Merge branch 'thunderx-features-fixes'

Aleksey Makarov says:

====================
net: thunderx: New features and fixes

v2:
  - The unused affinity_mask field of the structure cmp_queue
  has been deleted. (thanks to David Miller)
  - The unneeded initializers have been dropped. (thanks to Alexey Klimov)
  - The commit message "net: thunderx: Rework interrupt handling"
  has been fixed. (thanks to Alexey Klimov)
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

MAINTAINERS

@@ -928,7 +928,7 @@ M:	Sunil Goutham <sgoutham@cavium.com>
 M:	Robert Richter <rric@kernel.org>
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
 S:	Supported
-F:	drivers/net/ethernet/cavium/
+F:	drivers/net/ethernet/cavium/thunder/
 
 ARM/CIRRUS LOGIC CLPS711X ARM ARCHITECTURE
 M:	Alexander Shiyan <shc_work@mail.ru>
@@ -2543,7 +2543,6 @@ M:     Raghu Vatsavayi <raghu.vatsavayi@caviumnetworks.com>
 L:     netdev@vger.kernel.org
 W:     http://www.cavium.com
 S:     Supported
-F:     drivers/net/ethernet/cavium/
 F:     drivers/net/ethernet/cavium/liquidio/
 
 CC2520 IEEE-802.15.4 RADIO DRIVER

drivers/net/ethernet/cavium/thunder/nic.h

@@ -135,6 +135,7 @@
 #define	NICVF_TX_TIMEOUT		(50 * HZ)
 
 struct nicvf_cq_poll {
+	struct  nicvf *nicvf;
 	u8	cq_idx;		/* Completion queue index */
 	struct	napi_struct napi;
 };
@@ -190,10 +191,10 @@ enum tx_stats_reg_offset {
 };
 
 struct nicvf_hw_stats {
-	u64 rx_bytes_ok;
-	u64 rx_ucast_frames_ok;
-	u64 rx_bcast_frames_ok;
-	u64 rx_mcast_frames_ok;
+	u64 rx_bytes;
+	u64 rx_ucast_frames;
+	u64 rx_bcast_frames;
+	u64 rx_mcast_frames;
 	u64 rx_fcs_errors;
 	u64 rx_l2_errors;
 	u64 rx_drop_red;
@@ -204,6 +205,31 @@ struct nicvf_hw_stats {
 	u64 rx_drop_mcast;
 	u64 rx_drop_l3_bcast;
 	u64 rx_drop_l3_mcast;
+	u64 rx_bgx_truncated_pkts;
+	u64 rx_jabber_errs;
+	u64 rx_fcs_errs;
+	u64 rx_bgx_errs;
+	u64 rx_prel2_errs;
+	u64 rx_l2_hdr_malformed;
+	u64 rx_oversize;
+	u64 rx_undersize;
+	u64 rx_l2_len_mismatch;
+	u64 rx_l2_pclp;
+	u64 rx_ip_ver_errs;
+	u64 rx_ip_csum_errs;
+	u64 rx_ip_hdr_malformed;
+	u64 rx_ip_payload_malformed;
+	u64 rx_ip_ttl_errs;
+	u64 rx_l3_pclp;
+	u64 rx_l4_malformed;
+	u64 rx_l4_csum_errs;
+	u64 rx_udp_len_errs;
+	u64 rx_l4_port_errs;
+	u64 rx_tcp_flag_errs;
+	u64 rx_tcp_offset_errs;
+	u64 rx_l4_pclp;
+	u64 rx_truncated_pkts;
+
 	u64 tx_bytes_ok;
 	u64 tx_ucast_frames_ok;
 	u64 tx_bcast_frames_ok;
@@ -222,6 +248,7 @@ struct nicvf_drv_stats {
 	u64 rx_frames_1518;
 	u64 rx_frames_jumbo;
 	u64 rx_drops;
+
 	/* Tx */
 	u64 tx_frames_ok;
 	u64 tx_drops;
@@ -231,13 +258,24 @@ struct nicvf_drv_stats {
 };
 
 struct nicvf {
+	struct nicvf		*pnicvf;
 	struct net_device	*netdev;
 	struct pci_dev		*pdev;
 	u8			vf_id;
 	u8			node;
-	u8			tns_mode;
+	u8			tns_mode:1;
+	u8			sqs_mode:1;
+	u8			loopback_supported:1;
 	u16			mtu;
 	struct queue_set	*qs;
+#define	MAX_SQS_PER_VF_SINGLE_NODE		5
+#define	MAX_SQS_PER_VF				11
+	u8			sqs_id;
+	u8			sqs_count; /* Secondary Qset count */
+	struct nicvf		*snicvf[MAX_SQS_PER_VF];
+	u8			rx_queues;
+	u8			tx_queues;
+	u8			max_queues;
 	void __iomem		*reg_base;
 	bool			link_up;
 	u8			duplex;
@@ -257,7 +295,7 @@ struct nicvf {
 	u32			cq_coalesce_usecs;
 
 	u32			msg_enable;
-	struct nicvf_hw_stats   stats;
+	struct nicvf_hw_stats   hw_stats;
 	struct nicvf_drv_stats  drv_stats;
 	struct bgx_stats	bgx_stats;
 	struct work_struct	reset_task;
@@ -269,10 +307,9 @@ struct nicvf {
 	char			irq_name[NIC_VF_MSIX_VECTORS][20];
 	bool			irq_allocated[NIC_VF_MSIX_VECTORS];
 
-	bool			pf_ready_to_rcv_msg;
+	/* VF <-> PF mailbox communication */
 	bool			pf_acked;
 	bool			pf_nacked;
-	bool			bgx_stats_acked;
 	bool			set_mac_pending;
 } ____cacheline_aligned_in_smp;
 
@@ -304,14 +341,21 @@ struct nicvf {
 #define	NIC_MBOX_MSG_RQ_SW_SYNC		0x0F	/* Flush inflight pkts to RQ */
 #define	NIC_MBOX_MSG_BGX_STATS		0x10	/* Get stats from BGX */
 #define	NIC_MBOX_MSG_BGX_LINK_CHANGE	0x11	/* BGX:LMAC link status */
-#define NIC_MBOX_MSG_CFG_DONE		0x12	/* VF configuration done */
-#define NIC_MBOX_MSG_SHUTDOWN		0x13	/* VF is being shutdown */
+#define	NIC_MBOX_MSG_ALLOC_SQS		0x12	/* Allocate secondary Qset */
+#define	NIC_MBOX_MSG_NICVF_PTR		0x13	/* Send nicvf ptr to PF */
+#define	NIC_MBOX_MSG_PNICVF_PTR		0x14	/* Get primary qset nicvf ptr */
+#define	NIC_MBOX_MSG_SNICVF_PTR		0x15	/* Send sqet nicvf ptr to PVF */
+#define	NIC_MBOX_MSG_LOOPBACK		0x16	/* Set interface in loopback */
+#define	NIC_MBOX_MSG_CFG_DONE		0xF0	/* VF configuration done */
+#define	NIC_MBOX_MSG_SHUTDOWN		0xF1	/* VF is being shutdown */
 
 struct nic_cfg_msg {
 	u8    msg;
 	u8    vf_id;
-	u8    tns_mode;
 	u8    node_id;
+	u8    tns_mode:1;
+	u8    sqs_mode:1;
+	u8    loopback_supported:1;
 	u8    mac_addr[ETH_ALEN];
 };
 
@@ -319,6 +363,7 @@ struct nic_cfg_msg {
 struct qs_cfg_msg {
 	u8    msg;
 	u8    num;
+	u8    sqs_count;
 	u64   cfg;
 };
 
@@ -335,6 +380,7 @@ struct sq_cfg_msg {
 	u8    msg;
 	u8    qs_num;
 	u8    sq_num;
+	bool  sqs_mode;
 	u64   cfg;
 };
 
@@ -394,6 +440,28 @@ struct bgx_link_status {
 	u32   speed;
 };
 
+/* Get Extra Qset IDs */
+struct sqs_alloc {
+	u8    msg;
+	u8    vf_id;
+	u8    qs_count;
+};
+
+struct nicvf_ptr {
+	u8    msg;
+	u8    vf_id;
+	bool  sqs_mode;
+	u8    sqs_id;
+	u64   nicvf;
+};
+
+/* Set interface in loopback mode */
+struct set_loopback {
+	u8    msg;
+	u8    vf_id;
+	bool  enable;
+};
+
 /* 128 bit shared memory between PF and each VF */
 union nic_mbx {
 	struct { u8 msg; }	msg;
@@ -408,6 +476,9 @@ union nic_mbx {
 	struct rss_cfg_msg	rss_cfg;
 	struct bgx_stats_msg    bgx_stats;
 	struct bgx_link_status  link_status;
+	struct sqs_alloc        sqs_alloc;
+	struct nicvf_ptr	nicvf;
+	struct set_loopback	lbk;
 };
 
 #define NIC_NODE_ID_MASK	0x03

drivers/net/ethernet/cavium/thunder/nic_main.c

@@ -28,6 +28,11 @@ struct nicpf {
 	u8			num_vf_en;      /* No of VF enabled */
 	bool			vf_enabled[MAX_NUM_VFS_SUPPORTED];
 	void __iomem		*reg_base;       /* Register start address */
+	u8			num_sqs_en;	/* Secondary qsets enabled */
+	u64			nicvf[MAX_NUM_VFS_SUPPORTED];
+	u8			vf_sqs[MAX_NUM_VFS_SUPPORTED][MAX_SQS_PER_VF];
+	u8			pqs_vf[MAX_NUM_VFS_SUPPORTED];
+	bool			sqs_used[MAX_NUM_VFS_SUPPORTED];
 	struct pkind_cfg	pkind;
 #define	NIC_SET_VF_LMAC_MAP(bgx, lmac)	(((bgx & 0xF) << 4) | (lmac & 0xF))
 #define	NIC_GET_BGX_FROM_VF_LMAC_MAP(map)	((map >> 4) & 0xF)
@@ -139,14 +144,19 @@ static void nic_mbx_send_ready(struct nicpf *nic, int vf)
 
 	mbx.nic_cfg.tns_mode = NIC_TNS_BYPASS_MODE;
 
-	bgx_idx = NIC_GET_BGX_FROM_VF_LMAC_MAP(nic->vf_lmac_map[vf]);
-	lmac = NIC_GET_LMAC_FROM_VF_LMAC_MAP(nic->vf_lmac_map[vf]);
-
-	mac = bgx_get_lmac_mac(nic->node, bgx_idx, lmac);
-	if (mac)
-		ether_addr_copy((u8 *)&mbx.nic_cfg.mac_addr, mac);
+	if (vf < MAX_LMAC) {
+		bgx_idx = NIC_GET_BGX_FROM_VF_LMAC_MAP(nic->vf_lmac_map[vf]);
+		lmac = NIC_GET_LMAC_FROM_VF_LMAC_MAP(nic->vf_lmac_map[vf]);
 
+		mac = bgx_get_lmac_mac(nic->node, bgx_idx, lmac);
+		if (mac)
+			ether_addr_copy((u8 *)&mbx.nic_cfg.mac_addr, mac);
+	}
+	mbx.nic_cfg.sqs_mode = (vf >= nic->num_vf_en) ? true : false;
 	mbx.nic_cfg.node_id = nic->node;
+
+	mbx.nic_cfg.loopback_supported = vf < MAX_LMAC;
+
 	nic_send_msg_to_vf(nic, vf, &mbx);
 }
 
@@ -329,6 +339,10 @@ static void nic_init_hw(struct nicpf *nic)
 
 	/* Timer config */
 	nic_reg_write(nic, NIC_PF_INTR_TIMER_CFG, NICPF_CLK_PER_INT_TICK);
+
+	/* Enable VLAN ethertype matching and stripping */
+	nic_reg_write(nic, NIC_PF_RX_ETYPE_0_7,
+		      (2 << 19) | (ETYPE_ALG_VLAN_STRIP << 16) | ETH_P_8021Q);
 }
 
