
+static void qede_get_rxhash(struct sk_buff *skb, u8 bitfields, __le32 rss_hash)
 {
+	enum pkt_hash_types hash_type = PKT_HASH_TYPE_NONE;
 	enum rss_hash_type htype;
+	u32 hash = 0;
 
 	htype = GET_FIELD(bitfields, ETH_FAST_PATH_RX_REG_CQE_RSS_HASH_TYPE);
-
-	if ((edev->ndev->features & NETIF_F_RXHASH) && htype) {
-		*rxhash_type = ((htype == RSS_HASH_TYPE_IPV4) ||
-				(htype == RSS_HASH_TYPE_IPV6)) ?
-				PKT_HASH_TYPE_L3 : PKT_HASH_TYPE_L4;
-		return le32_to_cpu(rss_hash);
+	if (htype) {
+		hash_type = ((htype == RSS_HASH_TYPE_IPV4) ||
+			     (htype == RSS_HASH_TYPE_IPV6)) ?
+			    PKT_HASH_TYPE_L3 : PKT_HASH_TYPE_L4;
+		hash = le32_to_cpu(rss_hash);
 	}
-	*rxhash_type = PKT_HASH_TYPE_NONE;
-	return 0;
+	skb_set_hash(skb, hash, hash_type);
 }
 
 static void qede_set_skb_csum(struct sk_buff *skb, u8 csum_flag)
@@ -1025,12 +1116,14 @@ static void qede_set_skb_csum(struct sk_buff *skb, u8 csum_flag)
 
 static inline void qede_skb_receive(struct qede_dev *edev,
 				    struct qede_fastpath *fp,
+				    struct qede_rx_queue *rxq,
 				    struct sk_buff *skb, u16 vlan_tag)
 {
 	if (vlan_tag)
 		__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), vlan_tag);
 
 	napi_gro_receive(&fp->napi, skb);
+	fp->rxq->rcv_pkts++;
 }
 
 static void qede_set_gro_params(struct qede_dev *edev,
@@ -1058,7 +1151,7 @@ static int qede_fill_frag_skb(struct qede_dev *edev,
 	struct qede_agg_info *tpa_info = &rxq->tpa_info[tpa_agg_index];
 	struct sk_buff *skb = tpa_info->skb;
 
-	if (unlikely(tpa_info->agg_state != QEDE_AGG_STATE_START))
+	if (unlikely(tpa_info->state != QEDE_AGG_STATE_START))
 		goto out;
 
 	/* Add one frag and update the appropriate fields in the skb */
@@ -1066,7 +1159,7 @@ static int qede_fill_frag_skb(struct qede_dev *edev,
 			   current_bd->data, current_bd->page_offset,
 			   len_on_bd);
 
-	if (unlikely(qede_realloc_rx_buffer(edev, rxq, current_bd))) {
+	if (unlikely(qede_realloc_rx_buffer(rxq, current_bd))) {
 		/* Incr page ref count to reuse on allocation failure
 		 * so that it doesn't get freed while freeing SKB.
 		 */
@@ -1084,8 +1177,9 @@ static int qede_fill_frag_skb(struct qede_dev *edev,
 	return 0;
 
 out:
-	tpa_info->agg_state = QEDE_AGG_STATE_ERROR;
-	qede_recycle_rx_bd_ring(rxq, edev, 1);
+	tpa_info->state = QEDE_AGG_STATE_ERROR;
+	qede_recycle_rx_bd_ring(rxq, 1);
+
 	return -ENOMEM;
 }
 
@@ -1096,12 +1190,10 @@ static void qede_tpa_start(struct qede_dev *edev,
 	struct qede_agg_info *tpa_info = &rxq->tpa_info[cqe->tpa_agg_index];
 	struct eth_rx_bd *rx_bd_cons = qed_chain_consume(&rxq->rx_bd_ring);
 	struct eth_rx_bd *rx_bd_prod = qed_chain_produce(&rxq->rx_bd_ring);
-	struct sw_rx_data *replace_buf = &tpa_info->replace_buf;
-	dma_addr_t mapping = tpa_info->replace_buf_mapping;
+	struct sw_rx_data *replace_buf = &tpa_info->buffer;
+	dma_addr_t mapping = tpa_info->buffer_mapping;
 	struct sw_rx_data *sw_rx_data_cons;
 	struct sw_rx_data *sw_rx_data_prod;
-	enum pkt_hash_types rxhash_type;
-	u32 rxhash;
 
 	sw_rx_data_cons = &rxq->sw_rx_ring[rxq->sw_rx_cons & NUM_RX_BDS_MAX];
 	sw_rx_data_prod = &rxq->sw_rx_ring[rxq->sw_rx_prod & NUM_RX_BDS_MAX];
@@ -1122,11 +1214,11 @@ static void qede_tpa_start(struct qede_dev *edev,
 	/* move partial skb from cons to pool (don't unmap yet)
 	 * save mapping, incase we drop the packet later on.
 	 */
-	tpa_info->start_buf = *sw_rx_data_cons;
+	tpa_info->buffer = *sw_rx_data_cons;
 	mapping = HILO_U64(le32_to_cpu(rx_bd_cons->addr.hi),
 			   le32_to_cpu(rx_bd_cons->addr.lo));
 
-	tpa_info->start_buf_mapping = mapping;
+	tpa_info->buffer_mapping = mapping;
 	rxq->sw_rx_cons++;
 
 	/* set tpa state to start only if we are able to allocate skb
@@ -1137,27 +1229,27 @@ static void qede_tpa_start(struct qede_dev *edev,
 					 le16_to_cpu(cqe->len_on_first_bd));
 	if (unlikely(!tpa_info->skb)) {
 		DP_NOTICE(edev, "Failed to allocate SKB for gro\n");
-		tpa_info->agg_state = QEDE_AGG_STATE_ERROR;
+		tpa_info->state = QEDE_AGG_STATE_ERROR;
 		goto cons_buf;
 	}
 
-	skb_put(tpa_info->skb, le16_to_cpu(cqe->len_on_first_bd));
-	memcpy(&tpa_info->start_cqe, cqe, sizeof(tpa_info->start_cqe));
-
 	/* Start filling in the aggregation info */
+	skb_put(tpa_info->skb, le16_to_cpu(cqe->len_on_first_bd));
 	tpa_info->frag_id = 0;
-	tpa_info->agg_state = QEDE_AGG_STATE_START;
+	tpa_info->state = QEDE_AGG_STATE_START;
 
-	rxhash = qede_get_rxhash(edev, cqe->bitfields,
-				 cqe->rss_hash, &rxhash_type);
-	skb_set_hash(tpa_info->skb, rxhash, rxhash_type);
+	/* Store some information from first CQE */
+	tpa_info->start_cqe_placement_offset = cqe->placement_offset;
+	tpa_info->start_cqe_bd_len = le16_to_cpu(cqe->len_on_first_bd);
 	if ((le16_to_cpu(cqe->pars_flags.flags) >>
 	     PARSING_AND_ERR_FLAGS_TAG8021QEXIST_SHIFT) &
-		    PARSING_AND_ERR_FLAGS_TAG8021QEXIST_MASK)
+	    PARSING_AND_ERR_FLAGS_TAG8021QEXIST_MASK)
 		tpa_info->vlan_tag = le16_to_cpu(cqe->vlan_tag);
 	else
 		tpa_info->vlan_tag = 0;
 
+	qede_get_rxhash(tpa_info->skb, cqe->bitfields, cqe->rss_hash);
+
 	/* This is needed in order to enable forwarding support */
 	qede_set_gro_params(edev, tpa_info->skb, cqe);
 
@@ -1169,7 +1261,7 @@ cons_buf: /* We still need to handle bd_len_list to consume buffers */
 	if (unlikely(cqe->ext_bd_len_list[1])) {
 		DP_ERR(edev,
 		       "Unlikely - got a TPA aggregation with more than one ext_bd_len_list entry in the TPA start\n");
-		tpa_info->agg_state = QEDE_AGG_STATE_ERROR;
+		tpa_info->state = QEDE_AGG_STATE_ERROR;
 	}
 }
 
@@ -1239,7 +1331,7 @@ static void qede_gro_receive(struct qede_dev *edev,
 
 send_skb:
 	skb_record_rx_queue(skb, fp->rxq->rxq_id);
-	qede_skb_receive(edev, fp, skb, vlan_tag);
+	qede_skb_receive(edev, fp, fp->rxq, skb, vlan_tag);
 }
 
 static inline void qede_tpa_cont(struct qede_dev *edev,
@@ -1276,7 +1368,7 @@ static void qede_tpa_end(struct qede_dev *edev,
 		DP_ERR(edev,
 		       "Strange - TPA emd with more than a single len_list entry\n");
 
-	if (unlikely(tpa_info->agg_state != QEDE_AGG_STATE_START))
+	if (unlikely(tpa_info->state != QEDE_AGG_STATE_START))
 		goto err;
 
 	/* Sanity */
@@ -1290,14 +1382,9 @@ static void qede_tpa_end(struct qede_dev *edev,
 		       le16_to_cpu(cqe->total_packet_len), skb->len);
 
 	memcpy(skb->data,
-	       page_address(tpa_info->start_buf.data) +
-		tpa_info->start_cqe.placement_offset +
-		tpa_info->start_buf.page_offset,
-	       le16_to_cpu(tpa_info->start_cqe.len_on_first_bd));
-
-	/* Recycle [mapped] start buffer for the next replacement */
-	tpa_info->replace_buf = tpa_info->start_buf;
-	tpa_info->replace_buf_mapping = tpa_info->start_buf_mapping;
+	       page_address(tpa_info->buffer.data) +
+	       tpa_info->start_cqe_placement_offset +
+	       tpa_info->buffer.page_offset, tpa_info->start_cqe_bd_len);
 
 	/* Finalize the SKB */
 	skb->protocol = eth_type_trans(skb, edev->ndev);
@@ -1310,18 +1397,11 @@ static void qede_tpa_end(struct qede_dev *edev,
 
 	qede_gro_receive(edev, fp, skb, tpa_info->vlan_tag);
 
-	tpa_info->agg_state = QEDE_AGG_STATE_NONE;
+	tpa_info->state = QEDE_AGG_STATE_NONE;
 
 	return;
 err:
-	/* The BD starting the aggregation is still mapped; Re-use it for
-	 * future aggregations [as replacement buffer]
-	 */
-	memcpy(&tpa_info->replace_buf, &tpa_info->start_buf,
-	       sizeof(struct sw_rx_data));
-	tpa_info->replace_buf_mapping = tpa_info->start_buf_mapping;
-	tpa_info->start_buf.data = NULL;
-	tpa_info->agg_state = QEDE_AGG_STATE_NONE;
+	tpa_info->state = QEDE_AGG_STATE_NONE;
 	dev_kfree_skb_any(tpa_info->skb);
 	tpa_info->skb = NULL;
 }
@@ -1403,238 +1483,357 @@ static bool qede_pkt_is_ip_fragmented(struct eth_fast_path_rx_reg_cqe *cqe,
 	return false;
 }
 
-static int qede_rx_int(struct qede_fastpath *fp, int budget)
+/* Return true iff packet is to be passed to stack */
+static bool qede_rx_xdp(struct qede_dev *edev,
+			struct qede_fastpath *fp,
+			struct qede_rx_queue *rxq,
+			struct bpf_prog *prog,
+			struct sw_rx_data *bd,
+			struct eth_fast_path_rx_reg_cqe *cqe)
 {
-	struct qede_dev *edev = fp->edev;
-	struct qede_rx_queue *rxq = fp->rxq;
+	u16 len = le16_to_cpu(cqe->len_on_first_bd);
+	struct xdp_buff xdp;
+	enum xdp_action act;
 
-	u16 hw_comp_cons, sw_comp_cons, sw_rx_index, parse_flag;
-	int rx_pkt = 0;
-	u8 csum_flag;
+	xdp.data = page_address(bd->data) + cqe->placement_offset;
+	xdp.data_end = xdp.data + len;
+	act = bpf_prog_run_xdp(prog, &xdp);
 
