Adding upstream version 6.1.76.upstream/6.1.76upstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

58 files changed, 40287 insertions, 0 deletions

diff --git a/drivers/net/ethernet/cavium/Kconfig b/drivers/net/ethernet/cavium/Kconfig
new file mode 100644
index 000000000..1c76c95b0
--- /dev/null
+++ b/drivers/net/ethernet/cavium/Kconfig
@@ -0,0 +1,103 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# Cavium ethernet device configuration
+#
+
+config NET_VENDOR_CAVIUM
+	bool "Cavium ethernet drivers"
+	default y
+	help
+	  Select this option if you want enable Cavium network support.
+
+	  If you have a Cavium SoC or network adapter, say Y.
+
+if NET_VENDOR_CAVIUM
+
+config THUNDER_NIC_PF
+	tristate "Thunder Physical function driver"
+	depends on 64BIT && PCI
+	select THUNDER_NIC_BGX
+	help
+	  This driver supports Thunder's NIC physical function.
+	  The NIC provides the controller and DMA engines to
+	  move network traffic to/from the memory. The NIC
+	  works closely with TNS, BGX and SerDes to implement the
+	  functions replacing and virtualizing those of a typical
+	  standalone PCIe NIC chip.
+
+config THUNDER_NIC_VF
+	tristate "Thunder Virtual function driver"
+	imply CAVIUM_PTP
+	depends on 64BIT && PCI
+	help
+	  This driver supports Thunder's NIC virtual function
+
+config	THUNDER_NIC_BGX
+	tristate "Thunder MAC interface driver (BGX)"
+	depends on 64BIT && PCI
+	select PHYLIB
+	select MDIO_THUNDER if PCI
+	select THUNDER_NIC_RGX
+	help
+	  This driver supports programming and controlling of MAC
+	  interface from NIC physical function driver.
+
+config	THUNDER_NIC_RGX
+	tristate "Thunder MAC interface driver (RGX)"
+	depends on 64BIT && PCI
+	select PHYLIB
+	select MDIO_THUNDER if PCI
+	help
+	  This driver supports configuring XCV block of RGX interface
+	  present on CN81XX chip.
+
+config CAVIUM_PTP
+	tristate "Cavium PTP coprocessor as PTP clock"
+	depends on 64BIT && PCI
+	depends on PTP_1588_CLOCK
+	help
+	  This driver adds support for the Precision Time Protocol Clocks and
+	  Timestamping coprocessor (PTP) found on Cavium processors.
+	  PTP provides timestamping mechanism that is suitable for use in IEEE 1588
+	  Precision Time Protocol or other purposes.  Timestamps can be used in
+	  BGX, TNS, GTI, and NIC blocks.
+
+config LIQUIDIO
+	tristate "Cavium LiquidIO support"
+	depends on 64BIT && PCI
+	depends on PCI
+	depends on PTP_1588_CLOCK_OPTIONAL
+	select FW_LOADER
+	select LIBCRC32C
+	select NET_DEVLINK
+	help
+	  This driver supports Cavium LiquidIO Intelligent Server Adapters
+	  based on CN66XX, CN68XX and CN23XX chips.
+
+	  To compile this driver as a module, choose M here: the module
+	  will be called liquidio.  This is recommended.
+
+config OCTEON_MGMT_ETHERNET
+	tristate "Octeon Management port ethernet driver (CN5XXX, CN6XXX)"
+	depends on CAVIUM_OCTEON_SOC
+	select PHYLIB
+	select MDIO_OCTEON
+	default y
+	help
+	  Enable the ethernet driver for the management
+	  port on Cavium Networks' Octeon CN57XX, CN56XX, CN55XX,
+	  CN54XX, CN52XX, and CN6XXX chips.
+
+config LIQUIDIO_VF
+	tristate "Cavium LiquidIO VF support"
+	depends on 64BIT && PCI_MSI
+	depends on PTP_1588_CLOCK_OPTIONAL
+	help
+	  This driver supports Cavium LiquidIO Intelligent Server Adapter
+	  based on CN23XX chips.
+
+	  To compile this driver as a module, choose M here: The module
+	  will be called liquidio_vf. MSI-X interrupt support is required
+	  for this driver to work correctly
+
+endif # NET_VENDOR_CAVIUM
diff --git a/drivers/net/ethernet/cavium/Makefile b/drivers/net/ethernet/cavium/Makefile
new file mode 100644
index 000000000..5d3280821
--- /dev/null
+++ b/drivers/net/ethernet/cavium/Makefile
@@ -0,0 +1,8 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# Makefile for the Cavium ethernet device drivers.
+#
+obj-$(CONFIG_NET_VENDOR_CAVIUM) += common/
+obj-$(CONFIG_NET_VENDOR_CAVIUM) += thunder/
+obj-$(CONFIG_NET_VENDOR_CAVIUM) += liquidio/
+obj-$(CONFIG_NET_VENDOR_CAVIUM) += octeon/
diff --git a/drivers/net/ethernet/cavium/common/Makefile b/drivers/net/ethernet/cavium/common/Makefile
new file mode 100644
index 000000000..e3f87bd65
--- /dev/null
+++ b/drivers/net/ethernet/cavium/common/Makefile
@@ -0,0 +1,2 @@
+# SPDX-License-Identifier: GPL-2.0-only
+obj-$(CONFIG_CAVIUM_PTP) += cavium_ptp.o
diff --git a/drivers/net/ethernet/cavium/common/cavium_ptp.c b/drivers/net/ethernet/cavium/common/cavium_ptp.c
new file mode 100644
index 000000000..9fd717b9c
--- /dev/null
+++ b/drivers/net/ethernet/cavium/common/cavium_ptp.c
@@ -0,0 +1,348 @@
+// SPDX-License-Identifier: GPL-2.0
+/* cavium_ptp.c - PTP 1588 clock on Cavium hardware
+ * Copyright (c) 2003-2015, 2017 Cavium, Inc.
+ */
+
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/timecounter.h>
+#include <linux/pci.h>
+
+#include "cavium_ptp.h"
+
+#define DRV_NAME "cavium_ptp"
+
+#define PCI_DEVICE_ID_CAVIUM_PTP	0xA00C
+#define PCI_SUBSYS_DEVID_88XX_PTP	0xA10C
+#define PCI_SUBSYS_DEVID_81XX_PTP	0XA20C
+#define PCI_SUBSYS_DEVID_83XX_PTP	0xA30C
+#define PCI_DEVICE_ID_CAVIUM_RST	0xA00E
+
+#define PCI_PTP_BAR_NO	0
+#define PCI_RST_BAR_NO	0
+
+#define PTP_CLOCK_CFG		0xF00ULL
+#define  PTP_CLOCK_CFG_PTP_EN	BIT(0)
+#define PTP_CLOCK_LO		0xF08ULL
+#define PTP_CLOCK_HI		0xF10ULL
+#define PTP_CLOCK_COMP		0xF18ULL
+
+#define RST_BOOT	0x1600ULL
+#define CLOCK_BASE_RATE	50000000ULL
+
+static u64 ptp_cavium_clock_get(void)
+{
+	struct pci_dev *pdev;
+	void __iomem *base;
+	u64 ret = CLOCK_BASE_RATE * 16;
+
+	pdev = pci_get_device(PCI_VENDOR_ID_CAVIUM,
+			      PCI_DEVICE_ID_CAVIUM_RST, NULL);
+	if (!pdev)
+		goto error;
+
+	base = pci_ioremap_bar(pdev, PCI_RST_BAR_NO);
+	if (!base)
+		goto error_put_pdev;
+
+	ret = CLOCK_BASE_RATE * ((readq(base + RST_BOOT) >> 33) & 0x3f);
+
+	iounmap(base);
+
+error_put_pdev:
+	pci_dev_put(pdev);
+
+error:
+	return ret;
+}
+
+struct cavium_ptp *cavium_ptp_get(void)
+{
+	struct cavium_ptp *ptp;
+	struct pci_dev *pdev;
+
+	pdev = pci_get_device(PCI_VENDOR_ID_CAVIUM,
+			      PCI_DEVICE_ID_CAVIUM_PTP, NULL);
+	if (!pdev)
+		return ERR_PTR(-ENODEV);
+
+	ptp = pci_get_drvdata(pdev);
+	if (!ptp)
+		ptp = ERR_PTR(-EPROBE_DEFER);
+	if (IS_ERR(ptp))
+		pci_dev_put(pdev);
+
+	return ptp;
+}
+EXPORT_SYMBOL(cavium_ptp_get);
+
+void cavium_ptp_put(struct cavium_ptp *ptp)
+{
+	if (!ptp)
+		return;
+	pci_dev_put(ptp->pdev);
+}
+EXPORT_SYMBOL(cavium_ptp_put);
+
+/**
+ * cavium_ptp_adjfine() - Adjust ptp frequency
+ * @ptp_info: PTP clock info
+ * @scaled_ppm: how much to adjust by, in parts per million, but with a
+ *              16 bit binary fractional field
+ */
+static int cavium_ptp_adjfine(struct ptp_clock_info *ptp_info, long scaled_ppm)
+{
+	struct cavium_ptp *clock =
+		container_of(ptp_info, struct cavium_ptp, ptp_info);
+	unsigned long flags;
+	u64 comp;
+	u64 adj;
+	bool neg_adj = false;
+
+	if (scaled_ppm < 0) {
+		neg_adj = true;
+		scaled_ppm = -scaled_ppm;
+	}
+
+	/* The hardware adds the clock compensation value to the PTP clock
+	 * on every coprocessor clock cycle. Typical convention is that it
+	 * represent number of nanosecond betwen each cycle. In this
+	 * convention compensation value is in 64 bit fixed-point
+	 * representation where upper 32 bits are number of nanoseconds
+	 * and lower is fractions of nanosecond.
+	 * The scaled_ppm represent the ratio in "parts per bilion" by which the
+	 * compensation value should be corrected.
+	 * To calculate new compenstation value we use 64bit fixed point
+	 * arithmetic on following formula
+	 * comp = tbase + tbase * scaled_ppm / (1M * 2^16)
+	 * where tbase is the basic compensation value calculated initialy
+	 * in cavium_ptp_init() -> tbase = 1/Hz. Then we use endian
+	 * independent structure definition to write data to PTP register.
+	 */
+	comp = ((u64)1000000000ull << 32) / clock->clock_rate;
+	adj = comp * scaled_ppm;
+	adj >>= 16;
+	adj = div_u64(adj, 1000000ull);
+	comp = neg_adj ? comp - adj : comp + adj;
+
+	spin_lock_irqsave(&clock->spin_lock, flags);
+	writeq(comp, clock->reg_base + PTP_CLOCK_COMP);
+	spin_unlock_irqrestore(&clock->spin_lock, flags);
+
+	return 0;
+}
+
+/**
+ * cavium_ptp_adjtime() - Adjust ptp time
+ * @ptp_info:   PTP clock info
+ * @delta: how much to adjust by, in nanosecs
+ */
+static int cavium_ptp_adjtime(struct ptp_clock_info *ptp_info, s64 delta)
+{
+	struct cavium_ptp *clock =
+		container_of(ptp_info, struct cavium_ptp, ptp_info);
+	unsigned long flags;
+
+	spin_lock_irqsave(&clock->spin_lock, flags);
+	timecounter_adjtime(&clock->time_counter, delta);
+	spin_unlock_irqrestore(&clock->spin_lock, flags);
+
+	/* Sync, for network driver to get latest value */
+	smp_mb();
+
+	return 0;
+}
+
+/**
+ * cavium_ptp_gettime() - Get hardware clock time with adjustment
+ * @ptp_info: PTP clock info
+ * @ts:  timespec
+ */
+static int cavium_ptp_gettime(struct ptp_clock_info *ptp_info,
+			      struct timespec64 *ts)
+{
+	struct cavium_ptp *clock =
+		container_of(ptp_info, struct cavium_ptp, ptp_info);
+	unsigned long flags;
+	u64 nsec;
+
+	spin_lock_irqsave(&clock->spin_lock, flags);
+	nsec = timecounter_read(&clock->time_counter);
+	spin_unlock_irqrestore(&clock->spin_lock, flags);
+
+	*ts = ns_to_timespec64(nsec);
+
+	return 0;
+}
+
+/**
+ * cavium_ptp_settime() - Set hardware clock time. Reset adjustment
+ * @ptp_info: PTP clock info
+ * @ts:  timespec
+ */
+static int cavium_ptp_settime(struct ptp_clock_info *ptp_info,
+			      const struct timespec64 *ts)
+{
+	struct cavium_ptp *clock =
+		container_of(ptp_info, struct cavium_ptp, ptp_info);
+	unsigned long flags;
+	u64 nsec;
+
+	nsec = timespec64_to_ns(ts);
+
+	spin_lock_irqsave(&clock->spin_lock, flags);
+	timecounter_init(&clock->time_counter, &clock->cycle_counter, nsec);
+	spin_unlock_irqrestore(&clock->spin_lock, flags);
+
+	return 0;
+}
+
+/**
+ * cavium_ptp_enable() - Request to enable or disable an ancillary feature.
+ * @ptp_info: PTP clock info
+ * @rq:  request
+ * @on:  is it on
+ */
+static int cavium_ptp_enable(struct ptp_clock_info *ptp_info,
+			     struct ptp_clock_request *rq, int on)
+{
+	return -EOPNOTSUPP;
+}
+
+static u64 cavium_ptp_cc_read(const struct cyclecounter *cc)
+{
+	struct cavium_ptp *clock =
+		container_of(cc, struct cavium_ptp, cycle_counter);
+
+	return readq(clock->reg_base + PTP_CLOCK_HI);
+}
+
+static int cavium_ptp_probe(struct pci_dev *pdev,
+			    const struct pci_device_id *ent)
+{
+	struct device *dev = &pdev->dev;
+	struct cavium_ptp *clock;
+	struct cyclecounter *cc;
+	u64 clock_cfg;
+	u64 clock_comp;
+	int err;
+
+	clock = devm_kzalloc(dev, sizeof(*clock), GFP_KERNEL);
+	if (!clock) {
+		err = -ENOMEM;
+		goto error;
+	}
+
+	clock->pdev = pdev;
+
+	err = pcim_enable_device(pdev);
+	if (err)
+		goto error_free;
+
+	err = pcim_iomap_regions(pdev, 1 << PCI_PTP_BAR_NO, pci_name(pdev));
+	if (err)
+		goto error_free;
+
+	clock->reg_base = pcim_iomap_table(pdev)[PCI_PTP_BAR_NO];
+
+	spin_lock_init(&clock->spin_lock);
+
+	cc = &clock->cycle_counter;
+	cc->read = cavium_ptp_cc_read;
+	cc->mask = CYCLECOUNTER_MASK(64);
+	cc->mult = 1;
+	cc->shift = 0;
+
+	timecounter_init(&clock->time_counter, &clock->cycle_counter,
+			 ktime_to_ns(ktime_get_real()));
+
+	clock->clock_rate = ptp_cavium_clock_get();
+
+	clock->ptp_info = (struct ptp_clock_info) {
+		.owner		= THIS_MODULE,
+		.name		= "ThunderX PTP",
+		.max_adj	= 1000000000ull,
+		.n_ext_ts	= 0,
+		.n_pins		= 0,
+		.pps		= 0,
+		.adjfine	= cavium_ptp_adjfine,
+		.adjtime	= cavium_ptp_adjtime,
+		.gettime64	= cavium_ptp_gettime,
+		.settime64	= cavium_ptp_settime,
+		.enable		= cavium_ptp_enable,
+	};
+
+	clock_cfg = readq(clock->reg_base + PTP_CLOCK_CFG);
+	clock_cfg |= PTP_CLOCK_CFG_PTP_EN;
+	writeq(clock_cfg, clock->reg_base + PTP_CLOCK_CFG);
+
+	clock_comp = ((u64)1000000000ull << 32) / clock->clock_rate;
+	writeq(clock_comp, clock->reg_base + PTP_CLOCK_COMP);
+
+	clock->ptp_clock = ptp_clock_register(&clock->ptp_info, dev);
+	if (IS_ERR(clock->ptp_clock)) {
+		err = PTR_ERR(clock->ptp_clock);
+		goto error_stop;
+	}
+
+	pci_set_drvdata(pdev, clock);
+	return 0;
+
+error_stop:
+	clock_cfg = readq(clock->reg_base + PTP_CLOCK_CFG);
+	clock_cfg &= ~PTP_CLOCK_CFG_PTP_EN;
+	writeq(clock_cfg, clock->reg_base + PTP_CLOCK_CFG);
+	pcim_iounmap_regions(pdev, 1 << PCI_PTP_BAR_NO);
+
+error_free:
+	devm_kfree(dev, clock);
+
+error:
+	/* For `cavium_ptp_get()` we need to differentiate between the case
+	 * when the core has not tried to probe this device and the case when
+	 * the probe failed.  In the later case we pretend that the
+	 * initialization was successful and keep the error in
+	 * `dev->driver_data`.