 /* Channel parse index configuration */
@@ -429,6 +443,12 @@ static void nic_config_rss(struct nicpf *nic, struct rss_cfg_msg *cfg)
 	qset = cfg->vf_id;
 
 	for (; rssi < (rssi_base + cfg->tbl_len); rssi++) {
+		u8 svf = cfg->ind_tbl[idx] >> 3;
+
+		if (svf)
+			qset = nic->vf_sqs[cfg->vf_id][svf - 1];
+		else
+			qset = cfg->vf_id;
 		nic_reg_write(nic, NIC_PF_RSSI_0_4097_RQ | (rssi << 3),
 			      (qset << 3) | (cfg->ind_tbl[idx] & 0x7));
 		idx++;
@@ -452,19 +472,31 @@ static void nic_config_rss(struct nicpf *nic, struct rss_cfg_msg *cfg)
  * VNIC6-SQ0 -> TL4(528) -> TL3[132] -> TL2[33] -> TL1[1] -> BGX1
  * VNIC7-SQ0 -> TL4(536) -> TL3[134] -> TL2[33] -> TL1[1] -> BGX1
  */
-static void nic_tx_channel_cfg(struct nicpf *nic, u8 vnic, u8 sq_idx)
+static void nic_tx_channel_cfg(struct nicpf *nic, u8 vnic,
+			       struct sq_cfg_msg *sq)
 {
 	u32 bgx, lmac, chan;
 	u32 tl2, tl3, tl4;
 	u32 rr_quantum;
+	u8 sq_idx = sq->sq_num;
+	u8 pqs_vnic;
+
+	if (sq->sqs_mode)
+		pqs_vnic = nic->pqs_vf[vnic];
+	else
+		pqs_vnic = vnic;
+
+	bgx = NIC_GET_BGX_FROM_VF_LMAC_MAP(nic->vf_lmac_map[pqs_vnic]);
+	lmac = NIC_GET_LMAC_FROM_VF_LMAC_MAP(nic->vf_lmac_map[pqs_vnic]);
 
-	bgx = NIC_GET_BGX_FROM_VF_LMAC_MAP(nic->vf_lmac_map[vnic]);
-	lmac = NIC_GET_LMAC_FROM_VF_LMAC_MAP(nic->vf_lmac_map[vnic]);
 	/* 24 bytes for FCS, IPG and preamble */
 	rr_quantum = ((NIC_HW_MAX_FRS + 24) / 4);
 
 	tl4 = (lmac * NIC_TL4_PER_LMAC) + (bgx * NIC_TL4_PER_BGX);
 	tl4 += sq_idx;
+	if (sq->sqs_mode)
+		tl4 += vnic * 8;
+
 	tl3 = tl4 / (NIC_MAX_TL4 / NIC_MAX_TL3);
 	nic_reg_write(nic, NIC_PF_QSET_0_127_SQ_0_7_CFG2 |
 		      ((u64)vnic << NIC_QS_ID_SHIFT) |
@@ -485,6 +517,86 @@ static void nic_tx_channel_cfg(struct nicpf *nic, u8 vnic, u8 sq_idx)
 	nic_reg_write(nic, NIC_PF_TL2_0_63_PRI | (tl2 << 3), 0x00);
 }
 
+/* Send primary nicvf pointer to secondary QS's VF */
+static void nic_send_pnicvf(struct nicpf *nic, int sqs)
+{
+	union nic_mbx mbx = {};
+
+	mbx.nicvf.msg = NIC_MBOX_MSG_PNICVF_PTR;
+	mbx.nicvf.nicvf = nic->nicvf[nic->pqs_vf[sqs]];
+	nic_send_msg_to_vf(nic, sqs, &mbx);
+}
+
+/* Send SQS's nicvf pointer to primary QS's VF */
+static void nic_send_snicvf(struct nicpf *nic, struct nicvf_ptr *nicvf)
+{
+	union nic_mbx mbx = {};
+	int sqs_id = nic->vf_sqs[nicvf->vf_id][nicvf->sqs_id];
+
+	mbx.nicvf.msg = NIC_MBOX_MSG_SNICVF_PTR;
+	mbx.nicvf.sqs_id = nicvf->sqs_id;
+	mbx.nicvf.nicvf = nic->nicvf[sqs_id];
+	nic_send_msg_to_vf(nic, nicvf->vf_id, &mbx);
+}
+
+/* Find next available Qset that can be assigned as a
+ * secondary Qset to a VF.
+ */
+static int nic_nxt_avail_sqs(struct nicpf *nic)
+{
+	int sqs;
+
+	for (sqs = 0; sqs < nic->num_sqs_en; sqs++) {
+		if (!nic->sqs_used[sqs])
+			nic->sqs_used[sqs] = true;
+		else
+			continue;
+		return sqs + nic->num_vf_en;
+	}
+	return -1;
+}
+
+/* Allocate additional Qsets for requested VF */
+static void nic_alloc_sqs(struct nicpf *nic, struct sqs_alloc *sqs)
+{
+	union nic_mbx mbx = {};
+	int idx, alloc_qs = 0;
+	int sqs_id;
+
+	if (!nic->num_sqs_en)
+		goto send_mbox;
+
+	for (idx = 0; idx < sqs->qs_count; idx++) {
+		sqs_id = nic_nxt_avail_sqs(nic);
+		if (sqs_id < 0)
+			break;
+		nic->vf_sqs[sqs->vf_id][idx] = sqs_id;
+		nic->pqs_vf[sqs_id] = sqs->vf_id;
+		alloc_qs++;
+	}
+
+send_mbox:
+	mbx.sqs_alloc.msg = NIC_MBOX_MSG_ALLOC_SQS;
+	mbx.sqs_alloc.vf_id = sqs->vf_id;
+	mbx.sqs_alloc.qs_count = alloc_qs;
+	nic_send_msg_to_vf(nic, sqs->vf_id, &mbx);
+}
+
+static int nic_config_loopback(struct nicpf *nic, struct set_loopback *lbk)
+{
+	int bgx_idx, lmac_idx;
+
+	if (lbk->vf_id > MAX_LMAC)
+		return -1;
+
+	bgx_idx = NIC_GET_BGX_FROM_VF_LMAC_MAP(nic->vf_lmac_map[lbk->vf_id]);
+	lmac_idx = NIC_GET_LMAC_FROM_VF_LMAC_MAP(nic->vf_lmac_map[lbk->vf_id]);
+
+	bgx_lmac_internal_loopback(nic->node, bgx_idx, lmac_idx, lbk->enable);
+
+	return 0;
+}
+
 /* Interrupt handler to handle mailbox messages from VFs */
 static void nic_handle_mbx_intr(struct nicpf *nic, int vf)
 {
@@ -492,6 +604,7 @@ static void nic_handle_mbx_intr(struct nicpf *nic, int vf)
 	u64 *mbx_data;
 	u64 mbx_addr;
 	u64 reg_addr;
+	u64 cfg;
 	int bgx, lmac;
 	int i;
 	int ret = 0;
@@ -512,15 +625,24 @@ static void nic_handle_mbx_intr(struct nicpf *nic, int vf)
 	switch (mbx.msg.msg) {
 	case NIC_MBOX_MSG_READY:
 		nic_mbx_send_ready(nic, vf);
-		nic->link[vf] = 0;
-		nic->duplex[vf] = 0;
-		nic->speed[vf] = 0;
+		if (vf < MAX_LMAC) {
+			nic->link[vf] = 0;
+			nic->duplex[vf] = 0;
+			nic->speed[vf] = 0;
+		}
 		ret = 1;
 		break;
 	case NIC_MBOX_MSG_QS_CFG:
 		reg_addr = NIC_PF_QSET_0_127_CFG |
 			   (mbx.qs.num << NIC_QS_ID_SHIFT);
-		nic_reg_write(nic, reg_addr, mbx.qs.cfg);
+		cfg = mbx.qs.cfg;
+		/* Check if its a secondary Qset */
+		if (vf >= nic->num_vf_en) {
+			cfg = cfg & (~0x7FULL);
+			/* Assign this Qset to primary Qset's VF */
+			cfg |= nic->pqs_vf[vf];
+		}
+		nic_reg_write(nic, reg_addr, cfg);
 		break;
 	case NIC_MBOX_MSG_RQ_CFG:
 		reg_addr = NIC_PF_QSET_0_127_RQ_0_7_CFG |
@@ -548,9 +670,11 @@ static void nic_handle_mbx_intr(struct nicpf *nic, int vf)
 			   (mbx.sq.qs_num << NIC_QS_ID_SHIFT) |
 			   (mbx.sq.sq_num << NIC_Q_NUM_SHIFT);
 		nic_reg_write(nic, reg_addr, mbx.sq.cfg);
-		nic_tx_channel_cfg(nic, mbx.qs.num, mbx.sq.sq_num);
+		nic_tx_channel_cfg(nic, mbx.qs.num, &mbx.sq);
 		break;
 	case NIC_MBOX_MSG_SET_MAC:
+		if (vf >= nic->num_vf_en)
+			break;
 		lmac = mbx.mac.vf_id;
 		bgx = NIC_GET_BGX_FROM_VF_LMAC_MAP(nic->vf_lmac_map[lmac]);
 		lmac = NIC_GET_LMAC_FROM_VF_LMAC_MAP(nic->vf_lmac_map[lmac]);
@@ -577,10 +701,28 @@ static void nic_handle_mbx_intr(struct nicpf *nic, int vf)
 	case NIC_MBOX_MSG_SHUTDOWN:
 		/* First msg in VF teardown sequence */
 		nic->vf_enabled[vf] = false;
+		if (vf >= nic->num_vf_en)
+			nic->sqs_used[vf - nic->num_vf_en] = false;
+		nic->pqs_vf[vf] = 0;
 		break;
+	case NIC_MBOX_MSG_ALLOC_SQS:
+		nic_alloc_sqs(nic, &mbx.sqs_alloc);
+		goto unlock;
+	case NIC_MBOX_MSG_NICVF_PTR:
+		nic->nicvf[vf] = mbx.nicvf.nicvf;
+		break;
+	case NIC_MBOX_MSG_PNICVF_PTR:
+		nic_send_pnicvf(nic, vf);
+		goto unlock;
+	case NIC_MBOX_MSG_SNICVF_PTR:
+		nic_send_snicvf(nic, &mbx.nicvf);
+		goto unlock;
 	case NIC_MBOX_MSG_BGX_STATS:
 		nic_get_bgx_stats(nic, &mbx.bgx_stats);
 		goto unlock;
+	case NIC_MBOX_MSG_LOOPBACK:
+		ret = nic_config_loopback(nic, &mbx.lbk);
+		break;
 	default:
 		dev_err(&nic->pdev->dev,
 			"Invalid msg from VF%d, msg 0x%x\n", vf, mbx.msg.msg);
@@ -606,8 +748,7 @@ static void nic_mbx_intr_handler (struct nicpf *nic, int mbx)
 		if (intr & (1ULL << vf)) {
 			dev_dbg(&nic->pdev->dev, "Intr from VF %d\n",
 				vf + (mbx * vf_per_mbx_reg));
-			if ((vf + (mbx * vf_per_mbx_reg)) > nic->num_vf_en)
-				break;
+
 			nic_handle_mbx_intr(nic, vf + (mbx * vf_per_mbx_reg));
 			nic_clear_mbx_intr(nic, vf, mbx);
 		}
@@ -713,9 +854,24 @@ static void nic_unregister_interrupts(struct nicpf *nic)
 	nic_disable_msix(nic);
 }
 
+static int nic_num_sqs_en(struct nicpf *nic, int vf_en)
+{
+	int pos, sqs_per_vf = MAX_SQS_PER_VF_SINGLE_NODE;
+	u16 total_vf;
+
+	/* Check if its a multi-node environment */
+	if (nr_node_ids > 1)
+		sqs_per_vf = MAX_SQS_PER_VF;
+
+	pos = pci_find_ext_capability(nic->pdev, PCI_EXT_CAP_ID_SRIOV);
+	pci_read_config_word(nic->pdev, (pos + PCI_SRIOV_TOTAL_VF), &total_vf);
+	return min(total_vf - vf_en, vf_en * sqs_per_vf);
+}
+
 static int nic_sriov_init(struct pci_dev *pdev, struct nicpf *nic)
 {
 	int pos = 0;
+	int vf_en;
 	int err;
 	u16 total_vf_cnt;
 
@@ -732,16 +888,20 @@ static int nic_sriov_init(struct pci_dev *pdev, struct nicpf *nic)
 	if (!total_vf_cnt)
 		return 0;
 