-	hw_comp_cons = le16_to_cpu(*rxq->hw_cons_ptr);
-	sw_comp_cons = qed_chain_get_cons_idx(&rxq->rx_comp_ring);
+	if (act == XDP_PASS)
+		return true;
 
-	/* Memory barrier to prevent the CPU from doing speculative reads of CQE
-	 * / BD in the while-loop before reading hw_comp_cons. If the CQE is
-	 * read before it is written by FW, then FW writes CQE and SB, and then
-	 * the CPU reads the hw_comp_cons, it will use an old CQE.
-	 */
-	rmb();
+	/* Count number of packets not to be passed to stack */
+	rxq->xdp_no_pass++;
 
-	/* Loop to complete all indicated BDs */
-	while (sw_comp_cons != hw_comp_cons) {
-		struct eth_fast_path_rx_reg_cqe *fp_cqe;
-		enum pkt_hash_types rxhash_type;
-		enum eth_rx_cqe_type cqe_type;
-		struct sw_rx_data *sw_rx_data;
-		union eth_rx_cqe *cqe;
-		struct sk_buff *skb;
-		struct page *data;
-		__le16 flags;
-		u16 len, pad;
-		u32 rx_hash;
-
-		/* Get the CQE from the completion ring */
-		cqe = (union eth_rx_cqe *)
-			qed_chain_consume(&rxq->rx_comp_ring);
-		cqe_type = cqe->fast_path_regular.type;
-
-		if (unlikely(cqe_type == ETH_RX_CQE_TYPE_SLOW_PATH)) {
-			edev->ops->eth_cqe_completion(
-					edev->cdev, fp->id,
-					(struct eth_slow_path_rx_cqe *)cqe);
-			goto next_cqe;
+	switch (act) {
+	case XDP_TX:
+		/* We need the replacement buffer before transmit. */
+		if (qede_alloc_rx_buffer(rxq)) {
+			qede_recycle_rx_bd_ring(rxq, 1);
+			return false;
 		}
 
-		if (cqe_type != ETH_RX_CQE_TYPE_REGULAR) {
-			switch (cqe_type) {
-			case ETH_RX_CQE_TYPE_TPA_START:
-				qede_tpa_start(edev, rxq,
-					       &cqe->fast_path_tpa_start);
-				goto next_cqe;
-			case ETH_RX_CQE_TYPE_TPA_CONT:
-				qede_tpa_cont(edev, rxq,
-					      &cqe->fast_path_tpa_cont);
-				goto next_cqe;
-			case ETH_RX_CQE_TYPE_TPA_END:
-				qede_tpa_end(edev, fp,
-					     &cqe->fast_path_tpa_end);
-				goto next_rx_only;
-			default:
-				break;
-			}
+		/* Now if there's a transmission problem, we'd still have to
+		 * throw current buffer, as replacement was already allocated.
+		 */
+		if (qede_xdp_xmit(edev, fp, bd, cqe->placement_offset, len)) {
+			dma_unmap_page(rxq->dev, bd->mapping,
+				       PAGE_SIZE, DMA_BIDIRECTIONAL);
+			__free_page(bd->data);
 		}
 
-		/* Get the data from the SW ring */
-		sw_rx_index = rxq->sw_rx_cons & NUM_RX_BDS_MAX;
-		sw_rx_data = &rxq->sw_rx_ring[sw_rx_index];
-		data = sw_rx_data->data;
-
-		fp_cqe = &cqe->fast_path_regular;
-		len =  le16_to_cpu(fp_cqe->len_on_first_bd);
-		pad = fp_cqe->placement_offset;
-		flags = cqe->fast_path_regular.pars_flags.flags;
-
-		/* If this is an error packet then drop it */
-		parse_flag = le16_to_cpu(flags);
-
-		csum_flag = qede_check_csum(parse_flag);
-		if (unlikely(csum_flag == QEDE_CSUM_ERROR)) {
-			if (qede_pkt_is_ip_fragmented(&cqe->fast_path_regular,
-						      parse_flag)) {
-				rxq->rx_ip_frags++;
-				goto alloc_skb;
-			}
+		/* Regardless, we've consumed an Rx BD */
+		qede_rx_bd_ring_consume(rxq);
+		return false;
 
-			DP_NOTICE(edev,
-				  "CQE in CONS = %u has error, flags = %x, dropping incoming packet\n",
-				  sw_comp_cons, parse_flag);
-			rxq->rx_hw_errors++;
-			qede_recycle_rx_bd_ring(rxq, edev, fp_cqe->bd_num);
-			goto next_cqe;
-		}
+	default:
+		bpf_warn_invalid_xdp_action(act);
+	case XDP_ABORTED:
+	case XDP_DROP:
+		qede_recycle_rx_bd_ring(rxq, cqe->bd_num);
+	}
 
-alloc_skb:
-		skb = netdev_alloc_skb(edev->ndev, QEDE_RX_HDR_SIZE);
-		if (unlikely(!skb)) {
-			DP_NOTICE(edev,
-				  "skb allocation failed, dropping incoming packet\n");
-			qede_recycle_rx_bd_ring(rxq, edev, fp_cqe->bd_num);
-			rxq->rx_alloc_errors++;
-			goto next_cqe;
+	return false;
+}
+
+static struct sk_buff *qede_rx_allocate_skb(struct qede_dev *edev,
+					    struct qede_rx_queue *rxq,
+					    struct sw_rx_data *bd, u16 len,
+					    u16 pad)
+{
+	unsigned int offset = bd->page_offset;
+	struct skb_frag_struct *frag;
+	struct page *page = bd->data;
+	unsigned int pull_len;
+	struct sk_buff *skb;
+	unsigned char *va;
+
+	/* Allocate a new SKB with a sufficient large header len */
+	skb = netdev_alloc_skb(edev->ndev, QEDE_RX_HDR_SIZE);
+	if (unlikely(!skb))
+		return NULL;
+
+	/* Copy data into SKB - if it's small, we can simply copy it and
+	 * re-use the already allcoated & mapped memory.
+	 */
+	if (len + pad <= edev->rx_copybreak) {
+		memcpy(skb_put(skb, len),
+		       page_address(page) + pad + offset, len);
+		qede_reuse_page(rxq, bd);
+		goto out;
+	}
+
+	frag = &skb_shinfo(skb)->frags[0];
+
+	skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags,
+			page, pad + offset, len, rxq->rx_buf_seg_size);
+
+	va = skb_frag_address(frag);
+	pull_len = eth_get_headlen(va, QEDE_RX_HDR_SIZE);
+
+	/* Align the pull_len to optimize memcpy */
+	memcpy(skb->data, va, ALIGN(pull_len, sizeof(long)));
+
+	/* Correct the skb & frag sizes offset after the pull */
+	skb_frag_size_sub(frag, pull_len);
+	frag->page_offset += pull_len;
+	skb->data_len -= pull_len;
+	skb->tail += pull_len;
+
+	if (unlikely(qede_realloc_rx_buffer(rxq, bd))) {
+		/* Incr page ref count to reuse on allocation failure so
+		 * that it doesn't get freed while freeing SKB [as its
+		 * already mapped there].
+		 */
+		page_ref_inc(page);
+		dev_kfree_skb_any(skb);
+		return NULL;
+	}
+
+out:
+	/* We've consumed the first BD and prepared an SKB */
+	qede_rx_bd_ring_consume(rxq);
+	return skb;
+}
+
+static int qede_rx_build_jumbo(struct qede_dev *edev,
+			       struct qede_rx_queue *rxq,
+			       struct sk_buff *skb,
+			       struct eth_fast_path_rx_reg_cqe *cqe,
+			       u16 first_bd_len)
+{
+	u16 pkt_len = le16_to_cpu(cqe->pkt_len);
+	struct sw_rx_data *bd;
+	u16 bd_cons_idx;
+	u8 num_frags;
+
+	pkt_len -= first_bd_len;
+
+	/* We've already used one BD for the SKB. Now take care of the rest */
+	for (num_frags = cqe->bd_num - 1; num_frags > 0; num_frags--) {
+		u16 cur_size = pkt_len > rxq->rx_buf_size ? rxq->rx_buf_size :
+		    pkt_len;
+
+		if (unlikely(!cur_size)) {
+			DP_ERR(edev,
+			       "Still got %d BDs for mapping jumbo, but length became 0\n",
+			       num_frags);
+			goto out;
 		}
 
-		/* Copy data into SKB */
-		if (len + pad <= edev->rx_copybreak) {
-			memcpy(skb_put(skb, len),
-			       page_address(data) + pad +
-				sw_rx_data->page_offset, len);
-			qede_reuse_page(edev, rxq, sw_rx_data);
+		/* We need a replacement buffer for each BD */
+		if (unlikely(qede_alloc_rx_buffer(rxq)))
+			goto out;
+
+		/* Now that we've allocated the replacement buffer,
+		 * we can safely consume the next BD and map it to the SKB.
+		 */
+		bd_cons_idx = rxq->sw_rx_cons & NUM_RX_BDS_MAX;
+		bd = &rxq->sw_rx_ring[bd_cons_idx];
+		qede_rx_bd_ring_consume(rxq);
+
+		dma_unmap_page(rxq->dev, bd->mapping,
+			       PAGE_SIZE, DMA_FROM_DEVICE);
+
+		skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags++,
+				   bd->data, 0, cur_size);
+
+		skb->truesize += PAGE_SIZE;
+		skb->data_len += cur_size;
+		skb->len += cur_size;
+		pkt_len -= cur_size;
+	}
+
+	if (unlikely(pkt_len))
+		DP_ERR(edev,
+		       "Mapped all BDs of jumbo, but still have %d bytes\n",
+		       pkt_len);
+
+out:
+	return num_frags;
+}
+
+static int qede_rx_process_tpa_cqe(struct qede_dev *edev,
+				   struct qede_fastpath *fp,
+				   struct qede_rx_queue *rxq,
+				   union eth_rx_cqe *cqe,
+				   enum eth_rx_cqe_type type)
+{
+	switch (type) {
+	case ETH_RX_CQE_TYPE_TPA_START:
+		qede_tpa_start(edev, rxq, &cqe->fast_path_tpa_start);
+		return 0;
+	case ETH_RX_CQE_TYPE_TPA_CONT:
+		qede_tpa_cont(edev, rxq, &cqe->fast_path_tpa_cont);
+		return 0;
+	case ETH_RX_CQE_TYPE_TPA_END:
+		qede_tpa_end(edev, fp, &cqe->fast_path_tpa_end);
+		return 1;
+	default:
+		return 0;
+	}
+}
+
+static int qede_rx_process_cqe(struct qede_dev *edev,
+			       struct qede_fastpath *fp,
+			       struct qede_rx_queue *rxq)
+{
+	struct bpf_prog *xdp_prog = READ_ONCE(rxq->xdp_prog);
+	struct eth_fast_path_rx_reg_cqe *fp_cqe;
+	u16 len, pad, bd_cons_idx, parse_flag;
+	enum eth_rx_cqe_type cqe_type;
+	union eth_rx_cqe *cqe;
+	struct sw_rx_data *bd;
+	struct sk_buff *skb;
+	__le16 flags;
+	u8 csum_flag;
+
+	/* Get the CQE from the completion ring */
+	cqe = (union eth_rx_cqe *)qed_chain_consume(&rxq->rx_comp_ring);
+	cqe_type = cqe->fast_path_regular.type;
+
+	/* Process an unlikely slowpath event */
+	if (unlikely(cqe_type == ETH_RX_CQE_TYPE_SLOW_PATH)) {
+		struct eth_slow_path_rx_cqe *sp_cqe;
+
+		sp_cqe = (struct eth_slow_path_rx_cqe *)cqe;
+		edev->ops->eth_cqe_completion(edev->cdev, fp->id, sp_cqe);
+		return 0;
+	}
+
+	/* Handle TPA cqes */
+	if (cqe_type != ETH_RX_CQE_TYPE_REGULAR)
+		return qede_rx_process_tpa_cqe(edev, fp, rxq, cqe, cqe_type);
+
+	/* Get the data from the SW ring; Consume it only after it's evident
+	 * we wouldn't recycle it.
+	 */
+	bd_cons_idx = rxq->sw_rx_cons & NUM_RX_BDS_MAX;
+	bd = &rxq->sw_rx_ring[bd_cons_idx];
+
+	fp_cqe = &cqe->fast_path_regular;
+	len = le16_to_cpu(fp_cqe->len_on_first_bd);
+	pad = fp_cqe->placement_offset;
+
+	/* Run eBPF program if one is attached */
+	if (xdp_prog)
+		if (!qede_rx_xdp(edev, fp, rxq, xdp_prog, bd, fp_cqe))
+			return 1;
+
+	/* If this is an error packet then drop it */
+	flags = cqe->fast_path_regular.pars_flags.flags;
+	parse_flag = le16_to_cpu(flags);
+
+	csum_flag = qede_check_csum(parse_flag);
+	if (unlikely(csum_flag == QEDE_CSUM_ERROR)) {
+		if (qede_pkt_is_ip_fragmented(fp_cqe, parse_flag)) {
+			rxq->rx_ip_frags++;
 		} else {
-			struct skb_frag_struct *frag;
-			unsigned int pull_len;
-			unsigned char *va;
-
-			frag = &skb_shinfo(skb)->frags[0];
-
-			skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, data,
-					pad + sw_rx_data->page_offset,
-					len, rxq->rx_buf_seg_size);
-
-			va = skb_frag_address(frag);
-			pull_len = eth_get_headlen(va, QEDE_RX_HDR_SIZE);
-
-			/* Align the pull_len to optimize memcpy */
-			memcpy(skb->data, va, ALIGN(pull_len, sizeof(long)));
-
-			skb_frag_size_sub(frag, pull_len);
-			frag->page_offset += pull_len;
-			skb->data_len -= pull_len;
-			skb->tail += pull_len;
-
-			if (unlikely(qede_realloc_rx_buffer(edev, rxq,
-							    sw_rx_data))) {
-				DP_ERR(edev, "Failed to allocate rx buffer\n");
-				/* Incr page ref count to reuse on allocation
-				 * failure so that it doesn't get freed while
-				 * freeing SKB.
-				 */
-
-				page_ref_inc(sw_rx_data->data);
-				rxq->rx_alloc_errors++;
-				qede_recycle_rx_bd_ring(rxq, edev,
-							fp_cqe->bd_num);
-				dev_kfree_skb_any(skb);
-				goto next_cqe;
-			}
+			DP_NOTICE(edev,
+				  "CQE has error, flags = %x, dropping incoming packet\n",
+				  parse_flag);
+			rxq->rx_hw_errors++;
+			qede_recycle_rx_bd_ring(rxq, fp_cqe->bd_num);
+			return 0;
 		}
+	}
 