+	 */
+	pci_set_drvdata(pdev, ERR_PTR(err));
+	return 0;
+}
+
+static void cavium_ptp_remove(struct pci_dev *pdev)
+{
+	struct cavium_ptp *clock = pci_get_drvdata(pdev);
+	u64 clock_cfg;
+
+	if (IS_ERR_OR_NULL(clock))
+		return;
+
+	ptp_clock_unregister(clock->ptp_clock);
+
+	clock_cfg = readq(clock->reg_base + PTP_CLOCK_CFG);
+	clock_cfg &= ~PTP_CLOCK_CFG_PTP_EN;
+	writeq(clock_cfg, clock->reg_base + PTP_CLOCK_CFG);
+}
+
+static const struct pci_device_id cavium_ptp_id_table[] = {
+	{ PCI_DEVICE_SUB(PCI_VENDOR_ID_CAVIUM, PCI_DEVICE_ID_CAVIUM_PTP,
+			PCI_VENDOR_ID_CAVIUM, PCI_SUBSYS_DEVID_88XX_PTP) },
+	{ PCI_DEVICE_SUB(PCI_VENDOR_ID_CAVIUM, PCI_DEVICE_ID_CAVIUM_PTP,
+			PCI_VENDOR_ID_CAVIUM, PCI_SUBSYS_DEVID_81XX_PTP) },
+	{ PCI_DEVICE_SUB(PCI_VENDOR_ID_CAVIUM, PCI_DEVICE_ID_CAVIUM_PTP,
+			PCI_VENDOR_ID_CAVIUM, PCI_SUBSYS_DEVID_83XX_PTP) },
+	{ 0, }
+};
+
+static struct pci_driver cavium_ptp_driver = {
+	.name = DRV_NAME,
+	.id_table = cavium_ptp_id_table,
+	.probe = cavium_ptp_probe,
+	.remove = cavium_ptp_remove,
+};
+
+module_pci_driver(cavium_ptp_driver);
+
+MODULE_DESCRIPTION(DRV_NAME);
+MODULE_AUTHOR("Cavium Networks <support@cavium.com>");
+MODULE_LICENSE("GPL v2");
+MODULE_DEVICE_TABLE(pci, cavium_ptp_id_table);
diff --git a/drivers/net/ethernet/cavium/common/cavium_ptp.h b/drivers/net/ethernet/cavium/common/cavium_ptp.h
new file mode 100644
index 000000000..1e0ffe8f4
--- /dev/null
+++ b/drivers/net/ethernet/cavium/common/cavium_ptp.h
@@ -0,0 +1,70 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* cavium_ptp.h - PTP 1588 clock on Cavium hardware
+ * Copyright (c) 2003-2015, 2017 Cavium, Inc.
+ */
+
+#ifndef CAVIUM_PTP_H
+#define CAVIUM_PTP_H
+
+#include <linux/ptp_clock_kernel.h>
+#include <linux/timecounter.h>
+
+struct cavium_ptp {
+	struct pci_dev *pdev;
+
+	/* Serialize access to cycle_counter, time_counter and hw_registers */
+	spinlock_t spin_lock;
+	struct cyclecounter cycle_counter;
+	struct timecounter time_counter;
+	void __iomem *reg_base;
+
+	u32 clock_rate;
+
+	struct ptp_clock_info ptp_info;
+	struct ptp_clock *ptp_clock;
+};
+
+#if IS_REACHABLE(CONFIG_CAVIUM_PTP)
+
+struct cavium_ptp *cavium_ptp_get(void);
+void cavium_ptp_put(struct cavium_ptp *ptp);
+
+static inline u64 cavium_ptp_tstamp2time(struct cavium_ptp *ptp, u64 tstamp)
+{
+	unsigned long flags;
+	u64 ret;
+
+	spin_lock_irqsave(&ptp->spin_lock, flags);
+	ret = timecounter_cyc2time(&ptp->time_counter, tstamp);
+	spin_unlock_irqrestore(&ptp->spin_lock, flags);
+
+	return ret;
+}
+
+static inline int cavium_ptp_clock_index(struct cavium_ptp *clock)
+{
+	return ptp_clock_index(clock->ptp_clock);
+}
+
+#else
+
+static inline struct cavium_ptp *cavium_ptp_get(void)
+{
+	return ERR_PTR(-ENODEV);
+}
+
+static inline void cavium_ptp_put(struct cavium_ptp *ptp) {}
+
+static inline u64 cavium_ptp_tstamp2time(struct cavium_ptp *ptp, u64 tstamp)
+{
+	return 0;
+}
+
+static inline int cavium_ptp_clock_index(struct cavium_ptp *clock)
+{
+	return -1;
+}
+
+#endif
+
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/Makefile b/drivers/net/ethernet/cavium/liquidio/Makefile
new file mode 100644
index 000000000..bc9937502
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/Makefile
@@ -0,0 +1,24 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# Cavium Liquidio ethernet device driver
+#
+
+common-objs :=	lio_ethtool.o		\
+		lio_core.o		\
+		request_manager.o	\
+		response_manager.o	\
+		octeon_device.o		\
+		cn66xx_device.o		\
+		cn68xx_device.o		\
+		cn23xx_pf_device.o	\
+		cn23xx_vf_device.o	\
+		octeon_mailbox.o	\
+		octeon_mem_ops.o	\
+		octeon_droq.o		\
+		octeon_nic.o
+
+obj-$(CONFIG_LIQUIDIO) += liquidio.o
+liquidio-y := lio_main.o octeon_console.o lio_vf_rep.o $(common-objs)
+
+obj-$(CONFIG_LIQUIDIO_VF) += liquidio_vf.o
+liquidio_vf-y := lio_vf_main.o $(common-objs)
diff --git a/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c
new file mode 100644
index 000000000..9ed3d1ab2
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c
@@ -0,0 +1,1512 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/pci.h>
+#include <linux/vmalloc.h>
+#include <linux/etherdevice.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "cn23xx_pf_device.h"
+#include "octeon_main.h"
+#include "octeon_mailbox.h"
+
+#define RESET_NOTDONE 0
+#define RESET_DONE 1
+
+/* Change the value of SLI Packet Input Jabber Register to allow
+ * VXLAN TSO packets which can be 64424 bytes, exceeding the
+ * MAX_GSO_SIZE we supplied to the kernel
+ */
+#define CN23XX_INPUT_JABBER 64600
+
+void cn23xx_dump_pf_initialized_regs(struct octeon_device *oct)
+{
+	int i = 0;
+	u32 regval = 0;
+	struct octeon_cn23xx_pf *cn23xx = (struct octeon_cn23xx_pf *)oct->chip;
+
+	/*In cn23xx_soft_reset*/
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%llx\n",
+		"CN23XX_WIN_WR_MASK_REG", CVM_CAST64(CN23XX_WIN_WR_MASK_REG),
+		CVM_CAST64(octeon_read_csr64(oct, CN23XX_WIN_WR_MASK_REG)));
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%016llx\n",
+		"CN23XX_SLI_SCRATCH1", CVM_CAST64(CN23XX_SLI_SCRATCH1),
+		CVM_CAST64(octeon_read_csr64(oct, CN23XX_SLI_SCRATCH1)));
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%016llx\n",
+		"CN23XX_RST_SOFT_RST", CN23XX_RST_SOFT_RST,
+		lio_pci_readq(oct, CN23XX_RST_SOFT_RST));
+
+	/*In cn23xx_set_dpi_regs*/
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%016llx\n",
+		"CN23XX_DPI_DMA_CONTROL", CN23XX_DPI_DMA_CONTROL,
+		lio_pci_readq(oct, CN23XX_DPI_DMA_CONTROL));
+
+	for (i = 0; i < 6; i++) {
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_DPI_DMA_ENG_ENB", i,
+			CN23XX_DPI_DMA_ENG_ENB(i),
+			lio_pci_readq(oct, CN23XX_DPI_DMA_ENG_ENB(i)));
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_DPI_DMA_ENG_BUF", i,
+			CN23XX_DPI_DMA_ENG_BUF(i),
+			lio_pci_readq(oct, CN23XX_DPI_DMA_ENG_BUF(i)));
+	}
+
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%016llx\n", "CN23XX_DPI_CTL",
+		CN23XX_DPI_CTL, lio_pci_readq(oct, CN23XX_DPI_CTL));
+
+	/*In cn23xx_setup_pcie_mps and cn23xx_setup_pcie_mrrs */
+	pci_read_config_dword(oct->pci_dev, CN23XX_CONFIG_PCIE_DEVCTL, &regval);
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%016llx\n",
+		"CN23XX_CONFIG_PCIE_DEVCTL",
+		CVM_CAST64(CN23XX_CONFIG_PCIE_DEVCTL), CVM_CAST64(regval));
+
+	dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+		"CN23XX_DPI_SLI_PRTX_CFG", oct->pcie_port,
+		CN23XX_DPI_SLI_PRTX_CFG(oct->pcie_port),
+		lio_pci_readq(oct, CN23XX_DPI_SLI_PRTX_CFG(oct->pcie_port)));
+
+	/*In cn23xx_specific_regs_setup */
+	dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+		"CN23XX_SLI_S2M_PORTX_CTL", oct->pcie_port,
+		CVM_CAST64(CN23XX_SLI_S2M_PORTX_CTL(oct->pcie_port)),
+		CVM_CAST64(octeon_read_csr64(
+			oct, CN23XX_SLI_S2M_PORTX_CTL(oct->pcie_port))));
+
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%016llx\n",
+		"CN23XX_SLI_RING_RST", CVM_CAST64(CN23XX_SLI_PKT_IOQ_RING_RST),
+		(u64)octeon_read_csr64(oct, CN23XX_SLI_PKT_IOQ_RING_RST));
+
+	/*In cn23xx_setup_global_mac_regs*/
+	for (i = 0; i < CN23XX_MAX_MACS; i++) {
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_PKT_MAC_RINFO64", i,
+			CVM_CAST64(CN23XX_SLI_PKT_MAC_RINFO64(i, oct->pf_num)),
+			CVM_CAST64(octeon_read_csr64
+				(oct, CN23XX_SLI_PKT_MAC_RINFO64
+					(i, oct->pf_num))));
+	}
+
+	/*In cn23xx_setup_global_input_regs*/
+	for (i = 0; i < CN23XX_MAX_INPUT_QUEUES; i++) {
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_IQ_PKT_CONTROL64", i,
+			CVM_CAST64(CN23XX_SLI_IQ_PKT_CONTROL64(i)),
+			CVM_CAST64(octeon_read_csr64
+				(oct, CN23XX_SLI_IQ_PKT_CONTROL64(i))));
+	}
+
+	/*In cn23xx_setup_global_output_regs*/
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%016llx\n",
+		"CN23XX_SLI_OQ_WMARK", CVM_CAST64(CN23XX_SLI_OQ_WMARK),
+		CVM_CAST64(octeon_read_csr64(oct, CN23XX_SLI_OQ_WMARK)));
+
+	for (i = 0; i < CN23XX_MAX_OUTPUT_QUEUES; i++) {
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_OQ_PKT_CONTROL", i,
+			CVM_CAST64(CN23XX_SLI_OQ_PKT_CONTROL(i)),
+			CVM_CAST64(octeon_read_csr(
+				oct, CN23XX_SLI_OQ_PKT_CONTROL(i))));
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_OQ_PKT_INT_LEVELS", i,
+			CVM_CAST64(CN23XX_SLI_OQ_PKT_INT_LEVELS(i)),
+			CVM_CAST64(octeon_read_csr64(
+				oct, CN23XX_SLI_OQ_PKT_INT_LEVELS(i))));
+	}
+
+	/*In cn23xx_enable_interrupt and cn23xx_disable_interrupt*/
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%016llx\n",
+		"cn23xx->intr_enb_reg64",
+		CVM_CAST64((long)(cn23xx->intr_enb_reg64)),
+		CVM_CAST64(readq(cn23xx->intr_enb_reg64)));
+
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%016llx\n",
+		"cn23xx->intr_sum_reg64",
+		CVM_CAST64((long)(cn23xx->intr_sum_reg64)),
+		CVM_CAST64(readq(cn23xx->intr_sum_reg64)));
+
+	/*In cn23xx_setup_iq_regs*/
+	for (i = 0; i < CN23XX_MAX_INPUT_QUEUES; i++) {
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_IQ_BASE_ADDR64", i,
+			CVM_CAST64(CN23XX_SLI_IQ_BASE_ADDR64(i)),
+			CVM_CAST64(octeon_read_csr64(
+				oct, CN23XX_SLI_IQ_BASE_ADDR64(i))));
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_IQ_SIZE", i,
+			CVM_CAST64(CN23XX_SLI_IQ_SIZE(i)),
+			CVM_CAST64(octeon_read_csr
+				(oct, CN23XX_SLI_IQ_SIZE(i))));
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_IQ_DOORBELL", i,
+			CVM_CAST64(CN23XX_SLI_IQ_DOORBELL(i)),
+			CVM_CAST64(octeon_read_csr64(
+				oct, CN23XX_SLI_IQ_DOORBELL(i))));
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_IQ_INSTR_COUNT64", i,
+			CVM_CAST64(CN23XX_SLI_IQ_INSTR_COUNT64(i)),
+			CVM_CAST64(octeon_read_csr64(
+				oct, CN23XX_SLI_IQ_INSTR_COUNT64(i))));
+	}
+
+	/*In cn23xx_setup_oq_regs*/
+	for (i = 0; i < CN23XX_MAX_OUTPUT_QUEUES; i++) {
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_OQ_BASE_ADDR64", i,
+			CVM_CAST64(CN23XX_SLI_OQ_BASE_ADDR64(i)),
+			CVM_CAST64(octeon_read_csr64(
+				oct, CN23XX_SLI_OQ_BASE_ADDR64(i))));
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_OQ_SIZE", i,
+			CVM_CAST64(CN23XX_SLI_OQ_SIZE(i)),
+			CVM_CAST64(octeon_read_csr
+				(oct, CN23XX_SLI_OQ_SIZE(i))));
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_OQ_BUFF_INFO_SIZE", i,
+			CVM_CAST64(CN23XX_SLI_OQ_BUFF_INFO_SIZE(i)),
+			CVM_CAST64(octeon_read_csr(
+				oct, CN23XX_SLI_OQ_BUFF_INFO_SIZE(i))));
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_OQ_PKTS_SENT", i,
+			CVM_CAST64(CN23XX_SLI_OQ_PKTS_SENT(i)),
+			CVM_CAST64(octeon_read_csr64(
+				oct, CN23XX_SLI_OQ_PKTS_SENT(i))));
+		dev_dbg(&oct->pci_dev->dev, "%s(%d)[%llx] : 0x%016llx\n",