-	err = pci_enable_sriov(pdev, nic->num_vf_en);
+	vf_en = nic->num_vf_en;
+	nic->num_sqs_en = nic_num_sqs_en(nic, nic->num_vf_en);
+	vf_en += nic->num_sqs_en;
+
+	err = pci_enable_sriov(pdev, vf_en);
 	if (err) {
 		dev_err(&pdev->dev, "SRIOV enable failed, num VF is %d\n",
-			nic->num_vf_en);
+			vf_en);
 		nic->num_vf_en = 0;
 		return err;
 	}
 
 	dev_info(&pdev->dev, "SRIOV enabled, number of VF available %d\n",
-		 nic->num_vf_en);
+		 vf_en);
 
 	nic->flags |= NIC_SRIOV_ENABLED;
 	return 0;

drivers/net/ethernet/cavium/thunder/nicvf_ethtool.c

@@ -35,10 +35,10 @@ struct nicvf_stat {
 }
 
 static const struct nicvf_stat nicvf_hw_stats[] = {
-	NICVF_HW_STAT(rx_bytes_ok),
-	NICVF_HW_STAT(rx_ucast_frames_ok),
-	NICVF_HW_STAT(rx_bcast_frames_ok),
-	NICVF_HW_STAT(rx_mcast_frames_ok),
+	NICVF_HW_STAT(rx_bytes),
+	NICVF_HW_STAT(rx_ucast_frames),
+	NICVF_HW_STAT(rx_bcast_frames),
+	NICVF_HW_STAT(rx_mcast_frames),
 	NICVF_HW_STAT(rx_fcs_errors),
 	NICVF_HW_STAT(rx_l2_errors),
 	NICVF_HW_STAT(rx_drop_red),
@@ -49,6 +49,30 @@ static const struct nicvf_stat nicvf_hw_stats[] = {
 	NICVF_HW_STAT(rx_drop_mcast),
 	NICVF_HW_STAT(rx_drop_l3_bcast),
 	NICVF_HW_STAT(rx_drop_l3_mcast),
+	NICVF_HW_STAT(rx_bgx_truncated_pkts),
+	NICVF_HW_STAT(rx_jabber_errs),
+	NICVF_HW_STAT(rx_fcs_errs),
+	NICVF_HW_STAT(rx_bgx_errs),
+	NICVF_HW_STAT(rx_prel2_errs),
+	NICVF_HW_STAT(rx_l2_hdr_malformed),
+	NICVF_HW_STAT(rx_oversize),
+	NICVF_HW_STAT(rx_undersize),
+	NICVF_HW_STAT(rx_l2_len_mismatch),
+	NICVF_HW_STAT(rx_l2_pclp),
+	NICVF_HW_STAT(rx_ip_ver_errs),
+	NICVF_HW_STAT(rx_ip_csum_errs),
+	NICVF_HW_STAT(rx_ip_hdr_malformed),
+	NICVF_HW_STAT(rx_ip_payload_malformed),
+	NICVF_HW_STAT(rx_ip_ttl_errs),
+	NICVF_HW_STAT(rx_l3_pclp),
+	NICVF_HW_STAT(rx_l4_malformed),
+	NICVF_HW_STAT(rx_l4_csum_errs),
+	NICVF_HW_STAT(rx_udp_len_errs),
+	NICVF_HW_STAT(rx_l4_port_errs),
+	NICVF_HW_STAT(rx_tcp_flag_errs),
+	NICVF_HW_STAT(rx_tcp_offset_errs),
+	NICVF_HW_STAT(rx_l4_pclp),
+	NICVF_HW_STAT(rx_truncated_pkts),
 	NICVF_HW_STAT(tx_bytes_ok),
 	NICVF_HW_STAT(tx_ucast_frames_ok),
 	NICVF_HW_STAT(tx_bcast_frames_ok),
@@ -125,10 +149,33 @@ static void nicvf_set_msglevel(struct net_device *netdev, u32 lvl)
 	nic->msg_enable = lvl;
 }
 
+static void nicvf_get_qset_strings(struct nicvf *nic, u8 **data, int qset)
+{
+	int stats, qidx;
+	int start_qidx = qset * MAX_RCV_QUEUES_PER_QS;
+
+	for (qidx = 0; qidx < nic->qs->rq_cnt; qidx++) {
+		for (stats = 0; stats < nicvf_n_queue_stats; stats++) {
+			sprintf(*data, "rxq%d: %s", qidx + start_qidx,
+				nicvf_queue_stats[stats].name);
+			*data += ETH_GSTRING_LEN;
+		}
+	}
+
+	for (qidx = 0; qidx < nic->qs->sq_cnt; qidx++) {
+		for (stats = 0; stats < nicvf_n_queue_stats; stats++) {
+			sprintf(*data, "txq%d: %s", qidx + start_qidx,
+				nicvf_queue_stats[stats].name);
+			*data += ETH_GSTRING_LEN;
+		}
+	}
+}
+
 static void nicvf_get_strings(struct net_device *netdev, u32 sset, u8 *data)
 {
 	struct nicvf *nic = netdev_priv(netdev);
-	int stats, qidx;
+	int stats;
+	int sqs;
 
 	if (sset != ETH_SS_STATS)
 		return;
@@ -143,20 +190,12 @@ static void nicvf_get_strings(struct net_device *netdev, u32 sset, u8 *data)
 		data += ETH_GSTRING_LEN;
 	}
 
-	for (qidx = 0; qidx < nic->qs->rq_cnt; qidx++) {
-		for (stats = 0; stats < nicvf_n_queue_stats; stats++) {
-			sprintf(data, "rxq%d: %s", qidx,
-				nicvf_queue_stats[stats].name);
-			data += ETH_GSTRING_LEN;
-		}
-	}
+	nicvf_get_qset_strings(nic, &data, 0);
 
-	for (qidx = 0; qidx < nic->qs->sq_cnt; qidx++) {
-		for (stats = 0; stats < nicvf_n_queue_stats; stats++) {
-			sprintf(data, "txq%d: %s", qidx,
-				nicvf_queue_stats[stats].name);
-			data += ETH_GSTRING_LEN;
-		}
+	for (sqs = 0; sqs < nic->sqs_count; sqs++) {
+		if (!nic->snicvf[sqs])
+			continue;
+		nicvf_get_qset_strings(nic->snicvf[sqs], &data, sqs + 1);
 	}
 
 	for (stats = 0; stats < BGX_RX_STATS_COUNT; stats++) {
@@ -173,21 +212,58 @@ static void nicvf_get_strings(struct net_device *netdev, u32 sset, u8 *data)
 static int nicvf_get_sset_count(struct net_device *netdev, int sset)
 {
 	struct nicvf *nic = netdev_priv(netdev);
+	int qstats_count;
+	int sqs;
 
 	if (sset != ETH_SS_STATS)
 		return -EINVAL;
 
+	qstats_count = nicvf_n_queue_stats *
+		       (nic->qs->rq_cnt + nic->qs->sq_cnt);
+	for (sqs = 0; sqs < nic->sqs_count; sqs++) {
+		struct nicvf *snic;
+
+		snic = nic->snicvf[sqs];
+		if (!snic)
+			continue;
+		qstats_count += nicvf_n_queue_stats *
+				(snic->qs->rq_cnt + snic->qs->sq_cnt);
+	}
+
 	return nicvf_n_hw_stats + nicvf_n_drv_stats +
-		(nicvf_n_queue_stats *
-		 (nic->qs->rq_cnt + nic->qs->sq_cnt)) +
+		qstats_count +
 		BGX_RX_STATS_COUNT + BGX_TX_STATS_COUNT;
 }
 
+static void nicvf_get_qset_stats(struct nicvf *nic,
+				 struct ethtool_stats *stats, u64 **data)
+{
+	int stat, qidx;
+
+	if (!nic)
+		return;
+
+	for (qidx = 0; qidx < nic->qs->rq_cnt; qidx++) {
+		nicvf_update_rq_stats(nic, qidx);
+		for (stat = 0; stat < nicvf_n_queue_stats; stat++)
+			*((*data)++) = ((u64 *)&nic->qs->rq[qidx].stats)
+					[nicvf_queue_stats[stat].index];
+	}
+
+	for (qidx = 0; qidx < nic->qs->sq_cnt; qidx++) {
+		nicvf_update_sq_stats(nic, qidx);
+		for (stat = 0; stat < nicvf_n_queue_stats; stat++)
+			*((*data)++) = ((u64 *)&nic->qs->sq[qidx].stats)
+					[nicvf_queue_stats[stat].index];
+	}
+}
+
 static void nicvf_get_ethtool_stats(struct net_device *netdev,
 				    struct ethtool_stats *stats, u64 *data)
 {
 	struct nicvf *nic = netdev_priv(netdev);
-	int stat, qidx;
+	int stat;
+	int sqs;
 
 	nicvf_update_stats(nic);
 
@@ -195,22 +271,18 @@ static void nicvf_get_ethtool_stats(struct net_device *netdev,
 	nicvf_update_lmac_stats(nic);
 
 	for (stat = 0; stat < nicvf_n_hw_stats; stat++)
-		*(data++) = ((u64 *)&nic->stats)
+		*(data++) = ((u64 *)&nic->hw_stats)
 				[nicvf_hw_stats[stat].index];
 	for (stat = 0; stat < nicvf_n_drv_stats; stat++)
 		*(data++) = ((u64 *)&nic->drv_stats)
 				[nicvf_drv_stats[stat].index];
 
-	for (qidx = 0; qidx < nic->qs->rq_cnt; qidx++) {
-		for (stat = 0; stat < nicvf_n_queue_stats; stat++)
-			*(data++) = ((u64 *)&nic->qs->rq[qidx].stats)
-					[nicvf_queue_stats[stat].index];
-	}
+	nicvf_get_qset_stats(nic, stats, &data);
 
-	for (qidx = 0; qidx < nic->qs->sq_cnt; qidx++) {
-		for (stat = 0; stat < nicvf_n_queue_stats; stat++)
-			*(data++) = ((u64 *)&nic->qs->sq[qidx].stats)
-					[nicvf_queue_stats[stat].index];
+	for (sqs = 0; sqs < nic->sqs_count; sqs++) {
+		if (!nic->snicvf[sqs])
+			continue;
+		nicvf_get_qset_stats(nic->snicvf[sqs], stats, &data);
 	}
 
 	for (stat = 0; stat < BGX_RX_STATS_COUNT; stat++)
@@ -369,7 +441,7 @@ static int nicvf_get_rxnfc(struct net_device *dev,
 
 	switch (info->cmd) {
 	case ETHTOOL_GRXRINGS:
-		info->data = nic->qs->rq_cnt;
+		info->data = nic->rx_queues;
 		ret = 0;
 		break;
 	case ETHTOOL_GRXFH:
@@ -501,17 +573,15 @@ static int nicvf_set_rxfh(struct net_device *dev, const u32 *indir,
 	struct nicvf_rss_info *rss = &nic->rss_info;
 	int idx;
 
-	if ((nic->qs->rq_cnt <= 1) || (nic->cpi_alg != CPI_ALG_NONE)) {
-		rss->enable = false;
-		rss->hash_bits = 0;
-		return -EIO;
-	}
-
-	/* We do not allow change in unsupported parameters */
 	if (hfunc != ETH_RSS_HASH_NO_CHANGE && hfunc != ETH_RSS_HASH_TOP)
 		return -EOPNOTSUPP;
 
-	rss->enable = true;
+	if (!rss->enable) {
+		netdev_err(nic->netdev,
+			   "RSS is disabled, cannot change settings\n");
+		return -EIO;
+	}
+
 	if (indir) {
 		for (idx = 0; idx < rss->rss_size; idx++)
 			rss->ind_tbl[idx] = indir[idx];
@@ -534,11 +604,11 @@ static void nicvf_get_channels(struct net_device *dev,
 
 	memset(channel, 0, sizeof(*channel));
 
-	channel->max_rx = MAX_RCV_QUEUES_PER_QS;
-	channel->max_tx = MAX_SND_QUEUES_PER_QS;
+	channel->max_rx = nic->max_queues;
+	channel->max_tx = nic->max_queues;
 
-	channel->rx_count = nic->qs->rq_cnt;
-	channel->tx_count = nic->qs->sq_cnt;
+	channel->rx_count = nic->rx_queues;
+	channel->tx_count = nic->tx_queues;
 }
 