-		qede_rx_bd_ring_consume(rxq);
+	/* Basic validation passed; Need to prepare an SKB. This would also
+	 * guarantee to finally consume the first BD upon success.
+	 */
+	skb = qede_rx_allocate_skb(edev, rxq, bd, len, pad);
+	if (!skb) {
+		rxq->rx_alloc_errors++;
+		qede_recycle_rx_bd_ring(rxq, fp_cqe->bd_num);
+		return 0;
+	}
 
-		if (fp_cqe->bd_num != 1) {
-			u16 pkt_len = le16_to_cpu(fp_cqe->pkt_len);
-			u8 num_frags;
-
-			pkt_len -= len;
-
-			for (num_frags = fp_cqe->bd_num - 1; num_frags > 0;
-			     num_frags--) {
-				u16 cur_size = pkt_len > rxq->rx_buf_size ?
-						rxq->rx_buf_size : pkt_len;
-				if (unlikely(!cur_size)) {
-					DP_ERR(edev,
-					       "Still got %d BDs for mapping jumbo, but length became 0\n",
-					       num_frags);
-					qede_recycle_rx_bd_ring(rxq, edev,
-								num_frags);
-					dev_kfree_skb_any(skb);
-					goto next_cqe;
-				}
-
-				if (unlikely(qede_alloc_rx_buffer(edev, rxq))) {
-					qede_recycle_rx_bd_ring(rxq, edev,
-								num_frags);
-					dev_kfree_skb_any(skb);
-					goto next_cqe;
-				}
-
-				sw_rx_index = rxq->sw_rx_cons & NUM_RX_BDS_MAX;
-				sw_rx_data = &rxq->sw_rx_ring[sw_rx_index];
-				qede_rx_bd_ring_consume(rxq);
-
-				dma_unmap_page(&edev->pdev->dev,
-					       sw_rx_data->mapping,
-					       PAGE_SIZE, DMA_FROM_DEVICE);
-
-				skb_fill_page_desc(skb,
-						   skb_shinfo(skb)->nr_frags++,
-						   sw_rx_data->data, 0,
-						   cur_size);
-
-				skb->truesize += PAGE_SIZE;
-				skb->data_len += cur_size;
-				skb->len += cur_size;
-				pkt_len -= cur_size;
-			}
+	/* In case of Jumbo packet, several PAGE_SIZEd buffers will be pointed
+	 * by a single cqe.
+	 */
+	if (fp_cqe->bd_num > 1) {
+		u16 unmapped_frags = qede_rx_build_jumbo(edev, rxq, skb,
+							 fp_cqe, len);
 
-			if (unlikely(pkt_len))
-				DP_ERR(edev,
-				       "Mapped all BDs of jumbo, but still have %d bytes\n",
-				       pkt_len);
+		if (unlikely(unmapped_frags > 0)) {
+			qede_recycle_rx_bd_ring(rxq, unmapped_frags);
+			dev_kfree_skb_any(skb);
+			return 0;
 		}
+	}
 
-		skb->protocol = eth_type_trans(skb, edev->ndev);
+	/* The SKB contains all the data. Now prepare meta-magic */
+	skb->protocol = eth_type_trans(skb, edev->ndev);
+	qede_get_rxhash(skb, fp_cqe->bitfields, fp_cqe->rss_hash);
+	qede_set_skb_csum(skb, csum_flag);
+	skb_record_rx_queue(skb, rxq->rxq_id);
 
-		rx_hash = qede_get_rxhash(edev, fp_cqe->bitfields,
-					  fp_cqe->rss_hash, &rxhash_type);
+	/* SKB is prepared - pass it to stack */
+	qede_skb_receive(edev, fp, rxq, skb, le16_to_cpu(fp_cqe->vlan_tag));
 
-		skb_set_hash(skb, rx_hash, rxhash_type);
+	return 1;
+}
 
-		qede_set_skb_csum(skb, csum_flag);
+static int qede_rx_int(struct qede_fastpath *fp, int budget)
+{
+	struct qede_rx_queue *rxq = fp->rxq;
+	struct qede_dev *edev = fp->edev;
+	u16 hw_comp_cons, sw_comp_cons;
+	int work_done = 0;
 
-		skb_record_rx_queue(skb, fp->rxq->rxq_id);
+	hw_comp_cons = le16_to_cpu(*rxq->hw_cons_ptr);
+	sw_comp_cons = qed_chain_get_cons_idx(&rxq->rx_comp_ring);
 
-		qede_skb_receive(edev, fp, skb, le16_to_cpu(fp_cqe->vlan_tag));
-next_rx_only:
-		rx_pkt++;
+	/* Memory barrier to prevent the CPU from doing speculative reads of CQE
+	 * / BD in the while-loop before reading hw_comp_cons. If the CQE is
+	 * read before it is written by FW, then FW writes CQE and SB, and then
+	 * the CPU reads the hw_comp_cons, it will use an old CQE.
+	 */
+	rmb();
 
-next_cqe: /* don't consume bd rx buffer */
+	/* Loop to complete all indicated BDs */
+	while ((sw_comp_cons != hw_comp_cons) && (work_done < budget)) {
+		qede_rx_process_cqe(edev, fp, rxq);
 		qed_chain_recycle_consumed(&rxq->rx_comp_ring);
 		sw_comp_cons = qed_chain_get_cons_idx(&rxq->rx_comp_ring);
-		/* CR TPA - revisit how to handle budget in TPA perhaps
-		 * increase on "end"
-		 */
-		if (rx_pkt == budget)
-			break;
-	} /* repeat while sw_comp_cons != hw_comp_cons... */
+		work_done++;
+	}
 
 	/* Update producers */
 	qede_update_rx_prod(edev, rxq);
 
-	rxq->rcv_pkts += rx_pkt;
+	return work_done;
+}
 
-	return rx_pkt;
+static bool qede_poll_is_more_work(struct qede_fastpath *fp)
+{
+	qed_sb_update_sb_idx(fp->sb_info);
+
+	/* *_has_*_work() reads the status block, thus we need to ensure that
+	 * status block indices have been actually read (qed_sb_update_sb_idx)
+	 * prior to this check (*_has_*_work) so that we won't write the
+	 * "newer" value of the status block to HW (if there was a DMA right
+	 * after qede_has_rx_work and if there is no rmb, the memory reading
+	 * (qed_sb_update_sb_idx) may be postponed to right before *_ack_sb).
+	 * In this case there will never be another interrupt until there is
+	 * another update of the status block, while there is still unhandled
+	 * work.
+	 */
+	rmb();
+
+	if (likely(fp->type & QEDE_FASTPATH_RX))
+		if (qede_has_rx_work(fp->rxq))
+			return true;
+
+	if (fp->type & QEDE_FASTPATH_XDP)
+		if (qede_txq_has_work(fp->xdp_tx))
+			return true;
+
+	if (likely(fp->type & QEDE_FASTPATH_TX))
+		if (qede_txq_has_work(fp->txq))
+			return true;
+
+	return false;
 }
 
 static int qede_poll(struct napi_struct *napi, int budget)
@@ -1643,48 +1842,35 @@ static int qede_poll(struct napi_struct *napi, int budget)
 						napi);
 	struct qede_dev *edev = fp->edev;
 	int rx_work_done = 0;
-	u8 tc;
 
-	for (tc = 0; tc < edev->num_tc; tc++)
-		if (likely(fp->type & QEDE_FASTPATH_TX) &&
-		    qede_txq_has_work(&fp->txqs[tc]))
-			qede_tx_int(edev, &fp->txqs[tc]);
+	if (likely(fp->type & QEDE_FASTPATH_TX) && qede_txq_has_work(fp->txq))
+		qede_tx_int(edev, fp->txq);
+
+	if ((fp->type & QEDE_FASTPATH_XDP) && qede_txq_has_work(fp->xdp_tx))
+		qede_xdp_tx_int(edev, fp->xdp_tx);
 
 	rx_work_done = (likely(fp->type & QEDE_FASTPATH_RX) &&
 			qede_has_rx_work(fp->rxq)) ?
 			qede_rx_int(fp, budget) : 0;
 	if (rx_work_done < budget) {
-		qed_sb_update_sb_idx(fp->sb_info);
-		/* *_has_*_work() reads the status block,
-		 * thus we need to ensure that status block indices
-		 * have been actually read (qed_sb_update_sb_idx)
-		 * prior to this check (*_has_*_work) so that
-		 * we won't write the "newer" value of the status block
-		 * to HW (if there was a DMA right after
-		 * qede_has_rx_work and if there is no rmb, the memory
-		 * reading (qed_sb_update_sb_idx) may be postponed
-		 * to right before *_ack_sb). In this case there
-		 * will never be another interrupt until there is
-		 * another update of the status block, while there
-		 * is still unhandled work.
-		 */
-		rmb();
-
-		/* Fall out from the NAPI loop if needed */
-		if (!((likely(fp->type & QEDE_FASTPATH_RX) &&
-		       qede_has_rx_work(fp->rxq)) ||
-		      (likely(fp->type & QEDE_FASTPATH_TX) &&
-		       qede_has_tx_work(fp)))) {
+		if (!qede_poll_is_more_work(fp)) {
 			napi_complete(napi);
 