+			"CN23XX_SLI_OQ_PKTS_CREDIT", i,
+			CVM_CAST64(CN23XX_SLI_OQ_PKTS_CREDIT(i)),
+			CVM_CAST64(octeon_read_csr64(
+				oct, CN23XX_SLI_OQ_PKTS_CREDIT(i))));
+	}
+
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%016llx\n",
+		"CN23XX_SLI_PKT_TIME_INT",
+		CVM_CAST64(CN23XX_SLI_PKT_TIME_INT),
+		CVM_CAST64(octeon_read_csr64(oct, CN23XX_SLI_PKT_TIME_INT)));
+	dev_dbg(&oct->pci_dev->dev, "%s[%llx] : 0x%016llx\n",
+		"CN23XX_SLI_PKT_CNT_INT",
+		CVM_CAST64(CN23XX_SLI_PKT_CNT_INT),
+		CVM_CAST64(octeon_read_csr64(oct, CN23XX_SLI_PKT_CNT_INT)));
+}
+
+static int cn23xx_pf_soft_reset(struct octeon_device *oct)
+{
+	octeon_write_csr64(oct, CN23XX_WIN_WR_MASK_REG, 0xFF);
+
+	dev_dbg(&oct->pci_dev->dev, "OCTEON[%d]: BIST enabled for CN23XX soft reset\n",
+		oct->octeon_id);
+
+	octeon_write_csr64(oct, CN23XX_SLI_SCRATCH1, 0x1234ULL);
+
+	/* Initiate chip-wide soft reset */
+	lio_pci_readq(oct, CN23XX_RST_SOFT_RST);
+	lio_pci_writeq(oct, 1, CN23XX_RST_SOFT_RST);
+
+	/* Wait for 100ms as Octeon resets. */
+	mdelay(100);
+
+	if (octeon_read_csr64(oct, CN23XX_SLI_SCRATCH1)) {
+		dev_err(&oct->pci_dev->dev, "OCTEON[%d]: Soft reset failed\n",
+			oct->octeon_id);
+		return 1;
+	}
+
+	dev_dbg(&oct->pci_dev->dev, "OCTEON[%d]: Reset completed\n",
+		oct->octeon_id);
+
+	/* restore the  reset value*/
+	octeon_write_csr64(oct, CN23XX_WIN_WR_MASK_REG, 0xFF);
+
+	return 0;
+}
+
+static void cn23xx_enable_error_reporting(struct octeon_device *oct)
+{
+	u32 regval;
+	u32 uncorrectable_err_mask, corrtable_err_status;
+
+	pci_read_config_dword(oct->pci_dev, CN23XX_CONFIG_PCIE_DEVCTL, &regval);
+	if (regval & CN23XX_CONFIG_PCIE_DEVCTL_MASK) {
+		uncorrectable_err_mask = 0;
+		corrtable_err_status = 0;
+		pci_read_config_dword(oct->pci_dev,
+				      CN23XX_CONFIG_PCIE_UNCORRECT_ERR_MASK,
+				      &uncorrectable_err_mask);
+		pci_read_config_dword(oct->pci_dev,
+				      CN23XX_CONFIG_PCIE_CORRECT_ERR_STATUS,
+				      &corrtable_err_status);
+		dev_err(&oct->pci_dev->dev, "PCI-E Fatal error detected;\n"
+				 "\tdev_ctl_status_reg = 0x%08x\n"
+				 "\tuncorrectable_error_mask_reg = 0x%08x\n"
+				 "\tcorrectable_error_status_reg = 0x%08x\n",
+			    regval, uncorrectable_err_mask,
+			    corrtable_err_status);
+	}
+
+	regval |= 0xf; /* Enable Link error reporting */
+
+	dev_dbg(&oct->pci_dev->dev, "OCTEON[%d]: Enabling PCI-E error reporting..\n",
+		oct->octeon_id);
+	pci_write_config_dword(oct->pci_dev, CN23XX_CONFIG_PCIE_DEVCTL, regval);
+}
+
+static u32 cn23xx_coprocessor_clock(struct octeon_device *oct)
+{
+	/* Bits 29:24 of RST_BOOT[PNR_MUL] holds the ref.clock MULTIPLIER
+	 * for SLI.
+	 */
+
+	/* TBD: get the info in Hand-shake */
+	return (((lio_pci_readq(oct, CN23XX_RST_BOOT) >> 24) & 0x3f) * 50);
+}
+
+u32 cn23xx_pf_get_oq_ticks(struct octeon_device *oct, u32 time_intr_in_us)
+{
+	/* This gives the SLI clock per microsec */
+	u32 oqticks_per_us = cn23xx_coprocessor_clock(oct);
+
+	oct->pfvf_hsword.coproc_tics_per_us = oqticks_per_us;
+
+	/* This gives the clock cycles per millisecond */
+	oqticks_per_us *= 1000;
+
+	/* This gives the oq ticks (1024 core clock cycles) per millisecond */
+	oqticks_per_us /= 1024;
+
+	/* time_intr is in microseconds. The next 2 steps gives the oq ticks
+	 *  corressponding to time_intr.
+	 */
+	oqticks_per_us *= time_intr_in_us;
+	oqticks_per_us /= 1000;
+
+	return oqticks_per_us;
+}
+
+static void cn23xx_setup_global_mac_regs(struct octeon_device *oct)
+{
+	u16 mac_no = oct->pcie_port;
+	u16 pf_num = oct->pf_num;
+	u64 reg_val;
+	u64 temp;
+
+	/* programming SRN and TRS for each MAC(0..3)  */
+
+	dev_dbg(&oct->pci_dev->dev, "%s:Using pcie port %d\n",
+		__func__, mac_no);
+	/* By default, mapping all 64 IOQs to  a single MACs */
+
+	reg_val =
+	    octeon_read_csr64(oct, CN23XX_SLI_PKT_MAC_RINFO64(mac_no, pf_num));
+
+	if (oct->rev_id == OCTEON_CN23XX_REV_1_1) {
+		/* setting SRN <6:0>  */
+		reg_val = pf_num * CN23XX_MAX_RINGS_PER_PF_PASS_1_1;
+	} else {
+		/* setting SRN <6:0>  */
+		reg_val = pf_num * CN23XX_MAX_RINGS_PER_PF;
+	}
+
+	/* setting TRS <23:16> */
+	reg_val = reg_val |
+		  (oct->sriov_info.trs << CN23XX_PKT_MAC_CTL_RINFO_TRS_BIT_POS);
+	/* setting RPVF <39:32> */
+	temp = oct->sriov_info.rings_per_vf & 0xff;
+	reg_val |= (temp << CN23XX_PKT_MAC_CTL_RINFO_RPVF_BIT_POS);
+
+	/* setting NVFS <55:48> */
+	temp = oct->sriov_info.max_vfs & 0xff;
+	reg_val |= (temp << CN23XX_PKT_MAC_CTL_RINFO_NVFS_BIT_POS);
+
+	/* write these settings to MAC register */
+	octeon_write_csr64(oct, CN23XX_SLI_PKT_MAC_RINFO64(mac_no, pf_num),
+			   reg_val);
+
+	dev_dbg(&oct->pci_dev->dev, "SLI_PKT_MAC(%d)_PF(%d)_RINFO : 0x%016llx\n",
+		mac_no, pf_num, (u64)octeon_read_csr64
+		(oct, CN23XX_SLI_PKT_MAC_RINFO64(mac_no, pf_num)));
+}
+
+static int cn23xx_reset_io_queues(struct octeon_device *oct)
+{
+	int ret_val = 0;
+	u64 d64;
+	u32 q_no, srn, ern;
+	u32 loop = 1000;
+
+	srn = oct->sriov_info.pf_srn;
+	ern = srn + oct->sriov_info.num_pf_rings;
+
+	/*As per HRM reg description, s/w cant write 0 to ENB. */
+	/*to make the queue off, need to set the RST bit. */
+
+	/* Reset the Enable bit for all the 64 IQs.  */
+	for (q_no = srn; q_no < ern; q_no++) {
+		/* set RST bit to 1. This bit applies to both IQ and OQ */
+		d64 = octeon_read_csr64(oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no));
+		d64 = d64 | CN23XX_PKT_INPUT_CTL_RST;
+		octeon_write_csr64(oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no), d64);
+	}
+
+	/*wait until the RST bit is clear or the RST and quite bits are set*/
+	for (q_no = srn; q_no < ern; q_no++) {
+		u64 reg_val = octeon_read_csr64(oct,
+					CN23XX_SLI_IQ_PKT_CONTROL64(q_no));
+		while ((READ_ONCE(reg_val) & CN23XX_PKT_INPUT_CTL_RST) &&
+		       !(READ_ONCE(reg_val) & CN23XX_PKT_INPUT_CTL_QUIET) &&
+		       loop--) {
+			WRITE_ONCE(reg_val, octeon_read_csr64(
+			    oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no)));
+		}
+		if (!loop) {
+			dev_err(&oct->pci_dev->dev,
+				"clearing the reset reg failed or setting the quiet reg failed for qno: %u\n",
+				q_no);
+			return -1;
+		}
+		WRITE_ONCE(reg_val, READ_ONCE(reg_val) &
+			~CN23XX_PKT_INPUT_CTL_RST);
+		octeon_write_csr64(oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no),
+				   READ_ONCE(reg_val));
+
+		WRITE_ONCE(reg_val, octeon_read_csr64(
+			   oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no)));
+		if (READ_ONCE(reg_val) & CN23XX_PKT_INPUT_CTL_RST) {
+			dev_err(&oct->pci_dev->dev,
+				"clearing the reset failed for qno: %u\n",
+				q_no);
+			ret_val = -1;
+		}
+	}
+
+	return ret_val;
+}
+
+static int cn23xx_pf_setup_global_input_regs(struct octeon_device *oct)
+{
+	struct octeon_cn23xx_pf *cn23xx = (struct octeon_cn23xx_pf *)oct->chip;
+	struct octeon_instr_queue *iq;
+	u64 intr_threshold, reg_val;
+	u32 q_no, ern, srn;
+	u64 pf_num;
+	u64 vf_num;
+
+	pf_num = oct->pf_num;
+
+	srn = oct->sriov_info.pf_srn;
+	ern = srn + oct->sriov_info.num_pf_rings;
+
+	if (cn23xx_reset_io_queues(oct))
+		return -1;
+
+	/** Set the MAC_NUM and PVF_NUM in IQ_PKT_CONTROL reg
+	 * for all queues.Only PF can set these bits.
+	 * bits 29:30 indicate the MAC num.
+	 * bits 32:47 indicate the PVF num.
+	 */
+	for (q_no = 0; q_no < ern; q_no++) {
+		reg_val = (u64)oct->pcie_port << CN23XX_PKT_INPUT_CTL_MAC_NUM_POS;
+
+		/* for VF assigned queues. */
+		if (q_no < oct->sriov_info.pf_srn) {
+			vf_num = q_no / oct->sriov_info.rings_per_vf;
+			vf_num += 1; /* VF1, VF2,........ */
+		} else {
+			vf_num = 0;
+		}
+
+		reg_val |= vf_num << CN23XX_PKT_INPUT_CTL_VF_NUM_POS;
+		reg_val |= pf_num << CN23XX_PKT_INPUT_CTL_PF_NUM_POS;
+
+		octeon_write_csr64(oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no),
+				   reg_val);
+	}
+
+	/* Select ES, RO, NS, RDSIZE,DPTR Fomat#0 for
+	 * pf queues
+	 */
+	for (q_no = srn; q_no < ern; q_no++) {
+		void __iomem *inst_cnt_reg;
+
+		iq = oct->instr_queue[q_no];
+		if (iq)
+			inst_cnt_reg = iq->inst_cnt_reg;
+		else
+			inst_cnt_reg = (u8 *)oct->mmio[0].hw_addr +
+				       CN23XX_SLI_IQ_INSTR_COUNT64(q_no);
+
+		reg_val =
+		    octeon_read_csr64(oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no));
+
+		reg_val |= CN23XX_PKT_INPUT_CTL_MASK;
+
+		octeon_write_csr64(oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no),
+				   reg_val);
+
+		/* Set WMARK level for triggering PI_INT */
+		/* intr_threshold = CN23XX_DEF_IQ_INTR_THRESHOLD & */
+		intr_threshold = CFG_GET_IQ_INTR_PKT(cn23xx->conf) &
+				 CN23XX_PKT_IN_DONE_WMARK_MASK;
+
+		writeq((readq(inst_cnt_reg) &
+			~(CN23XX_PKT_IN_DONE_WMARK_MASK <<
+			  CN23XX_PKT_IN_DONE_WMARK_BIT_POS)) |
+		       (intr_threshold << CN23XX_PKT_IN_DONE_WMARK_BIT_POS),
+		       inst_cnt_reg);
+	}
+	return 0;
+}
+
+static void cn23xx_pf_setup_global_output_regs(struct octeon_device *oct)
+{
+	u32 reg_val;
+	u32 q_no, ern, srn;
+	u64 time_threshold;
+
+	struct octeon_cn23xx_pf *cn23xx = (struct octeon_cn23xx_pf *)oct->chip;
+
+	srn = oct->sriov_info.pf_srn;
+	ern = srn + oct->sriov_info.num_pf_rings;
+
+	if (CFG_GET_IS_SLI_BP_ON(cn23xx->conf)) {
+		octeon_write_csr64(oct, CN23XX_SLI_OQ_WMARK, 32);
+	} else {
+		/** Set Output queue watermark to 0 to disable backpressure */
+		octeon_write_csr64(oct, CN23XX_SLI_OQ_WMARK, 0);
+	}
+
+	for (q_no = srn; q_no < ern; q_no++) {
+		reg_val = octeon_read_csr(oct, CN23XX_SLI_OQ_PKT_CONTROL(q_no));
+
+		/* clear IPTR */
+		reg_val &= ~CN23XX_PKT_OUTPUT_CTL_IPTR;
+
+		/* set DPTR */
+		reg_val |= CN23XX_PKT_OUTPUT_CTL_DPTR;
+
+		/* reset BMODE */
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_BMODE);
+
+		/* No Relaxed Ordering, No Snoop, 64-bit Byte swap
+		 * for Output Queue ScatterList
+		 * reset ROR_P, NSR_P
+		 */
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_ROR_P);
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_NSR_P);
+
+#ifdef __LITTLE_ENDIAN_BITFIELD
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_ES_P);
+#else
+		reg_val |= (CN23XX_PKT_OUTPUT_CTL_ES_P);
+#endif
+		/* No Relaxed Ordering, No Snoop, 64-bit Byte swap
+		 * for Output Queue Data
+		 * reset ROR, NSR
+		 */
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_ROR);
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_NSR);
+		/* set the ES bit */
+		reg_val |= (CN23XX_PKT_OUTPUT_CTL_ES);
+
+		/* write all the selected settings */
+		octeon_write_csr(oct, CN23XX_SLI_OQ_PKT_CONTROL(q_no), reg_val);
+
+		/* Enabling these interrupt in oct->fn_list.enable_interrupt()
+		 * routine which called after IOQ init.