 /* Set no of Tx, Rx queues to be used */
@@ -548,22 +618,34 @@ static int nicvf_set_channels(struct net_device *dev,
 	struct nicvf *nic = netdev_priv(dev);
 	int err = 0;
 	bool if_up = netif_running(dev);
+	int cqcount;
 
 	if (!channel->rx_count || !channel->tx_count)
 		return -EINVAL;
-	if (channel->rx_count > MAX_RCV_QUEUES_PER_QS)
+	if (channel->rx_count > nic->max_queues)
 		return -EINVAL;
-	if (channel->tx_count > MAX_SND_QUEUES_PER_QS)
+	if (channel->tx_count > nic->max_queues)
 		return -EINVAL;
 
 	if (if_up)
 		nicvf_stop(dev);
 
-	nic->qs->rq_cnt = channel->rx_count;
-	nic->qs->sq_cnt = channel->tx_count;
+	cqcount = max(channel->rx_count, channel->tx_count);
+
+	if (cqcount > MAX_CMP_QUEUES_PER_QS) {
+		nic->sqs_count = roundup(cqcount, MAX_CMP_QUEUES_PER_QS);
+		nic->sqs_count = (nic->sqs_count / MAX_CMP_QUEUES_PER_QS) - 1;
+	} else {
+		nic->sqs_count = 0;
+	}
+
+	nic->qs->rq_cnt = min_t(u32, channel->rx_count, MAX_RCV_QUEUES_PER_QS);
+	nic->qs->sq_cnt = min_t(u32, channel->tx_count, MAX_SND_QUEUES_PER_QS);
 	nic->qs->cq_cnt = max(nic->qs->rq_cnt, nic->qs->sq_cnt);
 
-	err = nicvf_set_real_num_queues(dev, nic->qs->sq_cnt, nic->qs->rq_cnt);
+	nic->rx_queues = channel->rx_count;
+	nic->tx_queues = channel->tx_count;
+	err = nicvf_set_real_num_queues(dev, nic->tx_queues, nic->rx_queues);
 	if (err)
 		return err;
 
@@ -571,7 +653,7 @@ static int nicvf_set_channels(struct net_device *dev,
 		nicvf_open(dev);
 
 	netdev_info(dev, "Setting num Tx rings to %d, Rx rings to %d success\n",
-		    nic->qs->sq_cnt, nic->qs->rq_cnt);
+		    nic->tx_queues, nic->rx_queues);
 
 	return err;
 }

drivers/net/ethernet/cavium/thunder/nicvf_main.c

@@ -10,6 +10,7 @@
 #include <linux/interrupt.h>
 #include <linux/pci.h>
 #include <linux/netdevice.h>
+#include <linux/if_vlan.h>
 #include <linux/etherdevice.h>
 #include <linux/ethtool.h>
 #include <linux/log2.h>
@@ -50,6 +51,14 @@ module_param(cpi_alg, int, S_IRUGO);
 MODULE_PARM_DESC(cpi_alg,
 		 "PFC algorithm (0=none, 1=VLAN, 2=VLAN16, 3=IP Diffserv)");
 
+static inline u8 nicvf_netdev_qidx(struct nicvf *nic, u8 qidx)
+{
+	if (nic->sqs_mode)
+		return qidx + ((nic->sqs_id + 1) * MAX_CMP_QUEUES_PER_QS);
+	else
+		return qidx;
+}
+
 static inline void nicvf_set_rx_frame_cnt(struct nicvf *nic,
 					  struct sk_buff *skb)
 {
@@ -105,7 +114,6 @@ u64 nicvf_queue_reg_read(struct nicvf *nic, u64 offset, u64 qidx)
 }
 
 /* VF -> PF mailbox communication */
-
 static void nicvf_write_to_mbx(struct nicvf *nic, union nic_mbx *mbx)
 {
 	u64 *msg = (u64 *)mbx;
@@ -147,26 +155,15 @@ int nicvf_send_msg_to_pf(struct nicvf *nic, union nic_mbx *mbx)
 */
 static int nicvf_check_pf_ready(struct nicvf *nic)
 {
-	int timeout = 5000, sleep = 20;
 	union nic_mbx mbx = {};
 
 	mbx.msg.msg = NIC_MBOX_MSG_READY;
-
-	nic->pf_ready_to_rcv_msg = false;
-
-	nicvf_write_to_mbx(nic, &mbx);
-
-	while (!nic->pf_ready_to_rcv_msg) {
-		msleep(sleep);
-		if (nic->pf_ready_to_rcv_msg)
-			break;
-		timeout -= sleep;
-		if (!timeout) {
-			netdev_err(nic->netdev,
-				   "PF didn't respond to READY msg\n");
-			return 0;
-		}
+	if (nicvf_send_msg_to_pf(nic, &mbx)) {
+		netdev_err(nic->netdev,
+			   "PF didn't respond to READY msg\n");
+		return 0;
 	}
+
 	return 1;
 }
 
@@ -197,13 +194,15 @@ static void  nicvf_handle_mbx_intr(struct nicvf *nic)
 	netdev_dbg(nic->netdev, "Mbox message: msg: 0x%x\n", mbx.msg.msg);
 	switch (mbx.msg.msg) {
 	case NIC_MBOX_MSG_READY:
-		nic->pf_ready_to_rcv_msg = true;
+		nic->pf_acked = true;
 		nic->vf_id = mbx.nic_cfg.vf_id & 0x7F;
 		nic->tns_mode = mbx.nic_cfg.tns_mode & 0x7F;
 		nic->node = mbx.nic_cfg.node_id;
 		if (!nic->set_mac_pending)
 			ether_addr_copy(nic->netdev->dev_addr,
 					mbx.nic_cfg.mac_addr);
+		nic->sqs_mode = mbx.nic_cfg.sqs_mode;
+		nic->loopback_supported = mbx.nic_cfg.loopback_supported;
 		nic->link_up = false;
 		nic->duplex = 0;
 		nic->speed = 0;
@@ -221,7 +220,6 @@ static void  nicvf_handle_mbx_intr(struct nicvf *nic)
 	case NIC_MBOX_MSG_BGX_STATS:
 		nicvf_read_bgx_stats(nic, &mbx.bgx_stats);
 		nic->pf_acked = true;
-		nic->bgx_stats_acked = true;
 		break;
 	case NIC_MBOX_MSG_BGX_LINK_CHANGE:
 		nic->pf_acked = true;
@@ -242,6 +240,26 @@ static void  nicvf_handle_mbx_intr(struct nicvf *nic)
 			netif_tx_stop_all_queues(nic->netdev);
 		}
 		break;
+	case NIC_MBOX_MSG_ALLOC_SQS:
+		nic->sqs_count = mbx.sqs_alloc.qs_count;
+		nic->pf_acked = true;
+		break;
+	case NIC_MBOX_MSG_SNICVF_PTR:
+		/* Primary VF: make note of secondary VF's pointer
+		 * to be used while packet transmission.
+		 */
+		nic->snicvf[mbx.nicvf.sqs_id] =
+			(struct nicvf *)mbx.nicvf.nicvf;
+		nic->pf_acked = true;
+		break;
+	case NIC_MBOX_MSG_PNICVF_PTR:
+		/* Secondary VF/Qset: make note of primary VF's pointer
+		 * to be used while packet reception, to handover packet
+		 * to primary VF's netdev.
+		 */
+		nic->pnicvf = (struct nicvf *)mbx.nicvf.nicvf;
+		nic->pf_acked = true;
+		break;
 	default:
 		netdev_err(nic->netdev,
 			   "Invalid message from PF, msg 0x%x\n", mbx.msg.msg);
@@ -326,7 +344,7 @@ static int nicvf_rss_init(struct nicvf *nic)
 
 	nicvf_get_rss_size(nic);
 
-	if ((nic->qs->rq_cnt <= 1) || (cpi_alg != CPI_ALG_NONE)) {
+	if (cpi_alg != CPI_ALG_NONE) {
 		rss->enable = false;
 		rss->hash_bits = 0;
 		return 0;
@@ -350,11 +368,100 @@ static int nicvf_rss_init(struct nicvf *nic)
 
 	for (idx = 0; idx < rss->rss_size; idx++)
 		rss->ind_tbl[idx] = ethtool_rxfh_indir_default(idx,
-							       nic->qs->rq_cnt);
+							       nic->rx_queues);
 	nicvf_config_rss(nic);
 	return 1;
 }
 
+/* Request PF to allocate additional Qsets */
+static void nicvf_request_sqs(struct nicvf *nic)
+{
+	union nic_mbx mbx = {};
+	int sqs;
+	int sqs_count = nic->sqs_count;
+	int rx_queues = 0, tx_queues = 0;
+
+	/* Only primary VF should request */
+	if (nic->sqs_mode ||  !nic->sqs_count)
+		return;
+
+	mbx.sqs_alloc.msg = NIC_MBOX_MSG_ALLOC_SQS;
+	mbx.sqs_alloc.vf_id = nic->vf_id;
+	mbx.sqs_alloc.qs_count = nic->sqs_count;
+	if (nicvf_send_msg_to_pf(nic, &mbx)) {
+		/* No response from PF */
+		nic->sqs_count = 0;
+		return;
+	}
+
+	/* Return if no Secondary Qsets available */
+	if (!nic->sqs_count)
+		return;
+
+	if (nic->rx_queues > MAX_RCV_QUEUES_PER_QS)
+		rx_queues = nic->rx_queues - MAX_RCV_QUEUES_PER_QS;
+	if (nic->tx_queues > MAX_SND_QUEUES_PER_QS)
+		tx_queues = nic->tx_queues - MAX_SND_QUEUES_PER_QS;
+
+	/* Set no of Rx/Tx queues in each of the SQsets */
+	for (sqs = 0; sqs < nic->sqs_count; sqs++) {
+		mbx.nicvf.msg = NIC_MBOX_MSG_SNICVF_PTR;
+		mbx.nicvf.vf_id = nic->vf_id;
+		mbx.nicvf.sqs_id = sqs;
+		nicvf_send_msg_to_pf(nic, &mbx);
+
+		nic->snicvf[sqs]->sqs_id = sqs;
+		if (rx_queues > MAX_RCV_QUEUES_PER_QS) {
+			nic->snicvf[sqs]->qs->rq_cnt = MAX_RCV_QUEUES_PER_QS;
+			rx_queues -= MAX_RCV_QUEUES_PER_QS;
+		} else {
+			nic->snicvf[sqs]->qs->rq_cnt = rx_queues;
+			rx_queues = 0;
+		}
+
+		if (tx_queues > MAX_SND_QUEUES_PER_QS) {
+			nic->snicvf[sqs]->qs->sq_cnt = MAX_SND_QUEUES_PER_QS;
+			tx_queues -= MAX_SND_QUEUES_PER_QS;
+		} else {
+			nic->snicvf[sqs]->qs->sq_cnt = tx_queues;
+			tx_queues = 0;
+		}
+
+		nic->snicvf[sqs]->qs->cq_cnt =
+		max(nic->snicvf[sqs]->qs->rq_cnt, nic->snicvf[sqs]->qs->sq_cnt);
+
+		/* Initialize secondary Qset's queues and its interrupts */
+		nicvf_open(nic->snicvf[sqs]->netdev);
+	}
+
+	/* Update stack with actual Rx/Tx queue count allocated */
+	if (sqs_count != nic->sqs_count)
+		nicvf_set_real_num_queues(nic->netdev,
+					  nic->tx_queues, nic->rx_queues);
+}
+
+/* Send this Qset's nicvf pointer to PF.
+ * PF inturn sends primary VF's nicvf struct to secondary Qsets/VFs
+ * so that packets received by these Qsets can use primary VF's netdev
+ */
+static void nicvf_send_vf_struct(struct nicvf *nic)
+{
+	union nic_mbx mbx = {};
+
+	mbx.nicvf.msg = NIC_MBOX_MSG_NICVF_PTR;
+	mbx.nicvf.sqs_mode = nic->sqs_mode;
+	mbx.nicvf.nicvf = (u64)nic;
+	nicvf_send_msg_to_pf(nic, &mbx);
+}
+
+static void nicvf_get_primary_vf_struct(struct nicvf *nic)
+{
+	union nic_mbx mbx = {};
+
+	mbx.nicvf.msg = NIC_MBOX_MSG_PNICVF_PTR;
+	nicvf_send_msg_to_pf(nic, &mbx);
+}
+
 int nicvf_set_real_num_queues(struct net_device *netdev,
 			      int tx_queues, int rx_queues)
 {
@@ -429,6 +536,34 @@ static void nicvf_snd_pkt_handler(struct net_device *netdev,
 	}
 }
 