 			/* Update and reenable interrupts */
-			qed_sb_ack(fp->sb_info, IGU_INT_ENABLE,
-				   1 /*update*/);
+			qed_sb_ack(fp->sb_info, IGU_INT_ENABLE, 1);
 		} else {
 			rx_work_done = budget;
 		}
 	}
 
+	if (fp->xdp_xmit) {
+		u16 xdp_prod = qed_chain_get_prod_idx(&fp->xdp_tx->tx_pbl);
+
+		fp->xdp_xmit = 0;
+		fp->xdp_tx->tx_db.data.bd_prod = cpu_to_le16(xdp_prod);
+		qede_update_tx_producer(fp->xdp_tx);
+	}
+
 	return rx_work_done;
 }
 
@@ -1935,7 +2121,7 @@ static int qede_vlan_rx_add_vid(struct net_device *dev, __be16 proto, u16 vid)
 {
 	struct qede_dev *edev = netdev_priv(dev);
 	struct qede_vlan *vlan, *tmp;
-	int rc;
+	int rc = 0;
 
 	DP_VERBOSE(edev, NETIF_MSG_IFUP, "Adding vlan 0x%04x\n", vid);
 
@@ -1959,6 +2145,7 @@ static int qede_vlan_rx_add_vid(struct net_device *dev, __be16 proto, u16 vid)
 	}
 
 	/* If interface is down, cache this VLAN ID and return */
+	__qede_lock(edev);
 	if (edev->state != QEDE_STATE_OPEN) {
 		DP_VERBOSE(edev, NETIF_MSG_IFDOWN,
 			   "Interface is down, VLAN %d will be configured when interface is up\n",
@@ -1966,8 +2153,7 @@ static int qede_vlan_rx_add_vid(struct net_device *dev, __be16 proto, u16 vid)
 		if (vid != 0)
 			edev->non_configured_vlans++;
 		list_add(&vlan->list, &edev->vlan_list);
-
-		return 0;
+		goto out;
 	}
 
 	/* Check for the filter limit.
@@ -1983,7 +2169,7 @@ static int qede_vlan_rx_add_vid(struct net_device *dev, __be16 proto, u16 vid)
 			DP_ERR(edev, "Failed to configure VLAN %d\n",
 			       vlan->vid);
 			kfree(vlan);
-			return -EINVAL;
+			goto out;
 		}
 		vlan->configured = true;
 
@@ -2000,7 +2186,9 @@ static int qede_vlan_rx_add_vid(struct net_device *dev, __be16 proto, u16 vid)
 
 	list_add(&vlan->list, &edev->vlan_list);
 
-	return 0;
+out:
+	__qede_unlock(edev);
+	return rc;
 }
 
 static void qede_del_vlan_from_list(struct qede_dev *edev,
@@ -2077,11 +2265,12 @@ static int qede_vlan_rx_kill_vid(struct net_device *dev, __be16 proto, u16 vid)
 {
 	struct qede_dev *edev = netdev_priv(dev);
 	struct qede_vlan *vlan = NULL;
-	int rc;
+	int rc = 0;
 
 	DP_VERBOSE(edev, NETIF_MSG_IFDOWN, "Removing vlan 0x%04x\n", vid);
 
 	/* Find whether entry exists */
+	__qede_lock(edev);
 	list_for_each_entry(vlan, &edev->vlan_list, list)
 		if (vlan->vid == vid)
 			break;
@@ -2089,7 +2278,7 @@ static int qede_vlan_rx_kill_vid(struct net_device *dev, __be16 proto, u16 vid)
 	if (!vlan || (vlan->vid != vid)) {
 		DP_VERBOSE(edev, (NETIF_MSG_IFUP | NETIF_MSG_IFDOWN),
 			   "Vlan isn't configured\n");
-		return 0;
+		goto out;
 	}
 
 	if (edev->state != QEDE_STATE_OPEN) {
@@ -2099,7 +2288,7 @@ static int qede_vlan_rx_kill_vid(struct net_device *dev, __be16 proto, u16 vid)
 		DP_VERBOSE(edev, NETIF_MSG_IFDOWN,
 			   "Interface is down, removing VLAN from list only\n");
 		qede_del_vlan_from_list(edev, vlan);
-		return 0;
+		goto out;
 	}
 
 	/* Remove vlan */
@@ -2108,7 +2297,7 @@ static int qede_vlan_rx_kill_vid(struct net_device *dev, __be16 proto, u16 vid)
 					    vid);
 		if (rc) {
 			DP_ERR(edev, "Failed to remove VLAN %d\n", vid);
-			return -EINVAL;
+			goto out;
 		}
 	}
 
@@ -2119,6 +2308,8 @@ static int qede_vlan_rx_kill_vid(struct net_device *dev, __be16 proto, u16 vid)
 	 */
 	rc = qede_configure_vlan_filters(edev);
 
+out:
+	__qede_unlock(edev);
 	return rc;
 }
 
@@ -2148,7 +2339,13 @@ static void qede_vlan_mark_nonconfigured(struct qede_dev *edev)
 	edev->accept_any_vlan = false;
 }
 
-static int qede_set_features(struct net_device *dev, netdev_features_t features)
+static void qede_set_features_reload(struct qede_dev *edev,
+				     struct qede_reload_args *args)
+{
+	edev->ndev->features = args->u.features;
+}
+
+int qede_set_features(struct net_device *dev, netdev_features_t features)
 {
 	struct qede_dev *edev = netdev_priv(dev);
 	netdev_features_t changes = features ^ dev->features;
@@ -2162,9 +2359,23 @@ static int qede_set_features(struct net_device *dev, netdev_features_t features)
 			need_reload = edev->gro_disable;
 	}
 
-	if (need_reload && netif_running(edev->ndev)) {
-		dev->features = features;
-		qede_reload(edev, NULL, NULL);
+	if (need_reload) {
+		struct qede_reload_args args;
+
+		args.u.features = features;
+		args.func = &qede_set_features_reload;
+
+		/* Make sure that we definitely need to reload.
+		 * In case of an eBPF attached program, there will be no FW
+		 * aggregations, so no need to actually reload.
+		 */
+		__qede_lock(edev);
+		if (edev->xdp_prog)
+			args.func(edev, &args);
+		else
+			qede_reload(edev, &args, true);
+		__qede_unlock(edev);
+
 		return 1;
 	}
 
@@ -2275,6 +2486,43 @@ static netdev_features_t qede_features_check(struct sk_buff *skb,
 	return features;
 }
 
+static void qede_xdp_reload_func(struct qede_dev *edev,
+				 struct qede_reload_args *args)
+{
+	struct bpf_prog *old;
+
+	old = xchg(&edev->xdp_prog, args->u.new_prog);
+	if (old)
+		bpf_prog_put(old);
+}
+
+static int qede_xdp_set(struct qede_dev *edev, struct bpf_prog *prog)
+{
+	struct qede_reload_args args;
+
+	/* If we're called, there was already a bpf reference increment */
+	args.func = &qede_xdp_reload_func;
+	args.u.new_prog = prog;
+	qede_reload(edev, &args, false);
+
+	return 0;
+}
+
+static int qede_xdp(struct net_device *dev, struct netdev_xdp *xdp)
+{
+	struct qede_dev *edev = netdev_priv(dev);
+
+	switch (xdp->command) {
+	case XDP_SETUP_PROG:
+		return qede_xdp_set(edev, xdp->prog);
+	case XDP_QUERY_PROG:
+		xdp->prog_attached = !!edev->xdp_prog;
+		return 0;
+	default:
+		return -EINVAL;
+	}
+}
+
 static const struct net_device_ops qede_netdev_ops = {
 	.ndo_open = qede_open,
 	.ndo_stop = qede_close,
@@ -2300,6 +2548,7 @@ static const struct net_device_ops qede_netdev_ops = {
 	.ndo_udp_tunnel_add = qede_udp_tunnel_add,
 	.ndo_udp_tunnel_del = qede_udp_tunnel_del,
 	.ndo_features_check = qede_features_check,
+	.ndo_xdp = qede_xdp,
 };
 
 /* -------------------------------------------------------------------------
@@ -2340,8 +2589,6 @@ static struct qede_dev *qede_alloc_etherdev(struct qed_dev *cdev,
 	memset(&edev->stats, 0, sizeof(edev->stats));
 	memcpy(&edev->dev_info, info, sizeof(*info));
 
-	edev->num_tc = edev->dev_info.num_tc;
-
 	INIT_LIST_HEAD(&edev->vlan_list);
 
 	return edev;
@@ -2439,7 +2686,8 @@ static void qede_free_fp_array(struct qede_dev *edev)
 
 			kfree(fp->sb_info);
 			kfree(fp->rxq);
-			kfree(fp->txqs);
+			kfree(fp->xdp_tx);
+			kfree(fp->txq);
 		}
 		kfree(edev->fp_array);
 	}
@@ -2472,7 +2720,7 @@ static int qede_alloc_fp_array(struct qede_dev *edev)
 	for_each_queue(i) {
 		fp = &edev->fp_array[i];
 
-		fp->sb_info = kcalloc(1, sizeof(*fp->sb_info), GFP_KERNEL);
+		fp->sb_info = kzalloc(sizeof(*fp->sb_info), GFP_KERNEL);
 		if (!fp->sb_info) {
 			DP_NOTICE(edev, "sb info struct allocation failed\n");
 			goto err;
@@ -2489,21 +2737,22 @@ static int qede_alloc_fp_array(struct qede_dev *edev)
 		}
 
 		if (fp->type & QEDE_FASTPATH_TX) {
-			fp->txqs = kcalloc(edev->num_tc, sizeof(*fp->txqs),
-					   GFP_KERNEL);
-			if (!fp->txqs) {
-				DP_NOTICE(edev,
-					  "TXQ array allocation failed\n");
+			fp->txq = kzalloc(sizeof(*fp->txq), GFP_KERNEL);
+			if (!fp->txq)
 				goto err;
-			}
 		}
 
 		if (fp->type & QEDE_FASTPATH_RX) {
-			fp->rxq = kcalloc(1, sizeof(*fp->rxq), GFP_KERNEL);
-			if (!fp->rxq) {
-				DP_NOTICE(edev,
-					  "RXQ struct allocation failed\n");
+			fp->rxq = kzalloc(sizeof(*fp->rxq), GFP_KERNEL);
+			if (!fp->rxq)
 				goto err;
+
+			if (edev->xdp_prog) {
+				fp->xdp_tx = kzalloc(sizeof(*fp->xdp_tx),
+						     GFP_KERNEL);
+				if (!fp->xdp_tx)
+					goto err;
+				fp->type |= QEDE_FASTPATH_XDP;
 			}
 		}
 	}
@@ -2520,12 +2769,11 @@ static void qede_sp_task(struct work_struct *work)
 					     sp_task.work);
 	struct qed_dev *cdev = edev->cdev;
 
-	mutex_lock(&edev->qede_lock);
+	__qede_lock(edev);
 
-	if (edev->state == QEDE_STATE_OPEN) {
-		if (test_and_clear_bit(QEDE_SP_RX_MODE, &edev->sp_flags))
+	if (test_and_clear_bit(QEDE_SP_RX_MODE, &edev->sp_flags))
+		if (edev->state == QEDE_STATE_OPEN)
 			qede_config_rx_mode(edev->ndev);
-	}
 
 	if (test_and_clear_bit(QEDE_SP_VXLAN_PORT_CONFIG, &edev->sp_flags)) {
 		struct qed_tunn_params tunn_params;
@@ -2545,16 +2793,16 @@ static void qede_sp_task(struct work_struct *work)
 		qed_ops->tunn_config(cdev, &tunn_params);
 	}
 