+		 * Set up interrupt packet and time thresholds
+		 * for all the OQs
+		 */
+		time_threshold = cn23xx_pf_get_oq_ticks(
+		    oct, (u32)CFG_GET_OQ_INTR_TIME(cn23xx->conf));
+
+		octeon_write_csr64(oct, CN23XX_SLI_OQ_PKT_INT_LEVELS(q_no),
+				   (CFG_GET_OQ_INTR_PKT(cn23xx->conf) |
+				    (time_threshold << 32)));
+	}
+
+	/** Setting the water mark level for pko back pressure **/
+	writeq(0x40, (u8 *)oct->mmio[0].hw_addr + CN23XX_SLI_OQ_WMARK);
+
+	/** Disabling setting OQs in reset when ring has no dorebells
+	 * enabling this will cause of head of line blocking
+	 */
+	/* Do it only for pass1.1. and pass1.2 */
+	if ((oct->rev_id == OCTEON_CN23XX_REV_1_0) ||
+	    (oct->rev_id == OCTEON_CN23XX_REV_1_1))
+		writeq(readq((u8 *)oct->mmio[0].hw_addr +
+				     CN23XX_SLI_GBL_CONTROL) | 0x2,
+		       (u8 *)oct->mmio[0].hw_addr + CN23XX_SLI_GBL_CONTROL);
+
+	/** Enable channel-level backpressure */
+	if (oct->pf_num)
+		writeq(0xffffffffffffffffULL,
+		       (u8 *)oct->mmio[0].hw_addr + CN23XX_SLI_OUT_BP_EN2_W1S);
+	else
+		writeq(0xffffffffffffffffULL,
+		       (u8 *)oct->mmio[0].hw_addr + CN23XX_SLI_OUT_BP_EN_W1S);
+}
+
+static int cn23xx_setup_pf_device_regs(struct octeon_device *oct)
+{
+	cn23xx_enable_error_reporting(oct);
+
+	/* program the MAC(0..3)_RINFO before setting up input/output regs */
+	cn23xx_setup_global_mac_regs(oct);
+
+	if (cn23xx_pf_setup_global_input_regs(oct))
+		return -1;
+
+	cn23xx_pf_setup_global_output_regs(oct);
+
+	/* Default error timeout value should be 0x200000 to avoid host hang
+	 * when reads invalid register
+	 */
+	octeon_write_csr64(oct, CN23XX_SLI_WINDOW_CTL,
+			   CN23XX_SLI_WINDOW_CTL_DEFAULT);
+
+	/* set SLI_PKT_IN_JABBER to handle large VXLAN packets */
+	octeon_write_csr64(oct, CN23XX_SLI_PKT_IN_JABBER, CN23XX_INPUT_JABBER);
+	return 0;
+}
+
+static void cn23xx_setup_iq_regs(struct octeon_device *oct, u32 iq_no)
+{
+	struct octeon_instr_queue *iq = oct->instr_queue[iq_no];
+	u64 pkt_in_done;
+
+	iq_no += oct->sriov_info.pf_srn;
+
+	/* Write the start of the input queue's ring and its size  */
+	octeon_write_csr64(oct, CN23XX_SLI_IQ_BASE_ADDR64(iq_no),
+			   iq->base_addr_dma);
+	octeon_write_csr(oct, CN23XX_SLI_IQ_SIZE(iq_no), iq->max_count);
+
+	/* Remember the doorbell & instruction count register addr
+	 * for this queue
+	 */
+	iq->doorbell_reg =
+	    (u8 *)oct->mmio[0].hw_addr + CN23XX_SLI_IQ_DOORBELL(iq_no);
+	iq->inst_cnt_reg =
+	    (u8 *)oct->mmio[0].hw_addr + CN23XX_SLI_IQ_INSTR_COUNT64(iq_no);
+	dev_dbg(&oct->pci_dev->dev, "InstQ[%d]:dbell reg @ 0x%p instcnt_reg @ 0x%p\n",
+		iq_no, iq->doorbell_reg, iq->inst_cnt_reg);
+
+	/* Store the current instruction counter (used in flush_iq
+	 * calculation)
+	 */
+	pkt_in_done = readq(iq->inst_cnt_reg);
+
+	if (oct->msix_on) {
+		/* Set CINT_ENB to enable IQ interrupt   */
+		writeq((pkt_in_done | CN23XX_INTR_CINT_ENB),
+		       iq->inst_cnt_reg);
+	} else {
+		/* Clear the count by writing back what we read, but don't
+		 * enable interrupts
+		 */
+		writeq(pkt_in_done, iq->inst_cnt_reg);
+	}
+
+	iq->reset_instr_cnt = 0;
+}
+
+static void cn23xx_setup_oq_regs(struct octeon_device *oct, u32 oq_no)
+{
+	u32 reg_val;
+	struct octeon_droq *droq = oct->droq[oq_no];
+	struct octeon_cn23xx_pf *cn23xx = (struct octeon_cn23xx_pf *)oct->chip;
+	u64 time_threshold;
+	u64 cnt_threshold;
+
+	oq_no += oct->sriov_info.pf_srn;
+
+	octeon_write_csr64(oct, CN23XX_SLI_OQ_BASE_ADDR64(oq_no),
+			   droq->desc_ring_dma);
+	octeon_write_csr(oct, CN23XX_SLI_OQ_SIZE(oq_no), droq->max_count);
+
+	octeon_write_csr(oct, CN23XX_SLI_OQ_BUFF_INFO_SIZE(oq_no),
+			 droq->buffer_size);
+
+	/* Get the mapped address of the pkt_sent and pkts_credit regs */
+	droq->pkts_sent_reg =
+	    (u8 *)oct->mmio[0].hw_addr + CN23XX_SLI_OQ_PKTS_SENT(oq_no);
+	droq->pkts_credit_reg =
+	    (u8 *)oct->mmio[0].hw_addr + CN23XX_SLI_OQ_PKTS_CREDIT(oq_no);
+
+	if (!oct->msix_on) {
+		/* Enable this output queue to generate Packet Timer Interrupt
+		 */
+		reg_val =
+		    octeon_read_csr(oct, CN23XX_SLI_OQ_PKT_CONTROL(oq_no));
+		reg_val |= CN23XX_PKT_OUTPUT_CTL_TENB;
+		octeon_write_csr(oct, CN23XX_SLI_OQ_PKT_CONTROL(oq_no),
+				 reg_val);
+
+		/* Enable this output queue to generate Packet Count Interrupt
+		 */
+		reg_val =
+		    octeon_read_csr(oct, CN23XX_SLI_OQ_PKT_CONTROL(oq_no));
+		reg_val |= CN23XX_PKT_OUTPUT_CTL_CENB;
+		octeon_write_csr(oct, CN23XX_SLI_OQ_PKT_CONTROL(oq_no),
+				 reg_val);
+	} else {
+		time_threshold = cn23xx_pf_get_oq_ticks(
+		    oct, (u32)CFG_GET_OQ_INTR_TIME(cn23xx->conf));
+		cnt_threshold = (u32)CFG_GET_OQ_INTR_PKT(cn23xx->conf);
+
+		octeon_write_csr64(
+		    oct, CN23XX_SLI_OQ_PKT_INT_LEVELS(oq_no),
+		    ((time_threshold << 32 | cnt_threshold)));
+	}
+}
+
+static void cn23xx_pf_mbox_thread(struct work_struct *work)
+{
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct octeon_mbox *mbox = (struct octeon_mbox *)wk->ctxptr;
+	struct octeon_device *oct = mbox->oct_dev;
+	u64 mbox_int_val, val64;
+	u32 q_no, i;
+
+	if (oct->rev_id < OCTEON_CN23XX_REV_1_1) {
+		/*read and clear by writing 1*/
+		mbox_int_val = readq(mbox->mbox_int_reg);
+		writeq(mbox_int_val, mbox->mbox_int_reg);
+
+		for (i = 0; i < oct->sriov_info.num_vfs_alloced; i++) {
+			q_no = i * oct->sriov_info.rings_per_vf;
+
+			val64 = readq(oct->mbox[q_no]->mbox_write_reg);
+
+			if (val64 && (val64 != OCTEON_PFVFACK)) {
+				if (octeon_mbox_read(oct->mbox[q_no]))
+					octeon_mbox_process_message(
+					    oct->mbox[q_no]);
+			}
+		}
+
+		schedule_delayed_work(&wk->work, msecs_to_jiffies(10));
+	} else {
+		octeon_mbox_process_message(mbox);
+	}
+}
+
+static int cn23xx_setup_pf_mbox(struct octeon_device *oct)
+{
+	struct octeon_mbox *mbox = NULL;
+	u16 mac_no = oct->pcie_port;
+	u16 pf_num = oct->pf_num;
+	u32 q_no, i;
+
+	if (!oct->sriov_info.max_vfs)
+		return 0;
+
+	for (i = 0; i < oct->sriov_info.max_vfs; i++) {
+		q_no = i * oct->sriov_info.rings_per_vf;
+
+		mbox = vmalloc(sizeof(*mbox));
+		if (!mbox)
+			goto free_mbox;
+
+		memset(mbox, 0, sizeof(struct octeon_mbox));
+
+		spin_lock_init(&mbox->lock);
+
+		mbox->oct_dev = oct;
+
+		mbox->q_no = q_no;
+
+		mbox->state = OCTEON_MBOX_STATE_IDLE;
+
+		/* PF mbox interrupt reg */
+		mbox->mbox_int_reg = (u8 *)oct->mmio[0].hw_addr +
+				     CN23XX_SLI_MAC_PF_MBOX_INT(mac_no, pf_num);
+
+		/* PF writes into SIG0 reg */
+		mbox->mbox_write_reg = (u8 *)oct->mmio[0].hw_addr +
+				       CN23XX_SLI_PKT_PF_VF_MBOX_SIG(q_no, 0);
+
+		/* PF reads from SIG1 reg */
+		mbox->mbox_read_reg = (u8 *)oct->mmio[0].hw_addr +
+				      CN23XX_SLI_PKT_PF_VF_MBOX_SIG(q_no, 1);
+
+		/*Mail Box Thread creation*/
+		INIT_DELAYED_WORK(&mbox->mbox_poll_wk.work,
+				  cn23xx_pf_mbox_thread);
+		mbox->mbox_poll_wk.ctxptr = (void *)mbox;
+
+		oct->mbox[q_no] = mbox;
+
+		writeq(OCTEON_PFVFSIG, mbox->mbox_read_reg);
+	}
+
+	if (oct->rev_id < OCTEON_CN23XX_REV_1_1)
+		schedule_delayed_work(&oct->mbox[0]->mbox_poll_wk.work,
+				      msecs_to_jiffies(0));
+
+	return 0;
+
+free_mbox:
+	while (i) {
+		i--;
+		vfree(oct->mbox[i]);
+	}
+
+	return 1;
+}
+
+static int cn23xx_free_pf_mbox(struct octeon_device *oct)
+{
+	u32 q_no, i;
+
+	if (!oct->sriov_info.max_vfs)
+		return 0;
+
+	for (i = 0; i < oct->sriov_info.max_vfs; i++) {
+		q_no = i * oct->sriov_info.rings_per_vf;
+		cancel_delayed_work_sync(
+		    &oct->mbox[q_no]->mbox_poll_wk.work);
+		vfree(oct->mbox[q_no]);
+	}
+
+	return 0;
+}
+
+static int cn23xx_enable_io_queues(struct octeon_device *oct)
+{
+	u64 reg_val;
+	u32 srn, ern, q_no;
+	u32 loop = 1000;
+
+	srn = oct->sriov_info.pf_srn;
+	ern = srn + oct->num_iqs;
+
+	for (q_no = srn; q_no < ern; q_no++) {
+		/* set the corresponding IQ IS_64B bit */
+		if (oct->io_qmask.iq64B & BIT_ULL(q_no - srn)) {
+			reg_val = octeon_read_csr64(
+			    oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no));
+			reg_val = reg_val | CN23XX_PKT_INPUT_CTL_IS_64B;
+			octeon_write_csr64(
+			    oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no), reg_val);
+		}
+
+		/* set the corresponding IQ ENB bit */
+		if (oct->io_qmask.iq & BIT_ULL(q_no - srn)) {
+			/* IOQs are in reset by default in PEM2 mode,
+			 * clearing reset bit
+			 */
+			reg_val = octeon_read_csr64(
+			    oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no));
+
+			if (reg_val & CN23XX_PKT_INPUT_CTL_RST) {
+				while ((reg_val & CN23XX_PKT_INPUT_CTL_RST) &&
+				       !(reg_val &
+					 CN23XX_PKT_INPUT_CTL_QUIET) &&
+				       --loop) {
+					reg_val = octeon_read_csr64(
+					    oct,
+					    CN23XX_SLI_IQ_PKT_CONTROL64(q_no));
+				}
+				if (!loop) {
+					dev_err(&oct->pci_dev->dev,
+						"clearing the reset reg failed or setting the quiet reg failed for qno: %u\n",
+						q_no);
+					return -1;
+				}
+				reg_val = reg_val & ~CN23XX_PKT_INPUT_CTL_RST;
+				octeon_write_csr64(
+				    oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no),
+				    reg_val);
+
+				reg_val = octeon_read_csr64(
+				    oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no));
+				if (reg_val & CN23XX_PKT_INPUT_CTL_RST) {
+					dev_err(&oct->pci_dev->dev,
+						"clearing the reset failed for qno: %u\n",
+						q_no);
+					return -1;
+				}
+			}
+			reg_val = octeon_read_csr64(
+			    oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no));
+			reg_val = reg_val | CN23XX_PKT_INPUT_CTL_RING_ENB;
+			octeon_write_csr64(
+			    oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no), reg_val);
+		}
+	}
+	for (q_no = srn; q_no < ern; q_no++) {
+		u32 reg_val;
+		/* set the corresponding OQ ENB bit */
+		if (oct->io_qmask.oq & BIT_ULL(q_no - srn)) {
+			reg_val = octeon_read_csr(
+			    oct, CN23XX_SLI_OQ_PKT_CONTROL(q_no));
+			reg_val = reg_val | CN23XX_PKT_OUTPUT_CTL_RING_ENB;
+			octeon_write_csr(oct, CN23XX_SLI_OQ_PKT_CONTROL(q_no),
+					 reg_val);
+		}
+	}
+	return 0;
+}
+
+static void cn23xx_disable_io_queues(struct octeon_device *oct)
+{
+	int q_no, loop;
+	u64 d64;
+	u32 d32;
+	u32 srn, ern;
+
+	srn = oct->sriov_info.pf_srn;
+	ern = srn + oct->num_iqs;
+
+	/*** Disable Input Queues. ***/
+	for (q_no = srn; q_no < ern; q_no++) {
+		loop = HZ;
+
+		/* start the Reset for a particular ring */
+		WRITE_ONCE(d64, octeon_read_csr64(
+			   oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no)));
+		WRITE_ONCE(d64, READ_ONCE(d64) &
+					(~(CN23XX_PKT_INPUT_CTL_RING_ENB)));
+		WRITE_ONCE(d64, READ_ONCE(d64) | CN23XX_PKT_INPUT_CTL_RST);
+		octeon_write_csr64(oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no),
+				   READ_ONCE(d64));
+
+		/* Wait until hardware indicates that the particular IQ
+		 * is out of reset.