+static inline void nicvf_set_rxhash(struct net_device *netdev,
+				    struct cqe_rx_t *cqe_rx,
+				    struct sk_buff *skb)
+{
+	u8 hash_type;
+	u32 hash;
+
+	if (!(netdev->features & NETIF_F_RXHASH))
+		return;
+
+	switch (cqe_rx->rss_alg) {
+	case RSS_ALG_TCP_IP:
+	case RSS_ALG_UDP_IP:
+		hash_type = PKT_HASH_TYPE_L4;
+		hash = cqe_rx->rss_tag;
+		break;
+	case RSS_ALG_IP:
+		hash_type = PKT_HASH_TYPE_L3;
+		hash = cqe_rx->rss_tag;
+		break;
+	default:
+		hash_type = PKT_HASH_TYPE_NONE;
+		hash = 0;
+	}
+
+	skb_set_hash(skb, hash, hash_type);
+}
+
 static void nicvf_rcv_pkt_handler(struct net_device *netdev,
 				  struct napi_struct *napi,
 				  struct cmp_queue *cq,
@@ -437,6 +572,15 @@ static void nicvf_rcv_pkt_handler(struct net_device *netdev,
 	struct sk_buff *skb;
 	struct nicvf *nic = netdev_priv(netdev);
 	int err = 0;
+	int rq_idx;
+
+	rq_idx = nicvf_netdev_qidx(nic, cqe_rx->rq_idx);
+
+	if (nic->sqs_mode) {
+		/* Use primary VF's 'nicvf' struct */
+		nic = nic->pnicvf;
+		netdev = nic->netdev;
+	}
 
 	/* Check for errors */
 	err = nicvf_check_cqe_rx_errs(nic, cq, cqe_rx);
@@ -456,9 +600,17 @@ static void nicvf_rcv_pkt_handler(struct net_device *netdev,
 			       skb->data, skb->len, true);
 	}
 
+	/* If error packet, drop it here */
+	if (err) {
+		dev_kfree_skb_any(skb);
+		return;
+	}
+
 	nicvf_set_rx_frame_cnt(nic, skb);
 
-	skb_record_rx_queue(skb, cqe_rx->rq_idx);
+	nicvf_set_rxhash(netdev, cqe_rx, skb);
+
+	skb_record_rx_queue(skb, rq_idx);
 	if (netdev->hw_features & NETIF_F_RXCSUM) {
 		/* HW by default verifies TCP/UDP/SCTP checksums */
 		skb->ip_summed = CHECKSUM_UNNECESSARY;
@@ -468,6 +620,11 @@ static void nicvf_rcv_pkt_handler(struct net_device *netdev,
 
 	skb->protocol = eth_type_trans(skb, netdev);
 
+	/* Check for stripped VLAN */
+	if (cqe_rx->vlan_found && cqe_rx->vlan_stripped)
+		__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q),
+				       ntohs((__force __be16)cqe_rx->vlan_tci));
+
 	if (napi && (netdev->features & NETIF_F_GRO))
 		napi_gro_receive(napi, skb);
 	else
@@ -549,8 +706,11 @@ static int nicvf_cq_intr_handler(struct net_device *netdev, u8 cq_idx,
 done:
 	/* Wakeup TXQ if its stopped earlier due to SQ full */
 	if (tx_done) {
-		txq = netdev_get_tx_queue(netdev, cq_idx);
-		if (netif_tx_queue_stopped(txq)) {
+		netdev = nic->pnicvf->netdev;
+		txq = netdev_get_tx_queue(netdev,
+					  nicvf_netdev_qidx(nic, cq_idx));
+		nic = nic->pnicvf;
+		if (netif_tx_queue_stopped(txq) && netif_carrier_ok(netdev)) {
 			netif_tx_start_queue(txq);
 			nic->drv_stats.txq_wake++;
 			if (netif_msg_tx_err(nic))
@@ -624,11 +784,20 @@ static void nicvf_handle_qs_err(unsigned long data)
 	nicvf_enable_intr(nic, NICVF_INTR_QS_ERR, 0);
 }
 
+static void nicvf_dump_intr_status(struct nicvf *nic)
+{
+	if (netif_msg_intr(nic))
+		netdev_info(nic->netdev, "%s: interrupt status 0x%llx\n",
+			    nic->netdev->name, nicvf_reg_read(nic, NIC_VF_INT));
+}
+
 static irqreturn_t nicvf_misc_intr_handler(int irq, void *nicvf_irq)
 {
 	struct nicvf *nic = (struct nicvf *)nicvf_irq;
 	u64 intr;
 
+	nicvf_dump_intr_status(nic);
+
 	intr = nicvf_reg_read(nic, NIC_VF_INT);
 	/* Check for spurious interrupt */
 	if (!(intr & NICVF_INTR_MBOX_MASK))
@@ -639,59 +808,58 @@ static irqreturn_t nicvf_misc_intr_handler(int irq, void *nicvf_irq)
 	return IRQ_HANDLED;
 }
 
-static irqreturn_t nicvf_intr_handler(int irq, void *nicvf_irq)
+static irqreturn_t nicvf_intr_handler(int irq, void *cq_irq)
+{
+	struct nicvf_cq_poll *cq_poll = (struct nicvf_cq_poll *)cq_irq;
+	struct nicvf *nic = cq_poll->nicvf;
+	int qidx = cq_poll->cq_idx;
+
+	nicvf_dump_intr_status(nic);
+
+	/* Disable interrupts */
+	nicvf_disable_intr(nic, NICVF_INTR_CQ, qidx);
+
+	/* Schedule NAPI */
+	napi_schedule(&cq_poll->napi);
+
+	/* Clear interrupt */
+	nicvf_clear_intr(nic, NICVF_INTR_CQ, qidx);
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t nicvf_rbdr_intr_handler(int irq, void *nicvf_irq)
 {
-	u64 qidx, intr, clear_intr = 0;
-	u64 cq_intr, rbdr_intr, qs_err_intr;
 	struct nicvf *nic = (struct nicvf *)nicvf_irq;
-	struct queue_set *qs = nic->qs;
-	struct nicvf_cq_poll *cq_poll = NULL;
+	u8 qidx;
 
-	intr = nicvf_reg_read(nic, NIC_VF_INT);
-	if (netif_msg_intr(nic))
-		netdev_info(nic->netdev, "%s: interrupt status 0x%llx\n",
-			    nic->netdev->name, intr);
-
-	qs_err_intr = intr & NICVF_INTR_QS_ERR_MASK;
-	if (qs_err_intr) {
-		/* Disable Qset err interrupt and schedule softirq */
-		nicvf_disable_intr(nic, NICVF_INTR_QS_ERR, 0);
-		tasklet_hi_schedule(&nic->qs_err_task);
-		clear_intr |= qs_err_intr;
-	}
 
-	/* Disable interrupts and start polling */
-	cq_intr = (intr & NICVF_INTR_CQ_MASK) >> NICVF_INTR_CQ_SHIFT;
-	for (qidx = 0; qidx < qs->cq_cnt; qidx++) {
-		if (!(cq_intr & (1 << qidx)))
-			continue;
-		if (!nicvf_is_intr_enabled(nic, NICVF_INTR_CQ, qidx))
+	nicvf_dump_intr_status(nic);
+
+	/* Disable RBDR interrupt and schedule softirq */
+	for (qidx = 0; qidx < nic->qs->rbdr_cnt; qidx++) {
+		if (!nicvf_is_intr_enabled(nic, NICVF_INTR_RBDR, qidx))
 			continue;
+		nicvf_disable_intr(nic, NICVF_INTR_RBDR, qidx);
+		tasklet_hi_schedule(&nic->rbdr_task);
+		/* Clear interrupt */
+		nicvf_clear_intr(nic, NICVF_INTR_RBDR, qidx);
+	}
 
-		nicvf_disable_intr(nic, NICVF_INTR_CQ, qidx);
-		clear_intr |= ((1 << qidx) << NICVF_INTR_CQ_SHIFT);
+	return IRQ_HANDLED;
+}
 
-		cq_poll = nic->napi[qidx];
-		/* Schedule NAPI */
-		if (cq_poll)
-			napi_schedule(&cq_poll->napi);
-	}
+static irqreturn_t nicvf_qs_err_intr_handler(int irq, void *nicvf_irq)
+{
+	struct nicvf *nic = (struct nicvf *)nicvf_irq;
 
-	/* Handle RBDR interrupts */
-	rbdr_intr = (intr & NICVF_INTR_RBDR_MASK) >> NICVF_INTR_RBDR_SHIFT;
-	if (rbdr_intr) {
-		/* Disable RBDR interrupt and schedule softirq */
-		for (qidx = 0; qidx < qs->rbdr_cnt; qidx++) {
-			if (!nicvf_is_intr_enabled(nic, NICVF_INTR_RBDR, qidx))
-				continue;
-			nicvf_disable_intr(nic, NICVF_INTR_RBDR, qidx);
-			tasklet_hi_schedule(&nic->rbdr_task);
-			clear_intr |= ((1 << qidx) << NICVF_INTR_RBDR_SHIFT);
-		}
-	}
+	nicvf_dump_intr_status(nic);
+
+	/* Disable Qset err interrupt and schedule softirq */
+	nicvf_disable_intr(nic, NICVF_INTR_QS_ERR, 0);
+	tasklet_hi_schedule(&nic->qs_err_task);
+	nicvf_clear_intr(nic, NICVF_INTR_QS_ERR, 0);
 
-	/* Clear interrupts */
-	nicvf_reg_write(nic, NIC_VF_INT, clear_intr);
 	return IRQ_HANDLED;
 }
 
@@ -725,7 +893,7 @@ static void nicvf_disable_msix(struct nicvf *nic)
 
 static int nicvf_register_interrupts(struct nicvf *nic)
 {
-	int irq, free, ret = 0;
+	int irq, ret = 0;
 	int vector;
 
 	for_each_cq_irq(irq)
@@ -740,44 +908,42 @@ static int nicvf_register_interrupts(struct nicvf *nic)
 		sprintf(nic->irq_name[irq], "NICVF%d RBDR%d",
 			nic->vf_id, irq - NICVF_INTR_ID_RBDR);
 
-	/* Register all interrupts except mailbox */
-	for (irq = 0; irq < NICVF_INTR_ID_SQ; irq++) {
+	/* Register CQ interrupts */
+	for (irq = 0; irq < nic->qs->cq_cnt; irq++) {
 		vector = nic->msix_entries[irq].vector;
 		ret = request_irq(vector, nicvf_intr_handler,
-				  0, nic->irq_name[irq], nic);
+				  0, nic->irq_name[irq], nic->napi[irq]);
 		if (ret)
-			break;
+			goto err;
 		nic->irq_allocated[irq] = true;
 	}
 
-	for (irq = NICVF_INTR_ID_SQ; irq < NICVF_INTR_ID_MISC; irq++) {
+	/* Register RBDR interrupt */
+	for (irq = NICVF_INTR_ID_RBDR;
+	     irq < (NICVF_INTR_ID_RBDR + nic->qs->rbdr_cnt); irq++) {
 		vector = nic->msix_entries[irq].vector;
-		ret = request_irq(vector, nicvf_intr_handler,
+		ret = request_irq(vector, nicvf_rbdr_intr_handler,
 				  0, nic->irq_name[irq], nic);
 		if (ret)
-			break;
+			goto err;
 		nic->irq_allocated[irq] = true;
 	}
 
+	/* Register QS error interrupt */
 	sprintf(nic->irq_name[NICVF_INTR_ID_QS_ERR],
 		"NICVF%d Qset error", nic->vf_id);
-	if (!ret) {
-		vector = nic->msix_entries[NICVF_INTR_ID_QS_ERR].vector;
-		irq = NICVF_INTR_ID_QS_ERR;
-		ret = request_irq(vector, nicvf_intr_handler,
-				  0, nic->irq_name[irq], nic);
-		if (!ret)
-			nic->irq_allocated[irq] = true;
-	}
+	irq = NICVF_INTR_ID_QS_ERR;
+	ret = request_irq(nic->msix_entries[irq].vector,
+			  nicvf_qs_err_intr_handler,
+			  0, nic->irq_name[irq], nic);
+	if (!ret)
+		nic->irq_allocated[irq] = true;
 