-	mutex_unlock(&edev->qede_lock);
+	__qede_unlock(edev);
 }
 
 static void qede_update_pf_params(struct qed_dev *cdev)
 {
 	struct qed_pf_params pf_params;
 
-	/* 64 rx + 64 tx */
+	/* 64 rx + 64 tx + 64 XDP */
 	memset(&pf_params, 0, sizeof(struct qed_pf_params));
-	pf_params.eth_pf_params.num_cons = 128;
+	pf_params.eth_pf_params.num_cons = 192;
 	qed_ops->common->update_pf_params(cdev, &pf_params);
 }
 
@@ -2703,6 +2951,10 @@ static void __qede_remove(struct pci_dev *pdev, enum qede_remove_mode mode)
 
 	pci_set_drvdata(pdev, NULL);
 
+	/* Release edev's reference to XDP's bpf if such exist */
+	if (edev->xdp_prog)
+		bpf_prog_put(edev->xdp_prog);
+
 	free_netdev(ndev);
 
 	/* Use global ops since we've freed edev */
@@ -2807,7 +3059,7 @@ static void qede_free_rx_buffers(struct qede_dev *edev,
 		data = rx_buf->data;
 
 		dma_unmap_page(&edev->pdev->dev,
-			       rx_buf->mapping, PAGE_SIZE, DMA_FROM_DEVICE);
+			       rx_buf->mapping, PAGE_SIZE, rxq->data_direction);
 
 		rx_buf->data = NULL;
 		__free_page(data);
@@ -2823,7 +3075,7 @@ static void qede_free_sge_mem(struct qede_dev *edev, struct qede_rx_queue *rxq)
 
 	for (i = 0; i < ETH_TPA_MAX_AGGS_NUM; i++) {
 		struct qede_agg_info *tpa_info = &rxq->tpa_info[i];
-		struct sw_rx_data *replace_buf = &tpa_info->replace_buf;
+		struct sw_rx_data *replace_buf = &tpa_info->buffer;
 
 		if (replace_buf->data) {
 			dma_unmap_page(&edev->pdev->dev,
@@ -2849,52 +3101,15 @@ static void qede_free_mem_rxq(struct qede_dev *edev, struct qede_rx_queue *rxq)
 	edev->ops->common->chain_free(edev->cdev, &rxq->rx_comp_ring);
 }
 
-static int qede_alloc_rx_buffer(struct qede_dev *edev,
-				struct qede_rx_queue *rxq)
-{
-	struct sw_rx_data *sw_rx_data;
-	struct eth_rx_bd *rx_bd;
-	dma_addr_t mapping;
-	struct page *data;
-
-	data = alloc_pages(GFP_ATOMIC, 0);
-	if (unlikely(!data)) {
-		DP_NOTICE(edev, "Failed to allocate Rx data [page]\n");
-		return -ENOMEM;
-	}
-
-	/* Map the entire page as it would be used
-	 * for multiple RX buffer segment size mapping.
-	 */
-	mapping = dma_map_page(&edev->pdev->dev, data, 0,
-			       PAGE_SIZE, DMA_FROM_DEVICE);
-	if (unlikely(dma_mapping_error(&edev->pdev->dev, mapping))) {
-		__free_page(data);
-		DP_NOTICE(edev, "Failed to map Rx buffer\n");
-		return -ENOMEM;
-	}
-
-	sw_rx_data = &rxq->sw_rx_ring[rxq->sw_rx_prod & NUM_RX_BDS_MAX];
-	sw_rx_data->page_offset = 0;
-	sw_rx_data->data = data;
-	sw_rx_data->mapping = mapping;
-
-	/* Advance PROD and get BD pointer */
-	rx_bd = (struct eth_rx_bd *)qed_chain_produce(&rxq->rx_bd_ring);
-	WARN_ON(!rx_bd);
-	rx_bd->addr.hi = cpu_to_le32(upper_32_bits(mapping));
-	rx_bd->addr.lo = cpu_to_le32(lower_32_bits(mapping));
-
-	rxq->sw_rx_prod++;
-
-	return 0;
-}
-
 static int qede_alloc_sge_mem(struct qede_dev *edev, struct qede_rx_queue *rxq)
 {
 	dma_addr_t mapping;
 	int i;
 
+	/* Don't perform FW aggregations in case of XDP */
+	if (edev->xdp_prog)
+		edev->gro_disable = 1;
+
 	if (edev->gro_disable)
 		return 0;
 
@@ -2905,7 +3120,7 @@ static int qede_alloc_sge_mem(struct qede_dev *edev, struct qede_rx_queue *rxq)
 
 	for (i = 0; i < ETH_TPA_MAX_AGGS_NUM; i++) {
 		struct qede_agg_info *tpa_info = &rxq->tpa_info[i];
-		struct sw_rx_data *replace_buf = &tpa_info->replace_buf;
+		struct sw_rx_data *replace_buf = &tpa_info->buffer;
 
 		replace_buf->data = alloc_pages(GFP_ATOMIC, 0);
 		if (unlikely(!replace_buf->data)) {
@@ -2923,10 +3138,9 @@ static int qede_alloc_sge_mem(struct qede_dev *edev, struct qede_rx_queue *rxq)
 		}
 
 		replace_buf->mapping = mapping;
-		tpa_info->replace_buf.page_offset = 0;
-
-		tpa_info->replace_buf_mapping = mapping;
-		tpa_info->agg_state = QEDE_AGG_STATE_NONE;
+		tpa_info->buffer.page_offset = 0;
+		tpa_info->buffer_mapping = mapping;
+		tpa_info->state = QEDE_AGG_STATE_NONE;
 	}
 
 	return 0;
@@ -2948,8 +3162,13 @@ static int qede_alloc_mem_rxq(struct qede_dev *edev, struct qede_rx_queue *rxq)
 	if (rxq->rx_buf_size > PAGE_SIZE)
 		rxq->rx_buf_size = PAGE_SIZE;
 
-	/* Segment size to spilt a page in multiple equal parts */
-	rxq->rx_buf_seg_size = roundup_pow_of_two(rxq->rx_buf_size);
+	/* Segment size to spilt a page in multiple equal parts,
+	 * unless XDP is used in which case we'd use the entire page.
+	 */
+	if (!edev->xdp_prog)
+		rxq->rx_buf_seg_size = roundup_pow_of_two(rxq->rx_buf_size);
+	else
+		rxq->rx_buf_seg_size = PAGE_SIZE;
 
 	/* Allocate the parallel driver ring for Rx buffers */
 	size = sizeof(*rxq->sw_rx_ring) * RX_RING_SIZE;
@@ -2985,7 +3204,7 @@ static int qede_alloc_mem_rxq(struct qede_dev *edev, struct qede_rx_queue *rxq)
 
 	/* Allocate buffers for the Rx ring */
 	for (i = 0; i < rxq->num_rx_buffers; i++) {
-		rc = qede_alloc_rx_buffer(edev, rxq);
+		rc = qede_alloc_rx_buffer(rxq);
 		if (rc) {
 			DP_ERR(edev,
 			       "Rx buffers allocation failed at index %d\n", i);
@@ -3001,7 +3220,10 @@ err:
 static void qede_free_mem_txq(struct qede_dev *edev, struct qede_tx_queue *txq)
 {
 	/* Free the parallel SW ring */
-	kfree(txq->sw_tx_ring);
+	if (txq->is_xdp)
+		kfree(txq->sw_tx_ring.pages);
+	else
+		kfree(txq->sw_tx_ring.skbs);
 
 	/* Free the real RQ ring used by FW */
 	edev->ops->common->chain_free(edev->cdev, &txq->tx_pbl);
@@ -3010,17 +3232,22 @@ static void qede_free_mem_txq(struct qede_dev *edev, struct qede_tx_queue *txq)
 /* This function allocates all memory needed per Tx queue */
 static int qede_alloc_mem_txq(struct qede_dev *edev, struct qede_tx_queue *txq)
 {
-	int size, rc;
 	union eth_tx_bd_types *p_virt;
+	int size, rc;
 
 	txq->num_tx_buffers = edev->q_num_tx_buffers;
 
 	/* Allocate the parallel driver ring for Tx buffers */
-	size = sizeof(*txq->sw_tx_ring) * TX_RING_SIZE;
-	txq->sw_tx_ring = kzalloc(size, GFP_KERNEL);
-	if (!txq->sw_tx_ring) {
-		DP_NOTICE(edev, "Tx buffers ring allocation failed\n");
-		goto err;
+	if (txq->is_xdp) {
+		size = sizeof(*txq->sw_tx_ring.pages) * TX_RING_SIZE;
+		txq->sw_tx_ring.pages = kzalloc(size, GFP_KERNEL);
+		if (!txq->sw_tx_ring.pages)
+			goto err;
+	} else {
+		size = sizeof(*txq->sw_tx_ring.skbs) * TX_RING_SIZE;
+		txq->sw_tx_ring.skbs = kzalloc(size, GFP_KERNEL);
+		if (!txq->sw_tx_ring.skbs)
+			goto err;
 	}
 
 	rc = edev->ops->common->chain_alloc(edev->cdev,
@@ -3042,16 +3269,13 @@ err:
 /* This function frees all memory of a single fp */
 static void qede_free_mem_fp(struct qede_dev *edev, struct qede_fastpath *fp)
 {
-	int tc;
-
 	qede_free_mem_sb(edev, fp->sb_info);
 
 	if (fp->type & QEDE_FASTPATH_RX)
 		qede_free_mem_rxq(edev, fp->rxq);
 
 	if (fp->type & QEDE_FASTPATH_TX)
-		for (tc = 0; tc < edev->num_tc; tc++)
-			qede_free_mem_txq(edev, &fp->txqs[tc]);
+		qede_free_mem_txq(edev, fp->txq);
 }
 
 /* This function allocates all memory needed for a single fp (i.e. an entity
@@ -3059,28 +3283,31 @@ static void qede_free_mem_fp(struct qede_dev *edev, struct qede_fastpath *fp)
  */
 static int qede_alloc_mem_fp(struct qede_dev *edev, struct qede_fastpath *fp)
 {
-	int rc, tc;
+	int rc = 0;
 
 	rc = qede_alloc_mem_sb(edev, fp->sb_info, fp->id);
 	if (rc)
-		goto err;
+		goto out;
 
 	if (fp->type & QEDE_FASTPATH_RX) {
 		rc = qede_alloc_mem_rxq(edev, fp->rxq);
 		if (rc)
-			goto err;
+			goto out;
+	}
+
+	if (fp->type & QEDE_FASTPATH_XDP) {
+		rc = qede_alloc_mem_txq(edev, fp->xdp_tx);
+		if (rc)
+			goto out;
 	}
 
 	if (fp->type & QEDE_FASTPATH_TX) {
-		for (tc = 0; tc < edev->num_tc; tc++) {
-			rc = qede_alloc_mem_txq(edev, &fp->txqs[tc]);
-			if (rc)
-				goto err;
-		}
+		rc = qede_alloc_mem_txq(edev, fp->txq);
+		if (rc)
+			goto out;
 	}
 
-	return 0;
-err:
+out:
 	return rc;
 }
 
@@ -3119,7 +3346,7 @@ static int qede_alloc_mem_load(struct qede_dev *edev)
 /* This function inits fp content and resets the SB, RXQ and TXQ structures */
 static void qede_init_fp(struct qede_dev *edev)
 {
-	int queue_id, rxq_index = 0, txq_index = 0, tc;
+	int queue_id, rxq_index = 0, txq_index = 0;
 	struct qede_fastpath *fp;
 
 	for_each_queue(queue_id) {
@@ -3128,25 +3355,28 @@ static void qede_init_fp(struct qede_dev *edev)
 		fp->edev = edev;
 		fp->id = queue_id;
 