+		 */
+		WRITE_ONCE(d64, octeon_read_csr64(
+					oct, CN23XX_SLI_PKT_IOQ_RING_RST));
+		while (!(READ_ONCE(d64) & BIT_ULL(q_no)) && loop--) {
+			WRITE_ONCE(d64, octeon_read_csr64(
+					oct, CN23XX_SLI_PKT_IOQ_RING_RST));
+			schedule_timeout_uninterruptible(1);
+		}
+
+		/* Reset the doorbell register for this Input Queue. */
+		octeon_write_csr(oct, CN23XX_SLI_IQ_DOORBELL(q_no), 0xFFFFFFFF);
+		while (octeon_read_csr64(oct, CN23XX_SLI_IQ_DOORBELL(q_no)) &&
+		       loop--) {
+			schedule_timeout_uninterruptible(1);
+		}
+	}
+
+	/*** Disable Output Queues. ***/
+	for (q_no = srn; q_no < ern; q_no++) {
+		loop = HZ;
+
+		/* Wait until hardware indicates that the particular IQ
+		 * is out of reset.It given that SLI_PKT_RING_RST is
+		 * common for both IQs and OQs
+		 */
+		WRITE_ONCE(d64, octeon_read_csr64(
+					oct, CN23XX_SLI_PKT_IOQ_RING_RST));
+		while (!(READ_ONCE(d64) & BIT_ULL(q_no)) && loop--) {
+			WRITE_ONCE(d64, octeon_read_csr64(
+					oct, CN23XX_SLI_PKT_IOQ_RING_RST));
+			schedule_timeout_uninterruptible(1);
+		}
+
+		/* Reset the doorbell register for this Output Queue. */
+		octeon_write_csr(oct, CN23XX_SLI_OQ_PKTS_CREDIT(q_no),
+				 0xFFFFFFFF);
+		while (octeon_read_csr64(oct,
+					 CN23XX_SLI_OQ_PKTS_CREDIT(q_no)) &&
+		       loop--) {
+			schedule_timeout_uninterruptible(1);
+		}
+
+		/* clear the SLI_PKT(0..63)_CNTS[CNT] reg value */
+		WRITE_ONCE(d32, octeon_read_csr(
+					oct, CN23XX_SLI_OQ_PKTS_SENT(q_no)));
+		octeon_write_csr(oct, CN23XX_SLI_OQ_PKTS_SENT(q_no),
+				 READ_ONCE(d32));
+	}
+}
+
+static u64 cn23xx_pf_msix_interrupt_handler(void *dev)
+{
+	struct octeon_ioq_vector *ioq_vector = (struct octeon_ioq_vector *)dev;
+	struct octeon_device *oct = ioq_vector->oct_dev;
+	u64 pkts_sent;
+	u64 ret = 0;
+	struct octeon_droq *droq = oct->droq[ioq_vector->droq_index];
+
+	dev_dbg(&oct->pci_dev->dev, "In %s octeon_dev @ %p\n", __func__, oct);
+
+	if (!droq) {
+		dev_err(&oct->pci_dev->dev, "23XX bringup FIXME: oct pfnum:%d ioq_vector->ioq_num :%d droq is NULL\n",
+			oct->pf_num, ioq_vector->ioq_num);
+		return 0;
+	}
+
+	pkts_sent = readq(droq->pkts_sent_reg);
+
+	/* If our device has interrupted, then proceed. Also check
+	 * for all f's if interrupt was triggered on an error
+	 * and the PCI read fails.
+	 */
+	if (!pkts_sent || (pkts_sent == 0xFFFFFFFFFFFFFFFFULL))
+		return ret;
+
+	/* Write count reg in sli_pkt_cnts to clear these int.*/
+	if ((pkts_sent & CN23XX_INTR_PO_INT) ||
+	    (pkts_sent & CN23XX_INTR_PI_INT)) {
+		if (pkts_sent & CN23XX_INTR_PO_INT)
+			ret |= MSIX_PO_INT;
+	}
+
+	if (pkts_sent & CN23XX_INTR_PI_INT)
+		/* We will clear the count when we update the read_index. */
+		ret |= MSIX_PI_INT;
+
+	/* Never need to handle msix mbox intr for pf. They arrive on the last
+	 * msix
+	 */
+	return ret;
+}
+
+static void cn23xx_handle_pf_mbox_intr(struct octeon_device *oct)
+{
+	struct delayed_work *work;
+	u64 mbox_int_val;
+	u32 i, q_no;
+
+	mbox_int_val = readq(oct->mbox[0]->mbox_int_reg);
+
+	for (i = 0; i < oct->sriov_info.num_vfs_alloced; i++) {
+		q_no = i * oct->sriov_info.rings_per_vf;
+
+		if (mbox_int_val & BIT_ULL(q_no)) {
+			writeq(BIT_ULL(q_no),
+			       oct->mbox[0]->mbox_int_reg);
+			if (octeon_mbox_read(oct->mbox[q_no])) {
+				work = &oct->mbox[q_no]->mbox_poll_wk.work;
+				schedule_delayed_work(work,
+						      msecs_to_jiffies(0));
+			}
+		}
+	}
+}
+
+static irqreturn_t cn23xx_interrupt_handler(void *dev)
+{
+	struct octeon_device *oct = (struct octeon_device *)dev;
+	struct octeon_cn23xx_pf *cn23xx = (struct octeon_cn23xx_pf *)oct->chip;
+	u64 intr64;
+
+	dev_dbg(&oct->pci_dev->dev, "In %s octeon_dev @ %p\n", __func__, oct);
+	intr64 = readq(cn23xx->intr_sum_reg64);
+
+	oct->int_status = 0;
+
+	if (intr64 & CN23XX_INTR_ERR)
+		dev_err(&oct->pci_dev->dev, "OCTEON[%d]: Error Intr: 0x%016llx\n",
+			oct->octeon_id, CVM_CAST64(intr64));
+
+	/* When VFs write into MBOX_SIG2 reg,these intr is set in PF */
+	if (intr64 & CN23XX_INTR_VF_MBOX)
+		cn23xx_handle_pf_mbox_intr(oct);
+
+	if (oct->msix_on != LIO_FLAG_MSIX_ENABLED) {
+		if (intr64 & CN23XX_INTR_PKT_DATA)
+			oct->int_status |= OCT_DEV_INTR_PKT_DATA;
+	}
+
+	if (intr64 & (CN23XX_INTR_DMA0_FORCE))
+		oct->int_status |= OCT_DEV_INTR_DMA0_FORCE;
+	if (intr64 & (CN23XX_INTR_DMA1_FORCE))
+		oct->int_status |= OCT_DEV_INTR_DMA1_FORCE;
+
+	/* Clear the current interrupts */
+	writeq(intr64, cn23xx->intr_sum_reg64);
+
+	return IRQ_HANDLED;
+}
+
+static void cn23xx_bar1_idx_setup(struct octeon_device *oct, u64 core_addr,
+				  u32 idx, int valid)
+{
+	u64 bar1;
+	u64 reg_adr;
+
+	if (!valid) {
+		reg_adr = lio_pci_readq(
+			oct, CN23XX_PEM_BAR1_INDEX_REG(oct->pcie_port, idx));
+		WRITE_ONCE(bar1, reg_adr);
+		lio_pci_writeq(oct, (READ_ONCE(bar1) & 0xFFFFFFFEULL),
+			       CN23XX_PEM_BAR1_INDEX_REG(oct->pcie_port, idx));
+		reg_adr = lio_pci_readq(
+			oct, CN23XX_PEM_BAR1_INDEX_REG(oct->pcie_port, idx));
+		WRITE_ONCE(bar1, reg_adr);
+		return;
+	}
+
+	/*  The PEM(0..3)_BAR1_INDEX(0..15)[ADDR_IDX]<23:4> stores
+	 *  bits <41:22> of the Core Addr
+	 */
+	lio_pci_writeq(oct, (((core_addr >> 22) << 4) | PCI_BAR1_MASK),
+		       CN23XX_PEM_BAR1_INDEX_REG(oct->pcie_port, idx));
+
+	WRITE_ONCE(bar1, lio_pci_readq(
+		   oct, CN23XX_PEM_BAR1_INDEX_REG(oct->pcie_port, idx)));
+}
+
+static void cn23xx_bar1_idx_write(struct octeon_device *oct, u32 idx, u32 mask)
+{
+	lio_pci_writeq(oct, mask,
+		       CN23XX_PEM_BAR1_INDEX_REG(oct->pcie_port, idx));
+}
+
+static u32 cn23xx_bar1_idx_read(struct octeon_device *oct, u32 idx)
+{
+	return (u32)lio_pci_readq(
+	    oct, CN23XX_PEM_BAR1_INDEX_REG(oct->pcie_port, idx));
+}
+
+/* always call with lock held */
+static u32 cn23xx_update_read_index(struct octeon_instr_queue *iq)
+{
+	u32 new_idx;
+	u32 last_done;
+	u32 pkt_in_done = readl(iq->inst_cnt_reg);
+
+	last_done = pkt_in_done - iq->pkt_in_done;
+	iq->pkt_in_done = pkt_in_done;
+
+	/* Modulo of the new index with the IQ size will give us
+	 * the new index.  The iq->reset_instr_cnt is always zero for
+	 * cn23xx, so no extra adjustments are needed.
+	 */
+	new_idx = (iq->octeon_read_index +
+		   (u32)(last_done & CN23XX_PKT_IN_DONE_CNT_MASK)) %
+		  iq->max_count;
+
+	return new_idx;
+}
+
+static void cn23xx_enable_pf_interrupt(struct octeon_device *oct, u8 intr_flag)
+{
+	struct octeon_cn23xx_pf *cn23xx = (struct octeon_cn23xx_pf *)oct->chip;
+	u64 intr_val = 0;
+
+	/*  Divide the single write to multiple writes based on the flag. */
+	/* Enable Interrupt */
+	if (intr_flag == OCTEON_ALL_INTR) {
+		writeq(cn23xx->intr_mask64, cn23xx->intr_enb_reg64);
+	} else if (intr_flag & OCTEON_OUTPUT_INTR) {
+		intr_val = readq(cn23xx->intr_enb_reg64);
+		intr_val |= CN23XX_INTR_PKT_DATA;
+		writeq(intr_val, cn23xx->intr_enb_reg64);
+	} else if ((intr_flag & OCTEON_MBOX_INTR) &&
+		   (oct->sriov_info.max_vfs > 0)) {
+		if (oct->rev_id >= OCTEON_CN23XX_REV_1_1) {
+			intr_val = readq(cn23xx->intr_enb_reg64);
+			intr_val |= CN23XX_INTR_VF_MBOX;
+			writeq(intr_val, cn23xx->intr_enb_reg64);
+		}
+	}
+}
+
+static void cn23xx_disable_pf_interrupt(struct octeon_device *oct, u8 intr_flag)
+{
+	struct octeon_cn23xx_pf *cn23xx = (struct octeon_cn23xx_pf *)oct->chip;
+	u64 intr_val = 0;
+
+	/* Disable Interrupts */
+	if (intr_flag == OCTEON_ALL_INTR) {
+		writeq(0, cn23xx->intr_enb_reg64);
+	} else if (intr_flag & OCTEON_OUTPUT_INTR) {
+		intr_val = readq(cn23xx->intr_enb_reg64);
+		intr_val &= ~CN23XX_INTR_PKT_DATA;
+		writeq(intr_val, cn23xx->intr_enb_reg64);
+	} else if ((intr_flag & OCTEON_MBOX_INTR) &&
+		   (oct->sriov_info.max_vfs > 0)) {
+		if (oct->rev_id >= OCTEON_CN23XX_REV_1_1) {
+			intr_val = readq(cn23xx->intr_enb_reg64);
+			intr_val &= ~CN23XX_INTR_VF_MBOX;
+			writeq(intr_val, cn23xx->intr_enb_reg64);
+		}
+	}
+}
+
+static void cn23xx_get_pcie_qlmport(struct octeon_device *oct)
+{
+	oct->pcie_port = (octeon_read_csr(oct, CN23XX_SLI_MAC_NUMBER)) & 0xff;
+
+	dev_dbg(&oct->pci_dev->dev, "OCTEON: CN23xx uses PCIE Port %d\n",
+		oct->pcie_port);
+}
+
+static int cn23xx_get_pf_num(struct octeon_device *oct)
+{
+	u32 fdl_bit = 0;
+	u64 pkt0_in_ctl, d64;
+	int pfnum, mac, trs, ret;
+
+	ret = 0;
+
+	/** Read Function Dependency Link reg to get the function number */
+	if (pci_read_config_dword(oct->pci_dev, CN23XX_PCIE_SRIOV_FDL,
+				  &fdl_bit) == 0) {
+		oct->pf_num = ((fdl_bit >> CN23XX_PCIE_SRIOV_FDL_BIT_POS) &
+			       CN23XX_PCIE_SRIOV_FDL_MASK);
+	} else {
+		ret = -EINVAL;
+
+		/* Under some virtual environments, extended PCI regs are
+		 * inaccessible, in which case the above read will have failed.