-	if (ret) {
-		netdev_err(nic->netdev, "Request irq failed\n");
-		for (free = 0; free < irq; free++)
-			free_irq(nic->msix_entries[free].vector, nic);
-		return ret;
-	}
+err:
+	if (ret)
+		netdev_err(nic->netdev, "request_irq failed, vector %d\n", irq);
 
-	return 0;
+	return ret;
 }
 
 static void nicvf_unregister_interrupts(struct nicvf *nic)
@@ -786,8 +952,14 @@ static void nicvf_unregister_interrupts(struct nicvf *nic)
 
 	/* Free registered interrupts */
 	for (irq = 0; irq < nic->num_vec; irq++) {
-		if (nic->irq_allocated[irq])
+		if (!nic->irq_allocated[irq])
+			continue;
+
+		if (irq < NICVF_INTR_ID_SQ)
+			free_irq(nic->msix_entries[irq].vector, nic->napi[irq]);
+		else
 			free_irq(nic->msix_entries[irq].vector, nic);
+
 		nic->irq_allocated[irq] = false;
 	}
 
@@ -852,13 +1024,26 @@ static netdev_tx_t nicvf_xmit(struct sk_buff *skb, struct net_device *netdev)
 			netdev_warn(netdev,
 				    "%s: Transmit ring full, stopping SQ%d\n",
 				    netdev->name, qid);
-
 		return NETDEV_TX_BUSY;
 	}
 
 	return NETDEV_TX_OK;
 }
 
+static inline void nicvf_free_cq_poll(struct nicvf *nic)
+{
+	struct nicvf_cq_poll *cq_poll;
+	int qidx;
+
+	for (qidx = 0; qidx < nic->qs->cq_cnt; qidx++) {
+		cq_poll = nic->napi[qidx];
+		if (!cq_poll)
+			continue;
+		nic->napi[qidx] = NULL;
+		kfree(cq_poll);
+	}
+}
+
 int nicvf_stop(struct net_device *netdev)
 {
 	int irq, qidx;
@@ -871,6 +1056,17 @@ int nicvf_stop(struct net_device *netdev)
 	nicvf_send_msg_to_pf(nic, &mbx);
 
 	netif_carrier_off(netdev);
+	netif_tx_stop_all_queues(nic->netdev);
+
+	/* Teardown secondary qsets first */
+	if (!nic->sqs_mode) {
+		for (qidx = 0; qidx < nic->sqs_count; qidx++) {
+			if (!nic->snicvf[qidx])
+				continue;
+			nicvf_stop(nic->snicvf[qidx]->netdev);
+			nic->snicvf[qidx] = NULL;
+		}
+	}
 
 	/* Disable RBDR & QS error interrupts */
 	for (qidx = 0; qidx < qs->rbdr_cnt; qidx++) {
@@ -893,7 +1089,6 @@ int nicvf_stop(struct net_device *netdev)
 		cq_poll = nic->napi[qidx];
 		if (!cq_poll)
 			continue;
-		nic->napi[qidx] = NULL;
 		napi_synchronize(&cq_poll->napi);
 		/* CQ intr is enabled while napi_complete,
 		 * so disable it now
@@ -902,7 +1097,6 @@ int nicvf_stop(struct net_device *netdev)
 		nicvf_clear_intr(nic, NICVF_INTR_CQ, qidx);
 		napi_disable(&cq_poll->napi);
 		netif_napi_del(&cq_poll->napi);
-		kfree(cq_poll);
 	}
 
 	netif_tx_disable(netdev);
@@ -918,6 +1112,12 @@ int nicvf_stop(struct net_device *netdev)
 
 	nicvf_unregister_interrupts(nic);
 
+	nicvf_free_cq_poll(nic);
+
+	/* Clear multiqset info */
+	nic->pnicvf = nic;
+	nic->sqs_count = 0;
+
 	return 0;
 }
 
@@ -944,6 +1144,7 @@ int nicvf_open(struct net_device *netdev)
 			goto napi_del;
 		}
 		cq_poll->cq_idx = qidx;
+		cq_poll->nicvf = nic;
 		netif_napi_add(netdev, &cq_poll->napi, nicvf_poll,
 			       NAPI_POLL_WEIGHT);
 		napi_enable(&cq_poll->napi);
@@ -972,10 +1173,16 @@ int nicvf_open(struct net_device *netdev)
 
 	/* Configure CPI alorithm */
 	nic->cpi_alg = cpi_alg;
-	nicvf_config_cpi(nic);
+	if (!nic->sqs_mode)
+		nicvf_config_cpi(nic);
+
+	nicvf_request_sqs(nic);
+	if (nic->sqs_mode)
+		nicvf_get_primary_vf_struct(nic);
 
 	/* Configure receive side scaling */
-	nicvf_rss_init(nic);
+	if (!nic->sqs_mode)
+		nicvf_rss_init(nic);
 
 	err = nicvf_register_interrupts(nic);
 	if (err)
@@ -1011,6 +1218,8 @@ int nicvf_open(struct net_device *netdev)
 cleanup:
 	nicvf_disable_intr(nic, NICVF_INTR_MBOX, 0);
 	nicvf_unregister_interrupts(nic);
+	tasklet_kill(&nic->qs_err_task);
+	tasklet_kill(&nic->rbdr_task);
 napi_del:
 	for (qidx = 0; qidx < qs->cq_cnt; qidx++) {
 		cq_poll = nic->napi[qidx];
@@ -1018,9 +1227,8 @@ int nicvf_open(struct net_device *netdev)
 			continue;
 		napi_disable(&cq_poll->napi);
 		netif_napi_del(&cq_poll->napi);
-		kfree(cq_poll);
-		nic->napi[qidx] = NULL;
 	}
+	nicvf_free_cq_poll(nic);
 	return err;
 }
 
@@ -1077,7 +1285,6 @@ void nicvf_update_lmac_stats(struct nicvf *nic)
 {
 	int stat = 0;
 	union nic_mbx mbx = {};
-	int timeout;
 
 	if (!netif_running(nic->netdev))
 		return;
@@ -1087,14 +1294,9 @@ void nicvf_update_lmac_stats(struct nicvf *nic)
 	/* Rx stats */
 	mbx.bgx_stats.rx = 1;
 	while (stat < BGX_RX_STATS_COUNT) {
-		nic->bgx_stats_acked = 0;
 		mbx.bgx_stats.idx = stat;
-		nicvf_send_msg_to_pf(nic, &mbx);
-		timeout = 0;
-		while ((!nic->bgx_stats_acked) && (timeout < 10)) {
-			msleep(2);
-			timeout++;
-		}
+		if (nicvf_send_msg_to_pf(nic, &mbx))
+			return;
 		stat++;
 	}
 
@@ -1103,14 +1305,9 @@ void nicvf_update_lmac_stats(struct nicvf *nic)
 	/* Tx stats */
 	mbx.bgx_stats.rx = 0;
 	while (stat < BGX_TX_STATS_COUNT) {
-		nic->bgx_stats_acked = 0;
 		mbx.bgx_stats.idx = stat;
-		nicvf_send_msg_to_pf(nic, &mbx);
-		timeout = 0;
-		while ((!nic->bgx_stats_acked) && (timeout < 10)) {
-			msleep(2);
-			timeout++;
-		}
+		if (nicvf_send_msg_to_pf(nic, &mbx))
+			return;
 		stat++;
 	}
 }
@@ -1118,7 +1315,7 @@ void nicvf_update_lmac_stats(struct nicvf *nic)
 void nicvf_update_stats(struct nicvf *nic)
 {
 	int qidx;
-	struct nicvf_hw_stats *stats = &nic->stats;
+	struct nicvf_hw_stats *stats = &nic->hw_stats;
 	struct nicvf_drv_stats *drv_stats = &nic->drv_stats;
 	struct queue_set *qs = nic->qs;
 
@@ -1127,14 +1324,16 @@ void nicvf_update_stats(struct nicvf *nic)
 #define GET_TX_STATS(reg) \
 	nicvf_reg_read(nic, NIC_VNIC_TX_STAT_0_4 | (reg << 3))
 
-	stats->rx_bytes_ok = GET_RX_STATS(RX_OCTS);
-	stats->rx_ucast_frames_ok = GET_RX_STATS(RX_UCAST);
-	stats->rx_bcast_frames_ok = GET_RX_STATS(RX_BCAST);
-	stats->rx_mcast_frames_ok = GET_RX_STATS(RX_MCAST);
+	stats->rx_bytes = GET_RX_STATS(RX_OCTS);
+	stats->rx_ucast_frames = GET_RX_STATS(RX_UCAST);
+	stats->rx_bcast_frames = GET_RX_STATS(RX_BCAST);
+	stats->rx_mcast_frames = GET_RX_STATS(RX_MCAST);
 	stats->rx_fcs_errors = GET_RX_STATS(RX_FCS);
 	stats->rx_l2_errors = GET_RX_STATS(RX_L2ERR);
 	stats->rx_drop_red = GET_RX_STATS(RX_RED);
+	stats->rx_drop_red_bytes = GET_RX_STATS(RX_RED_OCTS);
 	stats->rx_drop_overrun = GET_RX_STATS(RX_ORUN);
+	stats->rx_drop_overrun_bytes = GET_RX_STATS(RX_ORUN_OCTS);
 	stats->rx_drop_bcast = GET_RX_STATS(RX_DRP_BCAST);
 	stats->rx_drop_mcast = GET_RX_STATS(RX_DRP_MCAST);
 	stats->rx_drop_l3_bcast = GET_RX_STATS(RX_DRP_L3BCAST);
@@ -1146,9 +1345,6 @@ void nicvf_update_stats(struct nicvf *nic)
 	stats->tx_mcast_frames_ok = GET_TX_STATS(TX_MCAST);
 	stats->tx_drops = GET_TX_STATS(TX_DROP);
 
-	drv_stats->rx_frames_ok = stats->rx_ucast_frames_ok +
-				  stats->rx_bcast_frames_ok +
-				  stats->rx_mcast_frames_ok;
 	drv_stats->tx_frames_ok = stats->tx_ucast_frames_ok +
 				  stats->tx_bcast_frames_ok +
 				  stats->tx_mcast_frames_ok;
@@ -1167,14 +1363,15 @@ static struct rtnl_link_stats64 *nicvf_get_stats64(struct net_device *netdev,
 					    struct rtnl_link_stats64 *stats)
 {
 	struct nicvf *nic = netdev_priv(netdev);
-	struct nicvf_hw_stats *hw_stats = &nic->stats;
+	struct nicvf_hw_stats *hw_stats = &nic->hw_stats;
 	struct nicvf_drv_stats *drv_stats = &nic->drv_stats;
 
 	nicvf_update_stats(nic);
 
-	stats->rx_bytes = hw_stats->rx_bytes_ok;
+	stats->rx_bytes = hw_stats->rx_bytes;
 	stats->rx_packets = drv_stats->rx_frames_ok;
 	stats->rx_dropped = drv_stats->rx_drops;
+	stats->multicast = hw_stats->rx_mcast_frames;
 
 	stats->tx_bytes = hw_stats->tx_bytes_ok;
 	stats->tx_packets = drv_stats->tx_frames_ok;
@@ -1208,6 +1405,45 @@ static void nicvf_reset_task(struct work_struct *work)
 	nic->netdev->trans_start = jiffies;
 }
 