-		memset((void *)&fp->napi, 0, sizeof(fp->napi));
-
-		memset((void *)fp->sb_info, 0, sizeof(*fp->sb_info));
+		if (fp->type & QEDE_FASTPATH_XDP) {
+			fp->xdp_tx->index = QEDE_TXQ_IDX_TO_XDP(edev,
+								rxq_index);
+			fp->xdp_tx->is_xdp = 1;
+		}
 
 		if (fp->type & QEDE_FASTPATH_RX) {
-			memset((void *)fp->rxq, 0, sizeof(*fp->rxq));
 			fp->rxq->rxq_id = rxq_index++;
+
+			/* Determine how to map buffers for this queue */
+			if (fp->type & QEDE_FASTPATH_XDP)
+				fp->rxq->data_direction = DMA_BIDIRECTIONAL;
+			else
+				fp->rxq->data_direction = DMA_FROM_DEVICE;
+			fp->rxq->dev = &edev->pdev->dev;
 		}
 
 		if (fp->type & QEDE_FASTPATH_TX) {
-			memset((void *)fp->txqs, 0,
-			       (edev->num_tc * sizeof(*fp->txqs)));
-			for (tc = 0; tc < edev->num_tc; tc++) {
-				fp->txqs[tc].index = txq_index +
-				    tc * QEDE_TSS_COUNT(edev);
-				if (edev->dev_info.is_legacy)
-					fp->txqs[tc].is_legacy = true;
-			}
-			txq_index++;
+			fp->txq->index = txq_index++;
+			if (edev->dev_info.is_legacy)
+				fp->txq->is_legacy = 1;
+			fp->txq->dev = &edev->pdev->dev;
 		}
 
 		snprintf(fp->name, sizeof(fp->name), "%s-fp-%d",
@@ -3314,11 +3544,18 @@ static int qede_drain_txq(struct qede_dev *edev,
 	return 0;
 }
 
+static int qede_stop_txq(struct qede_dev *edev,
+			 struct qede_tx_queue *txq, int rss_id)
+{
+	return edev->ops->q_tx_stop(edev->cdev, rss_id, txq->handle);
+}
+
 static int qede_stop_queues(struct qede_dev *edev)
 {
 	struct qed_update_vport_params vport_update_params;
 	struct qed_dev *cdev = edev->cdev;
-	int rc, tc, i;
+	struct qede_fastpath *fp;
+	int rc, i;
 
 	/* Disable the vport */
 	memset(&vport_update_params, 0, sizeof(vport_update_params));
@@ -3335,53 +3572,49 @@ static int qede_stop_queues(struct qede_dev *edev)
 
 	/* Flush Tx queues. If needed, request drain from MCP */
 	for_each_queue(i) {
-		struct qede_fastpath *fp = &edev->fp_array[i];
+		fp = &edev->fp_array[i];
 
 		if (fp->type & QEDE_FASTPATH_TX) {
-			for (tc = 0; tc < edev->num_tc; tc++) {
-				struct qede_tx_queue *txq = &fp->txqs[tc];
+			rc = qede_drain_txq(edev, fp->txq, true);
+			if (rc)
+				return rc;
+		}
 
-				rc = qede_drain_txq(edev, txq, true);
-				if (rc)
-					return rc;
-			}
+		if (fp->type & QEDE_FASTPATH_XDP) {
+			rc = qede_drain_txq(edev, fp->xdp_tx, true);
+			if (rc)
+				return rc;
 		}
 	}
 
 	/* Stop all Queues in reverse order */
 	for (i = QEDE_QUEUE_CNT(edev) - 1; i >= 0; i--) {
-		struct qed_stop_rxq_params rx_params;
+		fp = &edev->fp_array[i];
 
 		/* Stop the Tx Queue(s) */
-		if (edev->fp_array[i].type & QEDE_FASTPATH_TX) {
-			for (tc = 0; tc < edev->num_tc; tc++) {
-				struct qed_stop_txq_params tx_params;
-				u8 val;
-
-				tx_params.rss_id = i;
-				val = edev->fp_array[i].txqs[tc].index;
-				tx_params.tx_queue_id = val;
-				rc = edev->ops->q_tx_stop(cdev, &tx_params);
-				if (rc) {
-					DP_ERR(edev, "Failed to stop TXQ #%d\n",
-					       tx_params.tx_queue_id);
-					return rc;
-				}
-			}
+		if (fp->type & QEDE_FASTPATH_TX) {
+			rc = qede_stop_txq(edev, fp->txq, i);
+			if (rc)
+				return rc;
 		}
 
 		/* Stop the Rx Queue */
-		if (edev->fp_array[i].type & QEDE_FASTPATH_RX) {
-			memset(&rx_params, 0, sizeof(rx_params));
-			rx_params.rss_id = i;
-			rx_params.rx_queue_id = edev->fp_array[i].rxq->rxq_id;
-
-			rc = edev->ops->q_rx_stop(cdev, &rx_params);
+		if (fp->type & QEDE_FASTPATH_RX) {
+			rc = edev->ops->q_rx_stop(cdev, i, fp->rxq->handle);
 			if (rc) {
 				DP_ERR(edev, "Failed to stop RXQ #%d\n", i);
 				return rc;
 			}
 		}
+
+		/* Stop the XDP forwarding queue */
+		if (fp->type & QEDE_FASTPATH_XDP) {
+			rc = qede_stop_txq(edev, fp->xdp_tx, i);
+			if (rc)
+				return rc;
+
+			bpf_prog_put(fp->rxq->xdp_prog);
+		}
 	}
 
 	/* Stop the vport */
@@ -3392,9 +3625,55 @@ static int qede_stop_queues(struct qede_dev *edev)
 	return rc;
 }
 
+static int qede_start_txq(struct qede_dev *edev,
+			  struct qede_fastpath *fp,
+			  struct qede_tx_queue *txq, u8 rss_id, u16 sb_idx)
+{
+	dma_addr_t phys_table = qed_chain_get_pbl_phys(&txq->tx_pbl);
+	u32 page_cnt = qed_chain_get_page_cnt(&txq->tx_pbl);
+	struct qed_queue_start_common_params params;
+	struct qed_txq_start_ret_params ret_params;
+	int rc;
+
+	memset(&params, 0, sizeof(params));
+	memset(&ret_params, 0, sizeof(ret_params));
+
+	/* Let the XDP queue share the queue-zone with one of the regular txq.
+	 * We don't really care about its coalescing.
+	 */
+	if (txq->is_xdp)
+		params.queue_id = QEDE_TXQ_XDP_TO_IDX(edev, txq);
+	else
+		params.queue_id = txq->index;
+
+	params.sb = fp->sb_info->igu_sb_id;
+	params.sb_idx = sb_idx;
+
+	rc = edev->ops->q_tx_start(edev->cdev, rss_id, &params, phys_table,
+				   page_cnt, &ret_params);
+	if (rc) {
+		DP_ERR(edev, "Start TXQ #%d failed %d\n", txq->index, rc);
+		return rc;
+	}
+
+	txq->doorbell_addr = ret_params.p_doorbell;
+	txq->handle = ret_params.p_handle;
+
+	/* Determine the FW consumer address associated */
+	txq->hw_cons_ptr = &fp->sb_info->sb_virt->pi_array[sb_idx];
+
+	/* Prepare the doorbell parameters */
+	SET_FIELD(txq->tx_db.data.params, ETH_DB_DATA_DEST, DB_DEST_XCM);
+	SET_FIELD(txq->tx_db.data.params, ETH_DB_DATA_AGG_CMD, DB_AGG_CMD_SET);
+	SET_FIELD(txq->tx_db.data.params, ETH_DB_DATA_AGG_VAL_SEL,
+		  DQ_XCM_ETH_TX_BD_PROD_CMD);
+	txq->tx_db.data.agg_flags = DQ_XCM_ETH_DQ_CF_CMD;
+
+	return rc;
+}
+
 static int qede_start_queues(struct qede_dev *edev, bool clear_stats)
 {
-	int rc, tc, i;
 	int vlan_removal_en = 1;
 	struct qed_dev *cdev = edev->cdev;
 	struct qed_update_vport_params vport_update_params;
@@ -3402,6 +3681,7 @@ static int qede_start_queues(struct qede_dev *edev, bool clear_stats)
 	struct qed_dev_info *qed_info = &edev->dev_info.common;
 	struct qed_start_vport_params start = {0};
 	bool reset_rss_indir = false;
+	int rc, i;
 
 	if (!edev->num_queues) {
 		DP_ERR(edev,
@@ -3433,11 +3713,12 @@ static int qede_start_queues(struct qede_dev *edev, bool clear_stats)
 		u32 page_cnt;
 
 		if (fp->type & QEDE_FASTPATH_RX) {
+			struct qed_rxq_start_ret_params ret_params;
 			struct qede_rx_queue *rxq = fp->rxq;
 			__le16 *val;
 
+			memset(&ret_params, 0, sizeof(ret_params));
 			memset(&q_params, 0, sizeof(q_params));
-			q_params.rss_id = i;
 			q_params.queue_id = rxq->rxq_id;
 			q_params.vport_id = 0;
 			q_params.sb = fp->sb_info->igu_sb_id;
@@ -3447,60 +3728,44 @@ static int qede_start_queues(struct qede_dev *edev, bool clear_stats)
 			    qed_chain_get_pbl_phys(&rxq->rx_comp_ring);
 			page_cnt = qed_chain_get_page_cnt(&rxq->rx_comp_ring);
 
-			rc = edev->ops->q_rx_start(cdev, &q_params,
+			rc = edev->ops->q_rx_start(cdev, i, &q_params,
 						   rxq->rx_buf_size,
 						   rxq->rx_bd_ring.p_phys_addr,
 						   p_phys_table,
-						   page_cnt,
-						   &rxq->hw_rxq_prod_addr);
+						   page_cnt, &ret_params);
 			if (rc) {
 				DP_ERR(edev, "Start RXQ #%d failed %d\n", i,
 				       rc);
 				return rc;
 			}
 
+			/* Use the return parameters */
+			rxq->hw_rxq_prod_addr = ret_params.p_prod;
+			rxq->handle = ret_params.p_handle;
+
 			val = &fp->sb_info->sb_virt->pi_array[RX_PI];
 			rxq->hw_cons_ptr = val;
 
 			qede_update_rx_prod(edev, rxq);
 		}
 
-		if (!(fp->type & QEDE_FASTPATH_TX))
-			continue;
-
-		for (tc = 0; tc < edev->num_tc; tc++) {
-			struct qede_tx_queue *txq = &fp->txqs[tc];
-
-			p_phys_table = qed_chain_get_pbl_phys(&txq->tx_pbl);
-			page_cnt = qed_chain_get_page_cnt(&txq->tx_pbl);
-
-			memset(&q_params, 0, sizeof(q_params));
-			q_params.rss_id = i;
-			q_params.queue_id = txq->index;
-			q_params.vport_id = 0;
-			q_params.sb = fp->sb_info->igu_sb_id;
-			q_params.sb_idx = TX_PI(tc);
+		if (fp->type & QEDE_FASTPATH_XDP) {
+			rc = qede_start_txq(edev, fp, fp->xdp_tx, i, XDP_PI);
+			if (rc)
+				return rc;
 
-			rc = edev->ops->q_tx_start(cdev, &q_params,
-						   p_phys_table, page_cnt,
-						   &txq->doorbell_addr);
-			if (rc) {
-				DP_ERR(edev, "Start TXQ #%d failed %d\n",
-				       txq->index, rc);
+			fp->rxq->xdp_prog = bpf_prog_add(edev->xdp_prog, 1);
+			if (IS_ERR(fp->rxq->xdp_prog)) {
+				rc = PTR_ERR(fp->rxq->xdp_prog);
+				fp->rxq->xdp_prog = NULL;
 				return rc;
 			}
+		}
 