+		 * In this case, read the PF number from the
+		 * SLI_PKT0_INPUT_CONTROL reg (written by f/w)
+		 */
+		pkt0_in_ctl = octeon_read_csr64(oct,
+						CN23XX_SLI_IQ_PKT_CONTROL64(0));
+		pfnum = (pkt0_in_ctl >> CN23XX_PKT_INPUT_CTL_PF_NUM_POS) &
+			CN23XX_PKT_INPUT_CTL_PF_NUM_MASK;
+		mac = (octeon_read_csr(oct, CN23XX_SLI_MAC_NUMBER)) & 0xff;
+
+		/* validate PF num by reading RINFO; f/w writes RINFO.trs == 1*/
+		d64 = octeon_read_csr64(oct,
+					CN23XX_SLI_PKT_MAC_RINFO64(mac, pfnum));
+		trs = (int)(d64 >> CN23XX_PKT_MAC_CTL_RINFO_TRS_BIT_POS) & 0xff;
+		if (trs == 1) {
+			dev_err(&oct->pci_dev->dev,
+				"OCTEON: error reading PCI cfg space pfnum, re-read %u\n",
+				pfnum);
+			oct->pf_num = pfnum;
+			ret = 0;
+		} else {
+			dev_err(&oct->pci_dev->dev,
+				"OCTEON: error reading PCI cfg space pfnum; could not ascertain PF number\n");
+		}
+	}
+
+	return ret;
+}
+
+static void cn23xx_setup_reg_address(struct octeon_device *oct)
+{
+	u8 __iomem *bar0_pciaddr = oct->mmio[0].hw_addr;
+	struct octeon_cn23xx_pf *cn23xx = (struct octeon_cn23xx_pf *)oct->chip;
+
+	oct->reg_list.pci_win_wr_addr_hi =
+	    (u32 __iomem *)(bar0_pciaddr + CN23XX_WIN_WR_ADDR_HI);
+	oct->reg_list.pci_win_wr_addr_lo =
+	    (u32 __iomem *)(bar0_pciaddr + CN23XX_WIN_WR_ADDR_LO);
+	oct->reg_list.pci_win_wr_addr =
+	    (u64 __iomem *)(bar0_pciaddr + CN23XX_WIN_WR_ADDR64);
+
+	oct->reg_list.pci_win_rd_addr_hi =
+	    (u32 __iomem *)(bar0_pciaddr + CN23XX_WIN_RD_ADDR_HI);
+	oct->reg_list.pci_win_rd_addr_lo =
+	    (u32 __iomem *)(bar0_pciaddr + CN23XX_WIN_RD_ADDR_LO);
+	oct->reg_list.pci_win_rd_addr =
+	    (u64 __iomem *)(bar0_pciaddr + CN23XX_WIN_RD_ADDR64);
+
+	oct->reg_list.pci_win_wr_data_hi =
+	    (u32 __iomem *)(bar0_pciaddr + CN23XX_WIN_WR_DATA_HI);
+	oct->reg_list.pci_win_wr_data_lo =
+	    (u32 __iomem *)(bar0_pciaddr + CN23XX_WIN_WR_DATA_LO);
+	oct->reg_list.pci_win_wr_data =
+	    (u64 __iomem *)(bar0_pciaddr + CN23XX_WIN_WR_DATA64);
+
+	oct->reg_list.pci_win_rd_data_hi =
+	    (u32 __iomem *)(bar0_pciaddr + CN23XX_WIN_RD_DATA_HI);
+	oct->reg_list.pci_win_rd_data_lo =
+	    (u32 __iomem *)(bar0_pciaddr + CN23XX_WIN_RD_DATA_LO);
+	oct->reg_list.pci_win_rd_data =
+	    (u64 __iomem *)(bar0_pciaddr + CN23XX_WIN_RD_DATA64);
+
+	cn23xx_get_pcie_qlmport(oct);
+
+	cn23xx->intr_mask64 = CN23XX_INTR_MASK;
+	if (!oct->msix_on)
+		cn23xx->intr_mask64 |= CN23XX_INTR_PKT_TIME;
+	if (oct->rev_id >= OCTEON_CN23XX_REV_1_1)
+		cn23xx->intr_mask64 |= CN23XX_INTR_VF_MBOX;
+
+	cn23xx->intr_sum_reg64 =
+	    bar0_pciaddr +
+	    CN23XX_SLI_MAC_PF_INT_SUM64(oct->pcie_port, oct->pf_num);
+	cn23xx->intr_enb_reg64 =
+	    bar0_pciaddr +
+	    CN23XX_SLI_MAC_PF_INT_ENB64(oct->pcie_port, oct->pf_num);
+}
+
+int cn23xx_sriov_config(struct octeon_device *oct)
+{
+	struct octeon_cn23xx_pf *cn23xx = (struct octeon_cn23xx_pf *)oct->chip;
+	u32 max_rings, total_rings, max_vfs, rings_per_vf;
+	u32 pf_srn, num_pf_rings;
+	u32 max_possible_vfs;
+
+	cn23xx->conf =
+		(struct octeon_config *)oct_get_config_info(oct, LIO_23XX);
+	switch (oct->rev_id) {
+	case OCTEON_CN23XX_REV_1_0:
+		max_rings = CN23XX_MAX_RINGS_PER_PF_PASS_1_0;
+		max_possible_vfs = CN23XX_MAX_VFS_PER_PF_PASS_1_0;
+		break;
+	case OCTEON_CN23XX_REV_1_1:
+		max_rings = CN23XX_MAX_RINGS_PER_PF_PASS_1_1;
+		max_possible_vfs = CN23XX_MAX_VFS_PER_PF_PASS_1_1;
+		break;
+	default:
+		max_rings = CN23XX_MAX_RINGS_PER_PF;
+		max_possible_vfs = CN23XX_MAX_VFS_PER_PF;
+		break;
+	}
+
+	if (oct->sriov_info.num_pf_rings)
+		num_pf_rings = oct->sriov_info.num_pf_rings;
+	else
+		num_pf_rings = num_present_cpus();
+
+#ifdef CONFIG_PCI_IOV
+	max_vfs = min_t(u32,
+			(max_rings - num_pf_rings), max_possible_vfs);
+	rings_per_vf = 1;
+#else
+	max_vfs = 0;
+	rings_per_vf = 0;
+#endif
+
+	total_rings = num_pf_rings + max_vfs;
+
+	/* the first ring of the pf */
+	pf_srn = total_rings - num_pf_rings;
+
+	oct->sriov_info.trs = total_rings;
+	oct->sriov_info.max_vfs = max_vfs;
+	oct->sriov_info.rings_per_vf = rings_per_vf;
+	oct->sriov_info.pf_srn = pf_srn;
+	oct->sriov_info.num_pf_rings = num_pf_rings;
+	dev_notice(&oct->pci_dev->dev, "trs:%d max_vfs:%d rings_per_vf:%d pf_srn:%d num_pf_rings:%d\n",
+		   oct->sriov_info.trs, oct->sriov_info.max_vfs,
+		   oct->sriov_info.rings_per_vf, oct->sriov_info.pf_srn,
+		   oct->sriov_info.num_pf_rings);
+
+	oct->sriov_info.sriov_enabled = 0;
+
+	return 0;
+}
+
+int setup_cn23xx_octeon_pf_device(struct octeon_device *oct)
+{
+	u32 data32;
+	u64 BAR0, BAR1;
+
+	pci_read_config_dword(oct->pci_dev, PCI_BASE_ADDRESS_0, &data32);
+	BAR0 = (u64)(data32 & ~0xf);
+	pci_read_config_dword(oct->pci_dev, PCI_BASE_ADDRESS_1, &data32);
+	BAR0 |= ((u64)data32 << 32);
+	pci_read_config_dword(oct->pci_dev, PCI_BASE_ADDRESS_2, &data32);
+	BAR1 = (u64)(data32 & ~0xf);
+	pci_read_config_dword(oct->pci_dev, PCI_BASE_ADDRESS_3, &data32);
+	BAR1 |= ((u64)data32 << 32);
+
+	if (!BAR0 || !BAR1) {
+		if (!BAR0)
+			dev_err(&oct->pci_dev->dev, "device BAR0 unassigned\n");
+		if (!BAR1)
+			dev_err(&oct->pci_dev->dev, "device BAR1 unassigned\n");
+		return 1;
+	}
+
+	if (octeon_map_pci_barx(oct, 0, 0))
+		return 1;
+
+	if (octeon_map_pci_barx(oct, 1, MAX_BAR1_IOREMAP_SIZE)) {
+		dev_err(&oct->pci_dev->dev, "%s CN23XX BAR1 map failed\n",
+			__func__);
+		octeon_unmap_pci_barx(oct, 0);
+		return 1;
+	}
+
+	if (cn23xx_get_pf_num(oct) != 0)
+		return 1;
+
+	if (cn23xx_sriov_config(oct)) {
+		octeon_unmap_pci_barx(oct, 0);
+		octeon_unmap_pci_barx(oct, 1);
+		return 1;
+	}
+
+	octeon_write_csr64(oct, CN23XX_SLI_MAC_CREDIT_CNT, 0x3F802080802080ULL);
+
+	oct->fn_list.setup_iq_regs = cn23xx_setup_iq_regs;
+	oct->fn_list.setup_oq_regs = cn23xx_setup_oq_regs;
+	oct->fn_list.setup_mbox = cn23xx_setup_pf_mbox;
+	oct->fn_list.free_mbox = cn23xx_free_pf_mbox;
+
+	oct->fn_list.process_interrupt_regs = cn23xx_interrupt_handler;
+	oct->fn_list.msix_interrupt_handler = cn23xx_pf_msix_interrupt_handler;
+
+	oct->fn_list.soft_reset = cn23xx_pf_soft_reset;
+	oct->fn_list.setup_device_regs = cn23xx_setup_pf_device_regs;
+	oct->fn_list.update_iq_read_idx = cn23xx_update_read_index;
+
+	oct->fn_list.bar1_idx_setup = cn23xx_bar1_idx_setup;
+	oct->fn_list.bar1_idx_write = cn23xx_bar1_idx_write;
+	oct->fn_list.bar1_idx_read = cn23xx_bar1_idx_read;
+
+	oct->fn_list.enable_interrupt = cn23xx_enable_pf_interrupt;
+	oct->fn_list.disable_interrupt = cn23xx_disable_pf_interrupt;
+
+	oct->fn_list.enable_io_queues = cn23xx_enable_io_queues;
+	oct->fn_list.disable_io_queues = cn23xx_disable_io_queues;
+
+	cn23xx_setup_reg_address(oct);
+
+	oct->coproc_clock_rate = 1000000ULL * cn23xx_coprocessor_clock(oct);
+
+	return 0;
+}
+
+int validate_cn23xx_pf_config_info(struct octeon_device *oct,
+				   struct octeon_config *conf23xx)
+{
+	if (CFG_GET_IQ_MAX_Q(conf23xx) > CN23XX_MAX_INPUT_QUEUES) {
+		dev_err(&oct->pci_dev->dev, "%s: Num IQ (%d) exceeds Max (%d)\n",
+			__func__, CFG_GET_IQ_MAX_Q(conf23xx),
+			CN23XX_MAX_INPUT_QUEUES);
+		return 1;
+	}
+
+	if (CFG_GET_OQ_MAX_Q(conf23xx) > CN23XX_MAX_OUTPUT_QUEUES) {
+		dev_err(&oct->pci_dev->dev, "%s: Num OQ (%d) exceeds Max (%d)\n",
+			__func__, CFG_GET_OQ_MAX_Q(conf23xx),
+			CN23XX_MAX_OUTPUT_QUEUES);
+		return 1;
+	}
+
+	if (CFG_GET_IQ_INSTR_TYPE(conf23xx) != OCTEON_32BYTE_INSTR &&
+	    CFG_GET_IQ_INSTR_TYPE(conf23xx) != OCTEON_64BYTE_INSTR) {
+		dev_err(&oct->pci_dev->dev, "%s: Invalid instr type for IQ\n",
+			__func__);
+		return 1;
+	}
+
+	if (!CFG_GET_OQ_REFILL_THRESHOLD(conf23xx)) {
+		dev_err(&oct->pci_dev->dev, "%s: Invalid parameter for OQ\n",
+			__func__);
+		return 1;
+	}
+
+	if (!(CFG_GET_OQ_INTR_TIME(conf23xx))) {
+		dev_err(&oct->pci_dev->dev, "%s: Invalid parameter for OQ\n",
+			__func__);
+		return 1;
+	}
+
+	return 0;
+}
+
+int cn23xx_fw_loaded(struct octeon_device *oct)
+{
+	u64 val;
+
+	/* If there's more than one active PF on this NIC, then that
+	 * implies that the NIC firmware is loaded and running.  This check
+	 * prevents a rare false negative that might occur if we only relied
+	 * on checking the SCR2_BIT_FW_LOADED flag.  The false negative would
+	 * happen if the PF driver sees SCR2_BIT_FW_LOADED as cleared even
+	 * though the firmware was already loaded but still booting and has yet
+	 * to set SCR2_BIT_FW_LOADED.
+	 */
+	if (atomic_read(oct->adapter_refcount) > 1)
+		return 1;
+
+	val = octeon_read_csr64(oct, CN23XX_SLI_SCRATCH2);
+	return (val >> SCR2_BIT_FW_LOADED) & 1ULL;
+}
+
+void cn23xx_tell_vf_its_macaddr_changed(struct octeon_device *oct, int vfidx,
+					u8 *mac)
+{
+	if (oct->sriov_info.vf_drv_loaded_mask & BIT_ULL(vfidx)) {
+		struct octeon_mbox_cmd mbox_cmd;
+
+		mbox_cmd.msg.u64 = 0;
+		mbox_cmd.msg.s.type = OCTEON_MBOX_REQUEST;
+		mbox_cmd.msg.s.resp_needed = 0;
+		mbox_cmd.msg.s.cmd = OCTEON_PF_CHANGED_VF_MACADDR;
+		mbox_cmd.msg.s.len = 1;
+		mbox_cmd.recv_len = 0;
+		mbox_cmd.recv_status = 0;
+		mbox_cmd.fn = NULL;
+		mbox_cmd.fn_arg = NULL;
+		ether_addr_copy(mbox_cmd.msg.s.params, mac);
+		mbox_cmd.q_no = vfidx * oct->sriov_info.rings_per_vf;
+		octeon_mbox_write(oct, &mbox_cmd);
+	}
+}
+
+static void
+cn23xx_get_vf_stats_callback(struct octeon_device *oct,
+			     struct octeon_mbox_cmd *cmd, void *arg)
+{
+	struct oct_vf_stats_ctx *ctx = arg;
+
+	memcpy(ctx->stats, cmd->data, sizeof(struct oct_vf_stats));
+	atomic_set(&ctx->status, 1);
+}
+
+int cn23xx_get_vf_stats(struct octeon_device *oct, int vfidx,
+			struct oct_vf_stats *stats)
+{
+	u32 timeout = HZ; // 1sec
+	struct octeon_mbox_cmd mbox_cmd;
+	struct oct_vf_stats_ctx ctx;
+	u32 count = 0, ret;
+
+	if (!(oct->sriov_info.vf_drv_loaded_mask & (1ULL << vfidx)))
+		return -1;
+
+	if (sizeof(struct oct_vf_stats) > sizeof(mbox_cmd.data))
+		return -1;
+
+	mbox_cmd.msg.u64 = 0;
+	mbox_cmd.msg.s.type = OCTEON_MBOX_REQUEST;
+	mbox_cmd.msg.s.resp_needed = 1;
+	mbox_cmd.msg.s.cmd = OCTEON_GET_VF_STATS;
+	mbox_cmd.msg.s.len = 1;
+	mbox_cmd.q_no = vfidx * oct->sriov_info.rings_per_vf;
+	mbox_cmd.recv_len = 0;
+	mbox_cmd.recv_status = 0;
+	mbox_cmd.fn = (octeon_mbox_callback_t)cn23xx_get_vf_stats_callback;
+	ctx.stats = stats;
+	atomic_set(&ctx.status, 0);
+	mbox_cmd.fn_arg = (void *)&ctx;
+	memset(mbox_cmd.data, 0, sizeof(mbox_cmd.data));
+	octeon_mbox_write(oct, &mbox_cmd);
+
+	do {
+		schedule_timeout_uninterruptible(1);
+	} while ((atomic_read(&ctx.status) == 0) && (count++ < timeout));
+
+	ret = atomic_read(&ctx.status);
+	if (ret == 0) {
+		octeon_mbox_cancel(oct, 0);
+		dev_err(&oct->pci_dev->dev, "Unable to get stats from VF-%d, timedout\n",
+			vfidx);
+		return -1;
+	}
+
+	return 0;
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.h b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.h
new file mode 100644
index 000000000..e6f31d0d5
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.h
@@ -0,0 +1,73 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*! \file  cn23xx_device.h
+ * \brief Host Driver: Routines that perform CN23XX specific operations.