+static int nicvf_config_loopback(struct nicvf *nic,
+				 netdev_features_t features)
+{
+	union nic_mbx mbx = {};
+
+	mbx.lbk.msg = NIC_MBOX_MSG_LOOPBACK;
+	mbx.lbk.vf_id = nic->vf_id;
+	mbx.lbk.enable = (features & NETIF_F_LOOPBACK) != 0;
+
+	return nicvf_send_msg_to_pf(nic, &mbx);
+}
+
+static netdev_features_t nicvf_fix_features(struct net_device *netdev,
+					    netdev_features_t features)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+
+	if ((features & NETIF_F_LOOPBACK) &&
+	    netif_running(netdev) && !nic->loopback_supported)
+		features &= ~NETIF_F_LOOPBACK;
+
+	return features;
+}
+
+static int nicvf_set_features(struct net_device *netdev,
+			      netdev_features_t features)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+	netdev_features_t changed = features ^ netdev->features;
+
+	if (changed & NETIF_F_HW_VLAN_CTAG_RX)
+		nicvf_config_vlan_stripping(nic, features);
+
+	if ((changed & NETIF_F_LOOPBACK) && netif_running(netdev))
+		return nicvf_config_loopback(nic, features);
+
+	return 0;
+}
+
 static const struct net_device_ops nicvf_netdev_ops = {
 	.ndo_open		= nicvf_open,
 	.ndo_stop		= nicvf_stop,
@@ -1216,6 +1452,8 @@ static const struct net_device_ops nicvf_netdev_ops = {
 	.ndo_set_mac_address	= nicvf_set_mac_address,
 	.ndo_get_stats64	= nicvf_get_stats64,
 	.ndo_tx_timeout         = nicvf_tx_timeout,
+	.ndo_fix_features       = nicvf_fix_features,
+	.ndo_set_features       = nicvf_set_features,
 };
 
 static int nicvf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
@@ -1223,8 +1461,7 @@ static int nicvf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	struct device *dev = &pdev->dev;
 	struct net_device *netdev;
 	struct nicvf *nic;
-	struct queue_set *qs;
-	int    err;
+	int    err, qcount;
 
 	err = pci_enable_device(pdev);
 	if (err) {
@@ -1250,9 +1487,17 @@ static int nicvf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 		goto err_release_regions;
 	}
 
-	netdev = alloc_etherdev_mqs(sizeof(struct nicvf),
-				    MAX_RCV_QUEUES_PER_QS,
-				    MAX_SND_QUEUES_PER_QS);
+	qcount = MAX_CMP_QUEUES_PER_QS;
+
+	/* Restrict multiqset support only for host bound VFs */
+	if (pdev->is_virtfn) {
+		/* Set max number of queues per VF */
+		qcount = roundup(num_online_cpus(), MAX_CMP_QUEUES_PER_QS);
+		qcount = min(qcount,
+			     (MAX_SQS_PER_VF + 1) * MAX_CMP_QUEUES_PER_QS);
+	}
+
+	netdev = alloc_etherdev_mqs(sizeof(struct nicvf), qcount, qcount);
 	if (!netdev) {
 		err = -ENOMEM;
 		goto err_release_regions;
@@ -1265,6 +1510,8 @@ static int nicvf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	nic = netdev_priv(netdev);
 	nic->netdev = netdev;
 	nic->pdev = pdev;
+	nic->pnicvf = nic;
+	nic->max_queues = qcount;
 
 	/* MAP VF's configuration registers */
 	nic->reg_base = pcim_iomap(pdev, PCI_CFG_REG_BAR_NUM, 0);
@@ -1278,20 +1525,31 @@ static int nicvf_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (err)
 		goto err_free_netdev;
 
-	qs = nic->qs;
-
-	err = nicvf_set_real_num_queues(netdev, qs->sq_cnt, qs->rq_cnt);
-	if (err)
-		goto err_free_netdev;
-
 	/* Check if PF is alive and get MAC address for this VF */
 	err = nicvf_register_misc_interrupt(nic);
 	if (err)
 		goto err_free_netdev;
 
-	netdev->features |= (NETIF_F_RXCSUM | NETIF_F_IP_CSUM | NETIF_F_SG |
-			     NETIF_F_TSO | NETIF_F_GRO);
-	netdev->hw_features = netdev->features;
+	nicvf_send_vf_struct(nic);
+
+	/* Check if this VF is in QS only mode */
+	if (nic->sqs_mode)
+		return 0;
+
+	err = nicvf_set_real_num_queues(netdev, nic->tx_queues, nic->rx_queues);
+	if (err)
+		goto err_unregister_interrupts;
+
+	netdev->hw_features = (NETIF_F_RXCSUM | NETIF_F_IP_CSUM | NETIF_F_SG |
+			       NETIF_F_TSO | NETIF_F_GRO |
+			       NETIF_F_HW_VLAN_CTAG_RX);
+
+	netdev->hw_features |= NETIF_F_RXHASH;
+
+	netdev->features |= netdev->hw_features;
+	netdev->hw_features |= NETIF_F_LOOPBACK;
+
+	netdev->vlan_features = NETIF_F_SG | NETIF_F_IP_CSUM | NETIF_F_TSO;
 
 	netdev->netdev_ops = &nicvf_netdev_ops;
 	netdev->watchdog_timeo = NICVF_TX_TIMEOUT;
@@ -1326,8 +1584,13 @@ static void nicvf_remove(struct pci_dev *pdev)
 {
 	struct net_device *netdev = pci_get_drvdata(pdev);
 	struct nicvf *nic = netdev_priv(netdev);
+	struct net_device *pnetdev = nic->pnicvf->netdev;
 
-	unregister_netdev(netdev);
+	/* Check if this Qset is assigned to different VF.
+	 * If yes, clean primary and all secondary Qsets.
+	 */
+	if (pnetdev && (pnetdev->reg_state == NETREG_REGISTERED))
+		unregister_netdev(pnetdev);
 	nicvf_unregister_interrupts(nic);
 	pci_set_drvdata(pdev, NULL);
 	free_netdev(netdev);

drivers/net/ethernet/cavium/thunder/nicvf_queues.c

@@ -475,6 +475,27 @@ static void nicvf_reclaim_rbdr(struct nicvf *nic,
 		return;
 }
 
+void nicvf_config_vlan_stripping(struct nicvf *nic, netdev_features_t features)
+{
+	u64 rq_cfg;
+	int sqs;
+
+	rq_cfg = nicvf_queue_reg_read(nic, NIC_QSET_RQ_GEN_CFG, 0);
+
+	/* Enable first VLAN stripping */
+	if (features & NETIF_F_HW_VLAN_CTAG_RX)
+		rq_cfg |= (1ULL << 25);
+	else
+		rq_cfg &= ~(1ULL << 25);
+	nicvf_queue_reg_write(nic, NIC_QSET_RQ_GEN_CFG, 0, rq_cfg);
+
+	/* Configure Secondary Qsets, if any */
+	for (sqs = 0; sqs < nic->sqs_count; sqs++)
+		if (nic->snicvf[sqs])
+			nicvf_queue_reg_write(nic->snicvf[sqs],
+					      NIC_QSET_RQ_GEN_CFG, 0, rq_cfg);
+}
+
 /* Configures receive queue */
 static void nicvf_rcv_queue_config(struct nicvf *nic, struct queue_set *qs,
 				   int qidx, bool enable)
@@ -524,7 +545,9 @@ static void nicvf_rcv_queue_config(struct nicvf *nic, struct queue_set *qs,
 	mbx.rq.cfg = (1ULL << 62) | (RQ_CQ_DROP << 8);
 	nicvf_send_msg_to_pf(nic, &mbx);
 
-	nicvf_queue_reg_write(nic, NIC_QSET_RQ_GEN_CFG, qidx, 0x00);
+	nicvf_queue_reg_write(nic, NIC_QSET_RQ_GEN_CFG, 0, 0x00);
+	if (!nic->sqs_mode)
+		nicvf_config_vlan_stripping(nic, nic->netdev->features);
 
 	/* Enable Receive queue */
 	rq_cfg.ena = 1;
@@ -598,6 +621,7 @@ static void nicvf_snd_queue_config(struct nicvf *nic, struct queue_set *qs,
 	mbx.sq.msg = NIC_MBOX_MSG_SQ_CFG;
 	mbx.sq.qs_num = qs->vnic_id;
 	mbx.sq.sq_num = qidx;
+	mbx.sq.sqs_mode = nic->sqs_mode;
 	mbx.sq.cfg = (sq->cq_qs << 3) | sq->cq_idx;
 	nicvf_send_msg_to_pf(nic, &mbx);
 
@@ -679,6 +703,7 @@ void nicvf_qset_config(struct nicvf *nic, bool enable)
 	/* Send a mailbox msg to PF to config Qset */
 	mbx.qs.msg = NIC_MBOX_MSG_QS_CFG;
 	mbx.qs.num = qs->vnic_id;
+	mbx.qs.sqs_count = nic->sqs_count;
 
 	mbx.qs.cfg = 0;
 	qs_cfg = (struct qs_cfg *)&mbx.qs.cfg;
@@ -759,6 +784,10 @@ int nicvf_set_qset_resources(struct nicvf *nic)
 	qs->rbdr_len = RCV_BUF_COUNT;
 	qs->sq_len = SND_QUEUE_LEN;
 	qs->cq_len = CMP_QUEUE_LEN;
+
+	nic->rx_queues = qs->rq_cnt;
+	nic->tx_queues = qs->sq_cnt;
+
 	return 0;
 }
 
@@ -961,9 +990,6 @@ nicvf_sq_add_hdr_subdesc(struct snd_queue *sq, int qentry,
 
 	/* Offload checksum calculation to HW */
 	if (skb->ip_summed == CHECKSUM_PARTIAL) {
-		if (skb->protocol != htons(ETH_P_IP))
-			return;
-
 		hdr->csum_l3 = 1; /* Enable IP csum calculation */
 		hdr->l3_offset = skb_network_offset(skb);
 		hdr->l4_offset = skb_transport_offset(skb);
@@ -1005,7 +1031,7 @@ static inline void nicvf_sq_add_gather_subdesc(struct snd_queue *sq, int qentry,
  * them to SQ for transfer
  */
 static int nicvf_sq_append_tso(struct nicvf *nic, struct snd_queue *sq,
-			       int qentry, struct sk_buff *skb)
+			       int sq_num, int qentry, struct sk_buff *skb)
 {
 	struct tso_t tso;
 	int seg_subdescs = 0, desc_cnt = 0;
@@ -1065,7 +1091,7 @@ static int nicvf_sq_append_tso(struct nicvf *nic, struct snd_queue *sq,
 
 	/* Inform HW to xmit all TSO segments */
 	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_DOOR,
-			      skb_get_queue_mapping(skb), desc_cnt);
+			      sq_num, desc_cnt);
 	nic->drv_stats.tx_tso++;
 	return 1;
 }
@@ -1076,10 +1102,24 @@ int nicvf_sq_append_skb(struct nicvf *nic, struct sk_buff *skb)
 	int i, size;
 	int subdesc_cnt;
 	int sq_num, qentry;
-	struct queue_set *qs = nic->qs;
+	struct queue_set *qs;
 	struct snd_queue *sq;
 
 	sq_num = skb_get_queue_mapping(skb);
+	if (sq_num >= MAX_SND_QUEUES_PER_QS) {
+		/* Get secondary Qset's SQ structure */
+		i = sq_num / MAX_SND_QUEUES_PER_QS;
+		if (!nic->snicvf[i - 1]) {
+			netdev_warn(nic->netdev,
+				    "Secondary Qset#%d's ptr not initialized\n",
+				    i - 1);
+			return 1;
+		}
+		nic = (struct nicvf *)nic->snicvf[i - 1];
+		sq_num = sq_num % MAX_SND_QUEUES_PER_QS;
+	}
+
+	qs = nic->qs;
 	sq = &qs->sq[sq_num];
 
 	subdesc_cnt = nicvf_sq_subdesc_required(nic, skb);
@@ -1090,7 +1130,7 @@ int nicvf_sq_append_skb(struct nicvf *nic, struct sk_buff *skb)
 
 	/* Check if its a TSO packet */
 	if (skb_shinfo(skb)->gso_size)
-		return nicvf_sq_append_tso(nic, sq, qentry, skb);
+		return nicvf_sq_append_tso(nic, sq, sq_num, qentry, skb);
 