-			txq->hw_cons_ptr =
-				&fp->sb_info->sb_virt->pi_array[TX_PI(tc)];
-			SET_FIELD(txq->tx_db.data.params,
-				  ETH_DB_DATA_DEST, DB_DEST_XCM);
-			SET_FIELD(txq->tx_db.data.params, ETH_DB_DATA_AGG_CMD,
-				  DB_AGG_CMD_SET);
-			SET_FIELD(txq->tx_db.data.params,
-				  ETH_DB_DATA_AGG_VAL_SEL,
-				  DQ_XCM_ETH_TX_BD_PROD_CMD);
-
-			txq->tx_db.data.agg_flags = DQ_XCM_ETH_DQ_CF_CMD;
+		if (fp->type & QEDE_FASTPATH_TX) {
+			rc = qede_start_txq(edev, fp, fp->txq, i, TX_PI(0));
+			if (rc)
+				return rc;
 		}
 	}
 
@@ -3595,15 +3860,18 @@ enum qede_unload_mode {
 	QEDE_UNLOAD_NORMAL,
 };
 
-static void qede_unload(struct qede_dev *edev, enum qede_unload_mode mode)
+static void qede_unload(struct qede_dev *edev, enum qede_unload_mode mode,
+			bool is_locked)
 {
 	struct qed_link_params link_params;
 	int rc;
 
 	DP_INFO(edev, "Starting qede unload\n");
 
+	if (!is_locked)
+		__qede_lock(edev);
+
 	qede_roce_dev_event_close(edev);
-	mutex_lock(&edev->qede_lock);
 	edev->state = QEDE_STATE_CLOSED;
 
 	/* Close OS Tx */
@@ -3635,7 +3903,8 @@ static void qede_unload(struct qede_dev *edev, enum qede_unload_mode mode)
 	qede_free_fp_array(edev);
 
 out:
-	mutex_unlock(&edev->qede_lock);
+	if (!is_locked)
+		__qede_unlock(edev);
 	DP_INFO(edev, "Ending qede unload\n");
 }
 
@@ -3644,7 +3913,8 @@ enum qede_load_mode {
 	QEDE_LOAD_RELOAD,
 };
 
-static int qede_load(struct qede_dev *edev, enum qede_load_mode mode)
+static int qede_load(struct qede_dev *edev, enum qede_load_mode mode,
+		     bool is_locked)
 {
 	struct qed_link_params link_params;
 	struct qed_link_output link_output;
@@ -3652,21 +3922,24 @@ static int qede_load(struct qede_dev *edev, enum qede_load_mode mode)
 
 	DP_INFO(edev, "Starting qede load\n");
 
+	if (!is_locked)
+		__qede_lock(edev);
+
 	rc = qede_set_num_queues(edev);
 	if (rc)
-		goto err0;
+		goto out;
 
 	rc = qede_alloc_fp_array(edev);
 	if (rc)
-		goto err0;
+		goto out;
 
 	qede_init_fp(edev);
 
 	rc = qede_alloc_mem_load(edev);
 	if (rc)
 		goto err1;
-	DP_INFO(edev, "Allocated %d RSS queues on %d TC/s\n",
-		QEDE_QUEUE_CNT(edev), edev->num_tc);
+	DP_INFO(edev, "Allocated %d Rx, %d Tx queues\n",
+		QEDE_RSS_COUNT(edev), QEDE_TSS_COUNT(edev));
 
 	rc = qede_set_real_num_queues(edev);
 	if (rc)
@@ -3688,10 +3961,6 @@ static int qede_load(struct qede_dev *edev, enum qede_load_mode mode)
 	/* Add primary mac and set Rx filters */
 	ether_addr_copy(edev->primary_mac, edev->ndev->dev_addr);
 
-	mutex_lock(&edev->qede_lock);
-	edev->state = QEDE_STATE_OPEN;
-	mutex_unlock(&edev->qede_lock);
-
 	/* Program un-configured VLANs */
 	qede_configure_vlan_filters(edev);
 
@@ -3706,10 +3975,12 @@ static int qede_load(struct qede_dev *edev, enum qede_load_mode mode)
 	qede_roce_dev_event_open(edev);
 	qede_link_update(edev, &link_output);
 
+	edev->state = QEDE_STATE_OPEN;
+
 	DP_INFO(edev, "Ending successfully qede load\n");
 
-	return 0;
 
+	goto out;
 err4:
 	qede_sync_free_irqs(edev);
 	memset(&edev->int_info.msix_cnt, 0, sizeof(struct qed_int_info));
@@ -3723,26 +3994,40 @@ err1:
 	edev->num_queues = 0;
 	edev->fp_num_tx = 0;
 	edev->fp_num_rx = 0;
-err0:
+out:
+	if (!is_locked)
+		__qede_unlock(edev);
+
 	return rc;
 }
 
+/* 'func' should be able to run between unload and reload assuming interface
+ * is actually running, or afterwards in case it's currently DOWN.
+ */
 void qede_reload(struct qede_dev *edev,
-		 void (*func)(struct qede_dev *, union qede_reload_args *),
-		 union qede_reload_args *args)
+		 struct qede_reload_args *args, bool is_locked)
 {
-	qede_unload(edev, QEDE_UNLOAD_NORMAL);
-	/* Call function handler to update parameters
-	 * needed for function load.
-	 */
-	if (func)
-		func(edev, args);
+	if (!is_locked)
+		__qede_lock(edev);
 
-	qede_load(edev, QEDE_LOAD_RELOAD);
+	/* Since qede_lock is held, internal state wouldn't change even
+	 * if netdev state would start transitioning. Check whether current
+	 * internal configuration indicates device is up, then reload.
+	 */
+	if (edev->state == QEDE_STATE_OPEN) {
+		qede_unload(edev, QEDE_UNLOAD_NORMAL, true);
+		if (args)
+			args->func(edev, args);
+		qede_load(edev, QEDE_LOAD_RELOAD, true);
+
+		/* Since no one is going to do it for us, re-configure */
+		qede_config_rx_mode(edev->ndev);
+	} else if (args) {
+		args->func(edev, args);
+	}
 
-	mutex_lock(&edev->qede_lock);
-	qede_config_rx_mode(edev->ndev);
-	mutex_unlock(&edev->qede_lock);
+	if (!is_locked)
+		__qede_unlock(edev);
 }
 
 /* called with rtnl_lock */
@@ -3755,8 +4040,7 @@ static int qede_open(struct net_device *ndev)
 
 	edev->ops->common->set_power_state(edev->cdev, PCI_D0);
 
-	rc = qede_load(edev, QEDE_LOAD_NORMAL);
-
+	rc = qede_load(edev, QEDE_LOAD_NORMAL, false);
 	if (rc)
 		return rc;
 
@@ -3771,7 +4055,7 @@ static int qede_close(struct net_device *ndev)
 {
 	struct qede_dev *edev = netdev_priv(ndev);
 
-	qede_unload(edev, QEDE_UNLOAD_NORMAL);
+	qede_unload(edev, QEDE_UNLOAD_NORMAL, false);
 
 	edev->ops->common->update_drv_state(edev->cdev, false);
 
@@ -3903,15 +4187,8 @@ static void qede_set_rx_mode(struct net_device *ndev)
 {
 	struct qede_dev *edev = netdev_priv(ndev);
 
-	DP_INFO(edev, "qede_set_rx_mode called\n");
-
-	if (edev->state != QEDE_STATE_OPEN) {
-		DP_INFO(edev,
-			"qede_set_rx_mode called while interface is down\n");
-	} else {
-		set_bit(QEDE_SP_RX_MODE, &edev->sp_flags);
-		schedule_delayed_work(&edev->sp_task, 0);
-	}
+	set_bit(QEDE_SP_RX_MODE, &edev->sp_flags);
+	schedule_delayed_work(&edev->sp_task, 0);
 }
 
 /* Must be called with qede_lock held */
diff --git a/include/linux/qed/qed_chain.h b/include/linux/qed/qed_chain.h
index 72d88cf3ca25..37dfba101c6c 100644
--- a/include/linux/qed/qed_chain.h
+++ b/include/linux/qed/qed_chain.h
@@ -56,23 +56,6 @@ struct qed_chain_pbl_u32 {
 	u32 cons_page_idx;
 };
 
-struct qed_chain_pbl {
-	/* Base address of a pre-allocated buffer for pbl */
-	dma_addr_t	p_phys_table;
-	void		*p_virt_table;
-
-	/* Table for keeping the virtual addresses of the chain pages,
-	 * respectively to the physical addresses in the pbl table.
-	 */
-	void **pp_virt_addr_tbl;
-
-	/* Index to current used page by producer/consumer */
-	union {
-		struct qed_chain_pbl_u16 pbl16;
-		struct qed_chain_pbl_u32 pbl32;
-	} u;
-};
-
 struct qed_chain_u16 {
 	/* Cyclic index of next element to produce/consme */
 	u16 prod_idx;
@@ -86,46 +69,78 @@ struct qed_chain_u32 {
 };
 
 struct qed_chain {
-	void			*p_virt_addr;
-	dma_addr_t		p_phys_addr;
-	void			*p_prod_elem;
-	void			*p_cons_elem;
+	/* fastpath portion of the chain - required for commands such
+	 * as produce / consume.
+	 */
+	/* Point to next element to produce/consume */
+	void *p_prod_elem;
+	void *p_cons_elem;
+
+	/* Fastpath portions of the PBL [if exists] */
+	struct {
+		/* Table for keeping the virtual addresses of the chain pages,
+		 * respectively to the physical addresses in the pbl table.
+		 */
+		void **pp_virt_addr_tbl;
 
-	enum qed_chain_mode	mode;
-	enum qed_chain_use_mode intended_use; /* used to produce/consume */
-	enum qed_chain_cnt_type cnt_type;
+		union {
+			struct qed_chain_pbl_u16 u16;
+			struct qed_chain_pbl_u32 u32;
+		} c;
+	} pbl;
 
 	union {
 		struct qed_chain_u16 chain16;
 		struct qed_chain_u32 chain32;
 	} u;
 
+	/* Capacity counts only usable elements */
+	u32 capacity;
 	u32 page_cnt;
 
-	/* Number of elements - capacity is for usable elements only,
-	 * while size will contain total number of elements [for entire chain].
+	enum qed_chain_mode mode;
+
+	/* Elements information for fast calculations */
+	u16 elem_per_page;
+	u16 elem_per_page_mask;
+	u16 elem_size;
+	u16 next_page_mask;
+	u16 usable_per_page;
+	u8 elem_unusable;
+
+	u8 cnt_type;
+
+	/* Slowpath of the chain - required for initialization and destruction,
+	 * but isn't involved in regular functionality.
 	 */
-	u32 capacity;
+
+	/* Base address of a pre-allocated buffer for pbl */
+	struct {
+		dma_addr_t p_phys_table;
+		void *p_virt_table;
+	} pbl_sp;
+
+	/* Address of first page of the chain - the address is required
+	 * for fastpath operation [consume/produce] but only for the the SINGLE
+	 * flavour which isn't considered fastpath [== SPQ].
+	 */
+	void *p_virt_addr;
+	dma_addr_t p_phys_addr;
+
+	/* Total number of elements [for entire chain] */
 	u32 size;
 
-	/* Elements information for fast calculations */
-	u16			elem_per_page;
-	u16			elem_per_page_mask;
-	u16			elem_unusable;
-	u16			usable_per_page;
-	u16			elem_size;
-	u16			next_page_mask;
-	struct qed_chain_pbl	pbl;
+	u8 intended_use;
 };
 
 #define QED_CHAIN_PBL_ENTRY_SIZE        (8)
 #define QED_CHAIN_PAGE_SIZE             (0x1000)
 #define ELEMS_PER_PAGE(elem_size)       (QED_CHAIN_PAGE_SIZE / (elem_size))
 