+ */
+
+#ifndef __CN23XX_PF_DEVICE_H__
+#define __CN23XX_PF_DEVICE_H__
+
+#include "cn23xx_pf_regs.h"
+
+/* Register address and configuration for a CN23XX devices.
+ * If device specific changes need to be made then add a struct to include
+ * device specific fields as shown in the commented section
+ */
+struct octeon_cn23xx_pf {
+	/** PCI interrupt summary register */
+	u8 __iomem *intr_sum_reg64;
+
+	/** PCI interrupt enable register */
+	u8 __iomem *intr_enb_reg64;
+
+	/** The PCI interrupt mask used by interrupt handler */
+	u64 intr_mask64;
+
+	struct octeon_config *conf;
+};
+
+#define CN23XX_SLI_DEF_BP			0x40
+
+struct oct_vf_stats {
+	u64 rx_packets;
+	u64 tx_packets;
+	u64 rx_bytes;
+	u64 tx_bytes;
+	u64 broadcast;
+	u64 multicast;
+};
+
+int setup_cn23xx_octeon_pf_device(struct octeon_device *oct);
+
+int validate_cn23xx_pf_config_info(struct octeon_device *oct,
+				   struct octeon_config *conf23xx);
+
+u32 cn23xx_pf_get_oq_ticks(struct octeon_device *oct, u32 time_intr_in_us);
+
+void cn23xx_dump_pf_initialized_regs(struct octeon_device *oct);
+
+int cn23xx_sriov_config(struct octeon_device *oct);
+
+int cn23xx_fw_loaded(struct octeon_device *oct);
+
+void cn23xx_tell_vf_its_macaddr_changed(struct octeon_device *oct, int vfidx,
+					u8 *mac);
+
+int cn23xx_get_vf_stats(struct octeon_device *oct, int ifidx,
+			struct oct_vf_stats *stats);
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_regs.h b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_regs.h
new file mode 100644
index 000000000..a0fd32476
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_regs.h
@@ -0,0 +1,599 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*! \file cn23xx_regs.h
+ * \brief Host Driver: Register Address and Register Mask values for
+ * Octeon CN23XX devices.
+ */
+
+#ifndef __CN23XX_PF_REGS_H__
+#define __CN23XX_PF_REGS_H__
+
+#define     CN23XX_CONFIG_VENDOR_ID	0x00
+#define     CN23XX_CONFIG_DEVICE_ID	0x02
+
+#define     CN23XX_CONFIG_XPANSION_BAR             0x38
+
+#define     CN23XX_CONFIG_MSIX_CAP		   0x50
+#define     CN23XX_CONFIG_MSIX_LMSI		   0x54
+#define     CN23XX_CONFIG_MSIX_UMSI		   0x58
+#define     CN23XX_CONFIG_MSIX_MSIMD		   0x5C
+#define     CN23XX_CONFIG_MSIX_MSIMM		   0x60
+#define     CN23XX_CONFIG_MSIX_MSIMP		   0x64
+
+#define     CN23XX_CONFIG_PCIE_CAP                 0x70
+#define     CN23XX_CONFIG_PCIE_DEVCAP              0x74
+#define     CN23XX_CONFIG_PCIE_DEVCTL              0x78
+#define     CN23XX_CONFIG_PCIE_LINKCAP             0x7C
+#define     CN23XX_CONFIG_PCIE_LINKCTL             0x80
+#define     CN23XX_CONFIG_PCIE_SLOTCAP             0x84
+#define     CN23XX_CONFIG_PCIE_SLOTCTL             0x88
+#define     CN23XX_CONFIG_PCIE_DEVCTL2             0x98
+#define     CN23XX_CONFIG_PCIE_LINKCTL2            0xA0
+#define     CN23XX_CONFIG_PCIE_UNCORRECT_ERR_MASK  0x108
+#define     CN23XX_CONFIG_PCIE_CORRECT_ERR_STATUS  0x110
+#define     CN23XX_CONFIG_PCIE_DEVCTL_MASK         0x00040000
+
+#define     CN23XX_PCIE_SRIOV_FDL		   0x188
+#define     CN23XX_PCIE_SRIOV_FDL_BIT_POS	   0x10
+#define     CN23XX_PCIE_SRIOV_FDL_MASK		   0xFF
+
+#define     CN23XX_CONFIG_PCIE_FLTMSK              0x720
+
+#define     CN23XX_CONFIG_SRIOV_VFDEVID            0x190
+
+#define     CN23XX_CONFIG_SRIOV_BAR_START	   0x19C
+#define     CN23XX_CONFIG_SRIOV_BARX(i)		\
+		(CN23XX_CONFIG_SRIOV_BAR_START + ((i) * 4))
+#define     CN23XX_CONFIG_SRIOV_BAR_PF		   0x08
+#define     CN23XX_CONFIG_SRIOV_BAR_64BIT	   0x04
+#define     CN23XX_CONFIG_SRIOV_BAR_IO		   0x01
+
+/* ##############  BAR0 Registers ################ */
+
+#define    CN23XX_SLI_CTL_PORT_START               0x286E0
+#define    CN23XX_PORT_OFFSET                      0x10
+
+#define    CN23XX_SLI_CTL_PORT(p)                  \
+		(CN23XX_SLI_CTL_PORT_START + ((p) * CN23XX_PORT_OFFSET))
+
+/* 2 scatch registers (64-bit)  */
+#define    CN23XX_SLI_WINDOW_CTL                   0x282E0
+#define    CN23XX_SLI_SCRATCH1                     0x283C0
+#define    CN23XX_SLI_SCRATCH2                     0x283D0
+#define    CN23XX_SLI_WINDOW_CTL_DEFAULT           0x200000ULL
+
+/* 1 registers (64-bit)  - SLI_CTL_STATUS */
+#define    CN23XX_SLI_CTL_STATUS                   0x28570
+
+/* SLI Packet Input Jabber Register (64 bit register)
+ * <31:0> for Byte count for limiting sizes of packet sizes
+ * that are allowed for sli packet inbound packets.
+ * the default value is 0xFA00(=64000).
+ */
+#define    CN23XX_SLI_PKT_IN_JABBER                0x29170
+/* The input jabber is used to determine the TSO max size.
+ * Due to H/W limitation, this needs to be reduced to 60000
+ * in order to use H/W TSO and avoid the WQE malformation
+ * PKO_BUG_24989_WQE_LEN
+ */
+#define    CN23XX_DEFAULT_INPUT_JABBER             0xEA60 /*60000*/
+
+#define    CN23XX_WIN_WR_ADDR_LO                   0x20000
+#define    CN23XX_WIN_WR_ADDR_HI                   0x20004
+#define    CN23XX_WIN_WR_ADDR64                    CN23XX_WIN_WR_ADDR_LO
+
+#define    CN23XX_WIN_RD_ADDR_LO                   0x20010
+#define    CN23XX_WIN_RD_ADDR_HI                   0x20014
+#define    CN23XX_WIN_RD_ADDR64                    CN23XX_WIN_RD_ADDR_LO
+
+#define    CN23XX_WIN_WR_DATA_LO                   0x20020
+#define    CN23XX_WIN_WR_DATA_HI                   0x20024
+#define    CN23XX_WIN_WR_DATA64                    CN23XX_WIN_WR_DATA_LO
+
+#define    CN23XX_WIN_RD_DATA_LO                   0x20040
+#define    CN23XX_WIN_RD_DATA_HI                   0x20044
+#define    CN23XX_WIN_RD_DATA64                    CN23XX_WIN_RD_DATA_LO
+
+#define    CN23XX_WIN_WR_MASK_LO                   0x20030
+#define    CN23XX_WIN_WR_MASK_HI                   0x20034
+#define    CN23XX_WIN_WR_MASK_REG                  CN23XX_WIN_WR_MASK_LO
+#define    CN23XX_SLI_MAC_CREDIT_CNT               0x23D70
+
+/* 4 registers (64-bit) for mapping IOQs to MACs(PEMs)-
+ * SLI_PKT_MAC(0..3)_PF(0..1)_RINFO
+ */
+#define    CN23XX_SLI_PKT_MAC_RINFO_START64       0x29030
+
+/*1 register (64-bit) to determine whether IOQs are in reset. */
+#define    CN23XX_SLI_PKT_IOQ_RING_RST            0x291E0
+
+/* Each Input Queue register is at a 16-byte Offset in BAR0 */
+#define    CN23XX_IQ_OFFSET                       0x20000
+
+#define    CN23XX_MAC_RINFO_OFFSET                0x20
+#define    CN23XX_PF_RINFO_OFFSET                 0x10
+
+#define CN23XX_SLI_PKT_MAC_RINFO64(mac, pf)		\
+		(CN23XX_SLI_PKT_MAC_RINFO_START64 +     \
+		 ((mac) * CN23XX_MAC_RINFO_OFFSET) +	\
+		 ((pf) * CN23XX_PF_RINFO_OFFSET))
+
+/** mask for total rings, setting TRS to base */
+#define    CN23XX_PKT_MAC_CTL_RINFO_TRS               BIT_ULL(16)
+/** mask for starting ring number: setting SRN <6:0> = 0x7F */
+#define    CN23XX_PKT_MAC_CTL_RINFO_SRN               (0x7F)
+
+/* Starting bit of the TRS field in CN23XX_SLI_PKT_MAC_RINFO64 register */
+#define    CN23XX_PKT_MAC_CTL_RINFO_TRS_BIT_POS     16
+/* Starting bit of SRN field in CN23XX_SLI_PKT_MAC_RINFO64 register */
+#define    CN23XX_PKT_MAC_CTL_RINFO_SRN_BIT_POS     0
+/* Starting bit of RPVF field in CN23XX_SLI_PKT_MAC_RINFO64 register */
+#define    CN23XX_PKT_MAC_CTL_RINFO_RPVF_BIT_POS     32
+/* Starting bit of NVFS field in CN23XX_SLI_PKT_MAC_RINFO64 register */
+#define    CN23XX_PKT_MAC_CTL_RINFO_NVFS_BIT_POS     48
+
+/*###################### REQUEST QUEUE #########################*/
+
+/* 64 registers for Input Queue Instr Count - SLI_PKT_IN_DONE0_CNTS */
+#define    CN23XX_SLI_IQ_INSTR_COUNT_START64     0x10040
+
+/* 64 registers for Input Queues Start Addr - SLI_PKT0_INSTR_BADDR */
+#define    CN23XX_SLI_IQ_BASE_ADDR_START64       0x10010
+
+/* 64 registers for Input Doorbell - SLI_PKT0_INSTR_BAOFF_DBELL */
+#define    CN23XX_SLI_IQ_DOORBELL_START          0x10020
+
+/* 64 registers for Input Queue size - SLI_PKT0_INSTR_FIFO_RSIZE */
+#define    CN23XX_SLI_IQ_SIZE_START              0x10030
+
+/* 64 registers (64-bit) - ES, RO, NS, Arbitration for Input Queue Data &
+ * gather list fetches. SLI_PKT(0..63)_INPUT_CONTROL.
+ */
+#define    CN23XX_SLI_IQ_PKT_CONTROL_START64    0x10000
+
+/*------- Request Queue Macros ---------*/
+#define    CN23XX_SLI_IQ_PKT_CONTROL64(iq)          \
+		(CN23XX_SLI_IQ_PKT_CONTROL_START64 + ((iq) * CN23XX_IQ_OFFSET))
+
+#define    CN23XX_SLI_IQ_BASE_ADDR64(iq)          \
+		(CN23XX_SLI_IQ_BASE_ADDR_START64 + ((iq) * CN23XX_IQ_OFFSET))
+
+#define    CN23XX_SLI_IQ_SIZE(iq)                 \
+		(CN23XX_SLI_IQ_SIZE_START + ((iq) * CN23XX_IQ_OFFSET))
+
+#define    CN23XX_SLI_IQ_DOORBELL(iq)             \
+		(CN23XX_SLI_IQ_DOORBELL_START + ((iq) * CN23XX_IQ_OFFSET))
+
+#define    CN23XX_SLI_IQ_INSTR_COUNT64(iq)          \
+		(CN23XX_SLI_IQ_INSTR_COUNT_START64 + ((iq) * CN23XX_IQ_OFFSET))
+
+/*------------------ Masks ----------------*/
+#define    CN23XX_PKT_INPUT_CTL_VF_NUM                  BIT_ULL(32)
+#define    CN23XX_PKT_INPUT_CTL_MAC_NUM                 BIT(29)
+/* Number of instructions to be read in one MAC read request.