 	/* Add SQ header subdesc */
 	nicvf_sq_add_hdr_subdesc(sq, qentry, subdesc_cnt - 1, skb, skb->len);
@@ -1126,6 +1166,8 @@ int nicvf_sq_append_skb(struct nicvf *nic, struct sk_buff *skb)
 	return 1;
 
 append_fail:
+	/* Use original PCI dev for debug log */
+	nic = nic->pnicvf;
 	netdev_dbg(nic->netdev, "Not enough SQ descriptors to xmit pkt\n");
 	return 0;
 }
@@ -1371,10 +1413,11 @@ void nicvf_update_sq_stats(struct nicvf *nic, int sq_idx)
 int nicvf_check_cqe_rx_errs(struct nicvf *nic,
 			    struct cmp_queue *cq, struct cqe_rx_t *cqe_rx)
 {
-	struct cmp_queue_stats *stats = &cq->stats;
+	struct nicvf_hw_stats *stats = &nic->hw_stats;
+	struct nicvf_drv_stats *drv_stats = &nic->drv_stats;
 
 	if (!cqe_rx->err_level && !cqe_rx->err_opcode) {
-		stats->rx.errop.good++;
+		drv_stats->rx_frames_ok++;
 		return 0;
 	}
 
@@ -1384,111 +1427,78 @@ int nicvf_check_cqe_rx_errs(struct nicvf *nic,
 			   nic->netdev->name,
 			   cqe_rx->err_level, cqe_rx->err_opcode);
 
-	switch (cqe_rx->err_level) {
-	case CQ_ERRLVL_MAC:
-		stats->rx.errlvl.mac_errs++;
-		break;
-	case CQ_ERRLVL_L2:
-		stats->rx.errlvl.l2_errs++;
-		break;
-	case CQ_ERRLVL_L3:
-		stats->rx.errlvl.l3_errs++;
-		break;
-	case CQ_ERRLVL_L4:
-		stats->rx.errlvl.l4_errs++;
-		break;
-	}
-
 	switch (cqe_rx->err_opcode) {
 	case CQ_RX_ERROP_RE_PARTIAL:
-		stats->rx.errop.partial_pkts++;
+		stats->rx_bgx_truncated_pkts++;
 		break;
 	case CQ_RX_ERROP_RE_JABBER:
-		stats->rx.errop.jabber_errs++;
+		stats->rx_jabber_errs++;
 		break;
 	case CQ_RX_ERROP_RE_FCS:
-		stats->rx.errop.fcs_errs++;
-		break;
-	case CQ_RX_ERROP_RE_TERMINATE:
-		stats->rx.errop.terminate_errs++;
+		stats->rx_fcs_errs++;
 		break;
 	case CQ_RX_ERROP_RE_RX_CTL:
-		stats->rx.errop.bgx_rx_errs++;
+		stats->rx_bgx_errs++;
 		break;
 	case CQ_RX_ERROP_PREL2_ERR:
-		stats->rx.errop.prel2_errs++;
-		break;
-	case CQ_RX_ERROP_L2_FRAGMENT:
-		stats->rx.errop.l2_frags++;
-		break;
-	case CQ_RX_ERROP_L2_OVERRUN:
-		stats->rx.errop.l2_overruns++;
-		break;
-	case CQ_RX_ERROP_L2_PFCS:
-		stats->rx.errop.l2_pfcs++;
-		break;
-	case CQ_RX_ERROP_L2_PUNY:
-		stats->rx.errop.l2_puny++;
+		stats->rx_prel2_errs++;
 		break;
 	case CQ_RX_ERROP_L2_MAL:
-		stats->rx.errop.l2_hdr_malformed++;
+		stats->rx_l2_hdr_malformed++;
 		break;
 	case CQ_RX_ERROP_L2_OVERSIZE:
-		stats->rx.errop.l2_oversize++;
+		stats->rx_oversize++;
 		break;
 	case CQ_RX_ERROP_L2_UNDERSIZE:
-		stats->rx.errop.l2_undersize++;
+		stats->rx_undersize++;
 		break;
 	case CQ_RX_ERROP_L2_LENMISM:
-		stats->rx.errop.l2_len_mismatch++;
+		stats->rx_l2_len_mismatch++;
 		break;
 	case CQ_RX_ERROP_L2_PCLP:
-		stats->rx.errop.l2_pclp++;
+		stats->rx_l2_pclp++;
 		break;
 	case CQ_RX_ERROP_IP_NOT:
-		stats->rx.errop.non_ip++;
+		stats->rx_ip_ver_errs++;
 		break;
 	case CQ_RX_ERROP_IP_CSUM_ERR:
-		stats->rx.errop.ip_csum_err++;
+		stats->rx_ip_csum_errs++;
 		break;
 	case CQ_RX_ERROP_IP_MAL:
-		stats->rx.errop.ip_hdr_malformed++;
+		stats->rx_ip_hdr_malformed++;
 		break;
 	case CQ_RX_ERROP_IP_MALD:
-		stats->rx.errop.ip_payload_malformed++;
+		stats->rx_ip_payload_malformed++;
 		break;
 	case CQ_RX_ERROP_IP_HOP:
-		stats->rx.errop.ip_hop_errs++;
-		break;
-	case CQ_RX_ERROP_L3_ICRC:
-		stats->rx.errop.l3_icrc_errs++;
+		stats->rx_ip_ttl_errs++;
 		break;
 	case CQ_RX_ERROP_L3_PCLP:
-		stats->rx.errop.l3_pclp++;
+		stats->rx_l3_pclp++;
 		break;
 	case CQ_RX_ERROP_L4_MAL:
-		stats->rx.errop.l4_malformed++;
+		stats->rx_l4_malformed++;
 		break;
 	case CQ_RX_ERROP_L4_CHK:
-		stats->rx.errop.l4_csum_errs++;
+		stats->rx_l4_csum_errs++;
 		break;
 	case CQ_RX_ERROP_UDP_LEN:
-		stats->rx.errop.udp_len_err++;
+		stats->rx_udp_len_errs++;
 		break;
 	case CQ_RX_ERROP_L4_PORT:
-		stats->rx.errop.bad_l4_port++;
+		stats->rx_l4_port_errs++;
 		break;
 	case CQ_RX_ERROP_TCP_FLAG:
-		stats->rx.errop.bad_tcp_flag++;
+		stats->rx_tcp_flag_errs++;
 		break;
 	case CQ_RX_ERROP_TCP_OFFSET:
-		stats->rx.errop.tcp_offset_errs++;
+		stats->rx_tcp_offset_errs++;
 		break;
 	case CQ_RX_ERROP_L4_PCLP:
-		stats->rx.errop.l4_pclp++;
+		stats->rx_l4_pclp++;
 		break;
 	case CQ_RX_ERROP_RBDR_TRUNC:
-		stats->rx.errop.pkt_truncated++;
+		stats->rx_truncated_pkts++;
 		break;
 	}
 

drivers/net/ethernet/cavium/thunder/nicvf_queues.h

@@ -181,47 +181,6 @@ enum CQ_TX_ERROP_E {
 };
 
 struct cmp_queue_stats {
-	struct rx_stats {
-		struct {
-			u64 mac_errs;
-			u64 l2_errs;
-			u64 l3_errs;
-			u64 l4_errs;
-		} errlvl;
-		struct {
-			u64 good;
-			u64 partial_pkts;
-			u64 jabber_errs;
-			u64 fcs_errs;
-			u64 terminate_errs;
-			u64 bgx_rx_errs;
-			u64 prel2_errs;
-			u64 l2_frags;
-			u64 l2_overruns;
-			u64 l2_pfcs;
-			u64 l2_puny;
-			u64 l2_hdr_malformed;
-			u64 l2_oversize;
-			u64 l2_undersize;
-			u64 l2_len_mismatch;
-			u64 l2_pclp;
-			u64 non_ip;
-			u64 ip_csum_err;
-			u64 ip_hdr_malformed;
-			u64 ip_payload_malformed;
-			u64 ip_hop_errs;
-			u64 l3_icrc_errs;
-			u64 l3_pclp;
-			u64 l4_malformed;
-			u64 l4_csum_errs;
-			u64 udp_len_err;
-			u64 bad_l4_port;
-			u64 bad_tcp_flag;
-			u64 tcp_offset_errs;
-			u64 l4_pclp;
-			u64 pkt_truncated;
-		} errop;
-	} rx;
 	struct tx_stats {
 		u64 good;
 		u64 desc_fault;
@@ -292,6 +251,7 @@ struct cmp_queue {
 	void		*desc;
 	struct q_desc_mem   dmem;
 	struct cmp_queue_stats	stats;
+	int		irq;
 } ____cacheline_aligned_in_smp;
 
 struct snd_queue {
@@ -347,6 +307,8 @@ struct queue_set {
 
 #define	CQ_ERR_MASK	(CQ_WR_FULL | CQ_WR_DISABLE | CQ_WR_FAULT)
 
+void nicvf_config_vlan_stripping(struct nicvf *nic,
+				 netdev_features_t features);
 int nicvf_set_qset_resources(struct nicvf *nic);
 int nicvf_config_data_transfer(struct nicvf *nic, bool enable);
 void nicvf_qset_config(struct nicvf *nic, bool enable);

drivers/net/ethernet/cavium/thunder/thunder_bgx.c

@@ -329,6 +329,37 @@ static void bgx_flush_dmac_addrs(struct bgx *bgx, int lmac)
 	}
 }
 
+/* Configure BGX LMAC in internal loopback mode */
+void bgx_lmac_internal_loopback(int node, int bgx_idx,
+				int lmac_idx, bool enable)
+{
+	struct bgx *bgx;
+	struct lmac *lmac;
+	u64    cfg;
+
+	bgx = bgx_vnic[(node * MAX_BGX_PER_CN88XX) + bgx_idx];
+	if (!bgx)
+		return;
+
+	lmac = &bgx->lmac[lmac_idx];
+	if (lmac->is_sgmii) {
+		cfg = bgx_reg_read(bgx, lmac_idx, BGX_GMP_PCS_MRX_CTL);
+		if (enable)
+			cfg |= PCS_MRX_CTL_LOOPBACK1;
+		else
+			cfg &= ~PCS_MRX_CTL_LOOPBACK1;
+		bgx_reg_write(bgx, lmac_idx, BGX_GMP_PCS_MRX_CTL, cfg);
+	} else {
+		cfg = bgx_reg_read(bgx, lmac_idx, BGX_SPUX_CONTROL1);
+		if (enable)
+			cfg |= SPU_CTL_LOOPBACK;
+		else
+			cfg &= ~SPU_CTL_LOOPBACK;
+		bgx_reg_write(bgx, lmac_idx, BGX_SPUX_CONTROL1, cfg);
+	}
+}
+EXPORT_SYMBOL(bgx_lmac_internal_loopback);
+
 static int bgx_lmac_sgmii_init(struct bgx *bgx, int lmacid)
 {
 	u64 cfg;

drivers/net/ethernet/cavium/thunder/thunder_bgx.h

@@ -72,6 +72,7 @@
 
 #define BGX_SPUX_CONTROL1		0x10000
 #define  SPU_CTL_LOW_POWER			BIT_ULL(11)
+#define  SPU_CTL_LOOPBACK			BIT_ULL(14)
 #define  SPU_CTL_RESET				BIT_ULL(15)
 #define BGX_SPUX_STATUS1		0x10008
 #define  SPU_STATUS1_RCV_LNK			BIT_ULL(2)
@@ -126,6 +127,7 @@
 #define	 PCS_MRX_CTL_RST_AN			BIT_ULL(9)
 #define	 PCS_MRX_CTL_PWR_DN			BIT_ULL(11)
 #define	 PCS_MRX_CTL_AN_EN			BIT_ULL(12)
+#define	 PCS_MRX_CTL_LOOPBACK1			BIT_ULL(14)
 #define	 PCS_MRX_CTL_RESET			BIT_ULL(15)
 #define BGX_GMP_PCS_MRX_STATUS		0x30008
 #define	 PCS_MRX_STATUS_AN_CPT			BIT_ULL(5)
@@ -186,6 +188,8 @@ int bgx_get_lmac_count(int node, int bgx);
 const u8 *bgx_get_lmac_mac(int node, int bgx_idx, int lmacid);
 void bgx_set_lmac_mac(int node, int bgx_idx, int lmacid, const u8 *mac);
 void bgx_get_lmac_link_state(int node, int bgx_idx, int lmacid, void *status);
+void bgx_lmac_internal_loopback(int node, int bgx_idx,
+				int lmac_idx, bool enable);
 u64 bgx_get_rx_stats(int node, int bgx_idx, int lmac, int idx);
 u64 bgx_get_tx_stats(int node, int bgx_idx, int lmac, int idx);
 #define BGX_RX_STATS_COUNT 11