-#define UNUSABLE_ELEMS_PER_PAGE(elem_size, mode)     \
-	((mode == QED_CHAIN_MODE_NEXT_PTR) ?	     \
-	 (1 + ((sizeof(struct qed_chain_next) - 1) / \
-	       (elem_size))) : 0)
+#define UNUSABLE_ELEMS_PER_PAGE(elem_size, mode)	 \
+	(((mode) == QED_CHAIN_MODE_NEXT_PTR) ?		 \
+	 (u8)(1 + ((sizeof(struct qed_chain_next) - 1) / \
+		   (elem_size))) : 0)
 
 #define USABLE_ELEMS_PER_PAGE(elem_size, mode) \
 	((u32)(ELEMS_PER_PAGE(elem_size) -     \
@@ -186,7 +201,7 @@ static inline u16 qed_chain_get_usable_per_page(struct qed_chain *p_chain)
 	return p_chain->usable_per_page;
 }
 
-static inline u16 qed_chain_get_unusable_per_page(struct qed_chain *p_chain)
+static inline u8 qed_chain_get_unusable_per_page(struct qed_chain *p_chain)
 {
 	return p_chain->elem_unusable;
 }
@@ -198,7 +213,7 @@ static inline u32 qed_chain_get_page_cnt(struct qed_chain *p_chain)
 
 static inline dma_addr_t qed_chain_get_pbl_phys(struct qed_chain *p_chain)
 {
-	return p_chain->pbl.p_phys_table;
+	return p_chain->pbl_sp.p_phys_table;
 }
 
 /**
@@ -214,10 +229,10 @@ static inline dma_addr_t qed_chain_get_pbl_phys(struct qed_chain *p_chain)
 static inline void
 qed_chain_advance_page(struct qed_chain *p_chain,
 		       void **p_next_elem, void *idx_to_inc, void *page_to_inc)
-
 {
 	struct qed_chain_next *p_next = NULL;
 	u32 page_index = 0;
+
 	switch (p_chain->mode) {
 	case QED_CHAIN_MODE_NEXT_PTR:
 		p_next = *p_next_elem;
@@ -305,7 +320,7 @@ static inline void *qed_chain_produce(struct qed_chain *p_chain)
 		if ((p_chain->u.chain16.prod_idx &
 		     p_chain->elem_per_page_mask) == p_chain->next_page_mask) {
 			p_prod_idx = &p_chain->u.chain16.prod_idx;
-			p_prod_page_idx = &p_chain->pbl.u.pbl16.prod_page_idx;
+			p_prod_page_idx = &p_chain->pbl.c.u16.prod_page_idx;
 			qed_chain_advance_page(p_chain, &p_chain->p_prod_elem,
 					       p_prod_idx, p_prod_page_idx);
 		}
@@ -314,7 +329,7 @@ static inline void *qed_chain_produce(struct qed_chain *p_chain)
 		if ((p_chain->u.chain32.prod_idx &
 		     p_chain->elem_per_page_mask) == p_chain->next_page_mask) {
 			p_prod_idx = &p_chain->u.chain32.prod_idx;
-			p_prod_page_idx = &p_chain->pbl.u.pbl32.prod_page_idx;
+			p_prod_page_idx = &p_chain->pbl.c.u32.prod_page_idx;
 			qed_chain_advance_page(p_chain, &p_chain->p_prod_elem,
 					       p_prod_idx, p_prod_page_idx);
 		}
@@ -378,7 +393,7 @@ static inline void *qed_chain_consume(struct qed_chain *p_chain)
 		if ((p_chain->u.chain16.cons_idx &
 		     p_chain->elem_per_page_mask) == p_chain->next_page_mask) {
 			p_cons_idx = &p_chain->u.chain16.cons_idx;
-			p_cons_page_idx = &p_chain->pbl.u.pbl16.cons_page_idx;
+			p_cons_page_idx = &p_chain->pbl.c.u16.cons_page_idx;
 			qed_chain_advance_page(p_chain, &p_chain->p_cons_elem,
 					       p_cons_idx, p_cons_page_idx);
 		}
@@ -387,8 +402,8 @@ static inline void *qed_chain_consume(struct qed_chain *p_chain)
 		if ((p_chain->u.chain32.cons_idx &
 		     p_chain->elem_per_page_mask) == p_chain->next_page_mask) {
 			p_cons_idx = &p_chain->u.chain32.cons_idx;
-			p_cons_page_idx = &p_chain->pbl.u.pbl32.cons_page_idx;
-		qed_chain_advance_page(p_chain, &p_chain->p_cons_elem,
+			p_cons_page_idx = &p_chain->pbl.c.u32.cons_page_idx;
+			qed_chain_advance_page(p_chain, &p_chain->p_cons_elem,
 					       p_cons_idx, p_cons_page_idx);
 		}
 		p_chain->u.chain32.cons_idx++;
@@ -429,25 +444,26 @@ static inline void qed_chain_reset(struct qed_chain *p_chain)
 		u32 reset_val = p_chain->page_cnt - 1;
 
 		if (is_chain_u16(p_chain)) {
-			p_chain->pbl.u.pbl16.prod_page_idx = (u16)reset_val;
-			p_chain->pbl.u.pbl16.cons_page_idx = (u16)reset_val;
+			p_chain->pbl.c.u16.prod_page_idx = (u16)reset_val;
+			p_chain->pbl.c.u16.cons_page_idx = (u16)reset_val;
 		} else {
-			p_chain->pbl.u.pbl32.prod_page_idx = reset_val;
-			p_chain->pbl.u.pbl32.cons_page_idx = reset_val;
+			p_chain->pbl.c.u32.prod_page_idx = reset_val;
+			p_chain->pbl.c.u32.cons_page_idx = reset_val;
 		}
 	}
 
 	switch (p_chain->intended_use) {
-	case QED_CHAIN_USE_TO_CONSUME_PRODUCE:
-	case QED_CHAIN_USE_TO_PRODUCE:
-		/* Do nothing */
-		break;
-
 	case QED_CHAIN_USE_TO_CONSUME:
 		/* produce empty elements */
 		for (i = 0; i < p_chain->capacity; i++)
 			qed_chain_recycle_consumed(p_chain);
 		break;
+
+	case QED_CHAIN_USE_TO_CONSUME_PRODUCE:
+	case QED_CHAIN_USE_TO_PRODUCE:
+	default:
+		/* Do nothing */
+		break;
 	}
 }
 
@@ -473,13 +489,13 @@ static inline void qed_chain_init_params(struct qed_chain *p_chain,
 	p_chain->p_virt_addr = NULL;
 	p_chain->p_phys_addr = 0;
 	p_chain->elem_size	= elem_size;
-	p_chain->intended_use = intended_use;
+	p_chain->intended_use = (u8)intended_use;
 	p_chain->mode		= mode;
-	p_chain->cnt_type = cnt_type;
+	p_chain->cnt_type = (u8)cnt_type;
 
-	p_chain->elem_per_page		= ELEMS_PER_PAGE(elem_size);
+	p_chain->elem_per_page = ELEMS_PER_PAGE(elem_size);
 	p_chain->usable_per_page = USABLE_ELEMS_PER_PAGE(elem_size, mode);
-	p_chain->elem_per_page_mask	= p_chain->elem_per_page - 1;
+	p_chain->elem_per_page_mask = p_chain->elem_per_page - 1;
 	p_chain->elem_unusable = UNUSABLE_ELEMS_PER_PAGE(elem_size, mode);
 	p_chain->next_page_mask = (p_chain->usable_per_page &
 				   p_chain->elem_per_page_mask);
@@ -488,8 +504,8 @@ static inline void qed_chain_init_params(struct qed_chain *p_chain,
 	p_chain->capacity = p_chain->usable_per_page * page_cnt;
 	p_chain->size = p_chain->elem_per_page * page_cnt;
 
-	p_chain->pbl.p_phys_table = 0;
-	p_chain->pbl.p_virt_table = NULL;
+	p_chain->pbl_sp.p_phys_table = 0;
+	p_chain->pbl_sp.p_virt_table = NULL;
 	p_chain->pbl.pp_virt_addr_tbl = NULL;
 }
 
@@ -530,8 +546,8 @@ static inline void qed_chain_init_pbl_mem(struct qed_chain *p_chain,
 					  dma_addr_t p_phys_pbl,
 					  void **pp_virt_addr_tbl)
 {
-	p_chain->pbl.p_phys_table = p_phys_pbl;
-	p_chain->pbl.p_virt_table = p_virt_pbl;
+	p_chain->pbl_sp.p_phys_table = p_phys_pbl;
+	p_chain->pbl_sp.p_virt_table = p_virt_pbl;
 	p_chain->pbl.pp_virt_addr_tbl = pp_virt_addr_tbl;
 }
 
diff --git a/include/linux/qed/qed_eth_if.h b/include/linux/qed/qed_eth_if.h
index 9755a3feb52e..7a52f7c58c37 100644
--- a/include/linux/qed/qed_eth_if.h
+++ b/include/linux/qed/qed_eth_if.h
@@ -15,6 +15,29 @@
 #include <linux/qed/qed_if.h>
 #include <linux/qed/qed_iov_if.h>
 
+struct qed_queue_start_common_params {
+	/* Should always be relative to entity sending this. */
+	u8 vport_id;
+	u16 queue_id;
+
+	/* Relative, but relevant only for PFs */
+	u8 stats_id;
+
+	/* These are always absolute */
+	u16 sb;
+	u8 sb_idx;
+};
+
+struct qed_rxq_start_ret_params {
+	void __iomem *p_prod;
+	void *p_handle;
+};
+
+struct qed_txq_start_ret_params {
+	void __iomem *p_doorbell;
+	void *p_handle;
+};
+
 struct qed_dev_eth_info {
 	struct qed_dev_info common;
 
@@ -56,18 +79,6 @@ struct qed_start_vport_params {
 	bool clear_stats;
 };
 
-struct qed_stop_rxq_params {
-	u8 rss_id;
-	u8 rx_queue_id;
-	u8 vport_id;
-	bool eq_completion_only;
-};
-
-struct qed_stop_txq_params {
-	u8 rss_id;
-	u8 tx_queue_id;
-};
-
 enum qed_filter_rx_mode_type {
 	QED_FILTER_RX_MODE_TYPE_REGULAR,
 	QED_FILTER_RX_MODE_TYPE_MULTI_PROMISC,
@@ -112,15 +123,6 @@ struct qed_filter_params {
 	union qed_filter_type_params filter;
 };
 
-struct qed_queue_start_common_params {
-	u8 rss_id;
-	u8 queue_id;
-	u8 vport_id;
-	u16 sb;
-	u16 sb_idx;
-	u16 vf_qid;
-};
-
 struct qed_tunn_params {
 	u16 vxlan_port;
 	u8 update_vxlan_port;
@@ -220,24 +222,24 @@ struct qed_eth_ops {
 			    struct qed_update_vport_params *params);
 
 	int (*q_rx_start)(struct qed_dev *cdev,
+			  u8 rss_num,
 			  struct qed_queue_start_common_params *params,
 			  u16 bd_max_bytes,
 			  dma_addr_t bd_chain_phys_addr,
 			  dma_addr_t cqe_pbl_addr,
 			  u16 cqe_pbl_size,
-			  void __iomem **pp_prod);
+			  struct qed_rxq_start_ret_params *ret_params);
 
-	int (*q_rx_stop)(struct qed_dev *cdev,
-			 struct qed_stop_rxq_params *params);
+	int (*q_rx_stop)(struct qed_dev *cdev, u8 rss_id, void *handle);
 
 	int (*q_tx_start)(struct qed_dev *cdev,
+			  u8 rss_num,
 			  struct qed_queue_start_common_params *params,
 			  dma_addr_t pbl_addr,
 			  u16 pbl_size,
-			  void __iomem **pp_doorbell);
+			  struct qed_txq_start_ret_params *ret_params);
 
-	int (*q_tx_stop)(struct qed_dev *cdev,
-			 struct qed_stop_txq_params *params);
+	int (*q_tx_stop)(struct qed_dev *cdev, u8 rss_id, void *handle);
 
 	int (*filter_config)(struct qed_dev *cdev,
 			     struct qed_filter_params *params);