+ * setting to Max value(4)
+ */
+#define    CN23XX_PKT_INPUT_CTL_RDSIZE                  (3 << 25)
+#define    CN23XX_PKT_INPUT_CTL_IS_64B                  BIT(24)
+#define    CN23XX_PKT_INPUT_CTL_RST                     BIT(23)
+#define    CN23XX_PKT_INPUT_CTL_QUIET                   BIT(28)
+#define    CN23XX_PKT_INPUT_CTL_RING_ENB                BIT(22)
+#define    CN23XX_PKT_INPUT_CTL_DATA_NS                 BIT(8)
+#define    CN23XX_PKT_INPUT_CTL_DATA_ES_64B_SWAP        BIT(6)
+#define    CN23XX_PKT_INPUT_CTL_DATA_RO                 BIT(5)
+#define    CN23XX_PKT_INPUT_CTL_USE_CSR                 BIT(4)
+#define    CN23XX_PKT_INPUT_CTL_GATHER_NS               BIT(3)
+#define    CN23XX_PKT_INPUT_CTL_GATHER_ES_64B_SWAP      (2)
+#define    CN23XX_PKT_INPUT_CTL_GATHER_RO               (1)
+
+/** Rings per Virtual Function **/
+#define    CN23XX_PKT_INPUT_CTL_RPVF_MASK               (0x3F)
+#define    CN23XX_PKT_INPUT_CTL_RPVF_POS                (48)
+/** These bits[47:44] select the Physical function number within the MAC */
+#define    CN23XX_PKT_INPUT_CTL_PF_NUM_MASK             (0x7)
+#define    CN23XX_PKT_INPUT_CTL_PF_NUM_POS              (45)
+/** These bits[43:32] select the function number within the PF */
+#define    CN23XX_PKT_INPUT_CTL_VF_NUM_MASK             (0x1FFF)
+#define    CN23XX_PKT_INPUT_CTL_VF_NUM_POS              (32)
+#define    CN23XX_PKT_INPUT_CTL_MAC_NUM_MASK            (0x3)
+#define    CN23XX_PKT_INPUT_CTL_MAC_NUM_POS             (29)
+#define    CN23XX_PKT_IN_DONE_WMARK_MASK                (0xFFFFULL)
+#define    CN23XX_PKT_IN_DONE_WMARK_BIT_POS             (32)
+#define    CN23XX_PKT_IN_DONE_CNT_MASK                  (0x00000000FFFFFFFFULL)
+
+#ifdef __LITTLE_ENDIAN_BITFIELD
+#define    CN23XX_PKT_INPUT_CTL_MASK				\
+		(CN23XX_PKT_INPUT_CTL_RDSIZE		|	\
+		 CN23XX_PKT_INPUT_CTL_DATA_ES_64B_SWAP	|	\
+		 CN23XX_PKT_INPUT_CTL_USE_CSR)
+#else
+#define    CN23XX_PKT_INPUT_CTL_MASK				\
+		(CN23XX_PKT_INPUT_CTL_RDSIZE		|	\
+		 CN23XX_PKT_INPUT_CTL_DATA_ES_64B_SWAP	|	\
+		 CN23XX_PKT_INPUT_CTL_USE_CSR		|	\
+		 CN23XX_PKT_INPUT_CTL_GATHER_ES_64B_SWAP)
+#endif
+
+/** Masks for SLI_PKT_IN_DONE(0..63)_CNTS Register */
+#define    CN23XX_IN_DONE_CNTS_PI_INT               BIT_ULL(62)
+#define    CN23XX_IN_DONE_CNTS_CINT_ENB             BIT_ULL(48)
+
+/*############################ OUTPUT QUEUE #########################*/
+
+/* 64 registers for Output queue control - SLI_PKT(0..63)_OUTPUT_CONTROL */
+#define    CN23XX_SLI_OQ_PKT_CONTROL_START       0x10050
+
+/* 64 registers for Output queue buffer and info size - SLI_PKT0_OUT_SIZE */
+#define    CN23XX_SLI_OQ0_BUFF_INFO_SIZE         0x10060
+
+/* 64 registers for Output Queue Start Addr - SLI_PKT0_SLIST_BADDR */
+#define    CN23XX_SLI_OQ_BASE_ADDR_START64       0x10070
+
+/* 64 registers for Output Queue Packet Credits - SLI_PKT0_SLIST_BAOFF_DBELL */
+#define    CN23XX_SLI_OQ_PKT_CREDITS_START       0x10080
+
+/* 64 registers for Output Queue size - SLI_PKT0_SLIST_FIFO_RSIZE */
+#define    CN23XX_SLI_OQ_SIZE_START              0x10090
+
+/* 64 registers for Output Queue Packet Count - SLI_PKT0_CNTS */
+#define    CN23XX_SLI_OQ_PKT_SENT_START          0x100B0
+
+/* 64 registers for Output Queue INT Levels - SLI_PKT0_INT_LEVELS */
+#define    CN23XX_SLI_OQ_PKT_INT_LEVELS_START64   0x100A0
+
+/* Each Output Queue register is at a 16-byte Offset in BAR0 */
+#define    CN23XX_OQ_OFFSET                      0x20000
+
+/* 1 (64-bit register) for Output Queue backpressure across all rings. */
+#define    CN23XX_SLI_OQ_WMARK                   0x29180
+
+/* Global pkt control register */
+#define    CN23XX_SLI_GBL_CONTROL                0x29210
+
+/* Backpressure enable register for PF0  */
+#define    CN23XX_SLI_OUT_BP_EN_W1S              0x29260
+
+/* Backpressure enable register for PF1  */
+#define    CN23XX_SLI_OUT_BP_EN2_W1S             0x29270
+
+/* Backpressure disable register for PF0  */
+#define    CN23XX_SLI_OUT_BP_EN_W1C              0x29280
+
+/* Backpressure disable register for PF1  */
+#define    CN23XX_SLI_OUT_BP_EN2_W1C             0x29290
+
+/*------- Output Queue Macros ---------*/
+
+#define    CN23XX_SLI_OQ_PKT_CONTROL(oq)          \
+		(CN23XX_SLI_OQ_PKT_CONTROL_START + ((oq) * CN23XX_OQ_OFFSET))
+
+#define    CN23XX_SLI_OQ_BASE_ADDR64(oq)          \
+		(CN23XX_SLI_OQ_BASE_ADDR_START64 + ((oq) * CN23XX_OQ_OFFSET))
+
+#define    CN23XX_SLI_OQ_SIZE(oq)                 \
+		(CN23XX_SLI_OQ_SIZE_START + ((oq) * CN23XX_OQ_OFFSET))
+
+#define    CN23XX_SLI_OQ_BUFF_INFO_SIZE(oq)                 \
+		(CN23XX_SLI_OQ0_BUFF_INFO_SIZE + ((oq) * CN23XX_OQ_OFFSET))
+
+#define    CN23XX_SLI_OQ_PKTS_SENT(oq)            \
+		(CN23XX_SLI_OQ_PKT_SENT_START + ((oq) * CN23XX_OQ_OFFSET))
+
+#define    CN23XX_SLI_OQ_PKTS_CREDIT(oq)          \
+		(CN23XX_SLI_OQ_PKT_CREDITS_START + ((oq) * CN23XX_OQ_OFFSET))
+
+#define    CN23XX_SLI_OQ_PKT_INT_LEVELS(oq)		\
+		(CN23XX_SLI_OQ_PKT_INT_LEVELS_START64 +	\
+		 ((oq) * CN23XX_OQ_OFFSET))
+
+/*Macro's for accessing CNT and TIME separately from INT_LEVELS*/
+#define    CN23XX_SLI_OQ_PKT_INT_LEVELS_CNT(oq)		\
+		(CN23XX_SLI_OQ_PKT_INT_LEVELS_START64 + \
+		 ((oq) * CN23XX_OQ_OFFSET))
+
+#define    CN23XX_SLI_OQ_PKT_INT_LEVELS_TIME(oq)	\
+		(CN23XX_SLI_OQ_PKT_INT_LEVELS_START64 +	\
+		 ((oq) * CN23XX_OQ_OFFSET) + 4)
+
+/*------------------ Masks ----------------*/
+#define    CN23XX_PKT_OUTPUT_CTL_TENB                  BIT(13)
+#define    CN23XX_PKT_OUTPUT_CTL_CENB                  BIT(12)
+#define    CN23XX_PKT_OUTPUT_CTL_IPTR                  BIT(11)
+#define    CN23XX_PKT_OUTPUT_CTL_ES                    BIT(9)
+#define    CN23XX_PKT_OUTPUT_CTL_NSR                   BIT(8)
+#define    CN23XX_PKT_OUTPUT_CTL_ROR                   BIT(7)
+#define    CN23XX_PKT_OUTPUT_CTL_DPTR                  BIT(6)
+#define    CN23XX_PKT_OUTPUT_CTL_BMODE                 BIT(5)
+#define    CN23XX_PKT_OUTPUT_CTL_ES_P                  BIT(3)
+#define    CN23XX_PKT_OUTPUT_CTL_NSR_P                 BIT(2)
+#define    CN23XX_PKT_OUTPUT_CTL_ROR_P                 BIT(1)
+#define    CN23XX_PKT_OUTPUT_CTL_RING_ENB              BIT(0)
+
+/*######################### Mailbox Reg Macros ########################*/
+#define    CN23XX_SLI_PKT_MBOX_INT_START             0x10210
+#define    CN23XX_SLI_PKT_PF_VF_MBOX_SIG_START       0x10200
+#define    CN23XX_SLI_MAC_PF_MBOX_INT_START          0x27380
+
+#define    CN23XX_SLI_MBOX_OFFSET		     0x20000
+#define    CN23XX_SLI_MBOX_SIG_IDX_OFFSET	     0x8
+
+#define    CN23XX_SLI_PKT_MBOX_INT(q)          \
+		(CN23XX_SLI_PKT_MBOX_INT_START + ((q) * CN23XX_SLI_MBOX_OFFSET))
+
+#define    CN23XX_SLI_PKT_PF_VF_MBOX_SIG(q, idx)		\
+		(CN23XX_SLI_PKT_PF_VF_MBOX_SIG_START +		\
+		 ((q) * CN23XX_SLI_MBOX_OFFSET +		\
+		  (idx) * CN23XX_SLI_MBOX_SIG_IDX_OFFSET))
+
+#define    CN23XX_SLI_MAC_PF_MBOX_INT(mac, pf)		\
+		(CN23XX_SLI_MAC_PF_MBOX_INT_START +	\
+		 ((mac) * CN23XX_MAC_INT_OFFSET +	\
+		  (pf) * CN23XX_PF_INT_OFFSET))
+
+/*######################### DMA Counters #########################*/
+
+/* 2 registers (64-bit) - DMA Count - 1 for each DMA counter 0/1. */
+#define    CN23XX_DMA_CNT_START                   0x28400
+
+/* 2 registers (64-bit) - DMA Timer 0/1, contains DMA timer values */
+/* SLI_DMA_0_TIM */
+#define    CN23XX_DMA_TIM_START                   0x28420
+
+/* 2 registers (64-bit) - DMA count & Time Interrupt threshold -
+ * SLI_DMA_0_INT_LEVEL
+ */
+#define    CN23XX_DMA_INT_LEVEL_START             0x283E0
+
+/* Each DMA register is at a 16-byte Offset in BAR0 */
+#define    CN23XX_DMA_OFFSET                      0x10
+
+/*---------- DMA Counter Macros ---------*/
+#define    CN23XX_DMA_CNT(dq)                      \
+		(CN23XX_DMA_CNT_START + ((dq) * CN23XX_DMA_OFFSET))
+
+#define    CN23XX_DMA_INT_LEVEL(dq)                \
+		(CN23XX_DMA_INT_LEVEL_START + ((dq) * CN23XX_DMA_OFFSET))
+
+#define    CN23XX_DMA_PKT_INT_LEVEL(dq)            \
+		(CN23XX_DMA_INT_LEVEL_START + ((dq) * CN23XX_DMA_OFFSET))
+
+#define    CN23XX_DMA_TIME_INT_LEVEL(dq)           \
+		(CN23XX_DMA_INT_LEVEL_START + 4 + ((dq) * CN23XX_DMA_OFFSET))
+
+#define    CN23XX_DMA_TIM(dq)                     \
+		(CN23XX_DMA_TIM_START + ((dq) * CN23XX_DMA_OFFSET))
+
+/*######################## MSIX TABLE #########################*/
+
+#define	CN23XX_MSIX_TABLE_ADDR_START		0x0
+#define	CN23XX_MSIX_TABLE_DATA_START		0x8
+
+#define	CN23XX_MSIX_TABLE_SIZE			0x10
+#define	CN23XX_MSIX_TABLE_ENTRIES		0x41
+
+#define CN23XX_MSIX_ENTRY_VECTOR_CTL	BIT_ULL(32)
+
+#define	CN23XX_MSIX_TABLE_ADDR(idx)		\
+	(CN23XX_MSIX_TABLE_ADDR_START + ((idx) * CN23XX_MSIX_TABLE_SIZE))
+
+#define	CN23XX_MSIX_TABLE_DATA(idx)		\
+	(CN23XX_MSIX_TABLE_DATA_START + ((idx) * CN23XX_MSIX_TABLE_SIZE))
+
+/*######################## INTERRUPTS #########################*/
+#define CN23XX_MAC_INT_OFFSET   0x20
+#define CN23XX_PF_INT_OFFSET    0x10
+
+/* 1 register (64-bit) for Interrupt Summary */
+#define    CN23XX_SLI_INT_SUM64            0x27000
+
+/* 4 registers (64-bit) for Interrupt Enable for each Port */
+#define    CN23XX_SLI_INT_ENB64            0x27080
+
+#define    CN23XX_SLI_MAC_PF_INT_SUM64(mac, pf)			\
+		(CN23XX_SLI_INT_SUM64 +				\
+		 ((mac) * CN23XX_MAC_INT_OFFSET) +		\
+		 ((pf) * CN23XX_PF_INT_OFFSET))
+
+#define    CN23XX_SLI_MAC_PF_INT_ENB64(mac, pf)		\
+		(CN23XX_SLI_INT_ENB64 +			\
+		 ((mac) * CN23XX_MAC_INT_OFFSET) +	\
+		 ((pf) * CN23XX_PF_INT_OFFSET))
+
+/* 1 register (64-bit) to indicate which Output Queue reached pkt threshold */
+#define    CN23XX_SLI_PKT_CNT_INT                0x29130
+
+/* 1 register (64-bit) to indicate which Output Queue reached time threshold */
+#define    CN23XX_SLI_PKT_TIME_INT               0x29140
+
+/*------------------ Interrupt Masks ----------------*/
+
+#define    CN23XX_INTR_PO_INT			BIT_ULL(63)
+#define    CN23XX_INTR_PI_INT			BIT_ULL(62)
+#define    CN23XX_INTR_MBOX_INT			BIT_ULL(61)
+#define    CN23XX_INTR_RESEND			BIT_ULL(60)
+
+#define    CN23XX_INTR_CINT_ENB                 BIT_ULL(48)
+#define    CN23XX_INTR_MBOX_ENB                 BIT(0)
+
+#define    CN23XX_INTR_RML_TIMEOUT_ERR           (1)
+
+#define    CN23XX_INTR_MIO_INT                   BIT(1)
+
+#define    CN23XX_INTR_RESERVED1                 (3 << 2)
+
+#define    CN23XX_INTR_PKT_COUNT                 BIT(4)
+#define    CN23XX_INTR_PKT_TIME                  BIT(5)
+
+#define    CN23XX_INTR_RESERVED2                 (3 << 6)
+
+#define    CN23XX_INTR_M0UPB0_ERR                BIT(8)
+#define    CN23XX_INTR_M0UPWI_ERR                BIT(9)
+#define    CN23XX_INTR_M0UNB0_ERR                BIT(10)
+#define    CN23XX_INTR_M0UNWI_ERR                BIT(11)
+
+#define    CN23XX_INTR_RESERVED3                 (0xFFFFFULL << 12)
+
+#define    CN23XX_INTR_DMA0_FORCE                BIT_ULL(32)
+#define    CN23XX_INTR_DMA1_FORCE                BIT_ULL(33)
+
+#define    CN23XX_INTR_DMA0_COUNT                BIT_ULL(34)
+#define    CN23XX_INTR_DMA1_COUNT                BIT_ULL(35)
+
+#define    CN23XX_INTR_DMA0_TIME                 BIT_ULL(36)
+#define    CN23XX_INTR_DMA1_TIME                 BIT_ULL(37)
+
+#define    CN23XX_INTR_RESERVED4                 (0x7FFFFULL << 38)
+
+#define    CN23XX_INTR_VF_MBOX                   BIT_ULL(57)
+#define    CN23XX_INTR_DMAVF_ERR                 BIT_ULL(58)
+#define    CN23XX_INTR_DMAPF_ERR                 BIT_ULL(59)
+
+#define    CN23XX_INTR_PKTVF_ERR                 BIT_ULL(60)
+#define    CN23XX_INTR_PKTPF_ERR                 BIT_ULL(61)
+#define    CN23XX_INTR_PPVF_ERR                  BIT_ULL(62)
+#define    CN23XX_INTR_PPPF_ERR                  BIT_ULL(63)
+
+#define    CN23XX_INTR_DMA0_DATA                 (CN23XX_INTR_DMA0_TIME)
+#define    CN23XX_INTR_DMA1_DATA                 (CN23XX_INTR_DMA1_TIME)
+
+#define    CN23XX_INTR_DMA_DATA                  \
+		(CN23XX_INTR_DMA0_DATA | CN23XX_INTR_DMA1_DATA)
+
+/* By fault only TIME based */
+#define    CN23XX_INTR_PKT_DATA                  (CN23XX_INTR_PKT_TIME)
+/* For both COUNT and TIME based */
+/* #define    CN23XX_INTR_PKT_DATA                  \
+ * (CN23XX_INTR_PKT_COUNT | CN23XX_INTR_PKT_TIME)
+ */
+
+/* Sum of interrupts for all PCI-Express Data Interrupts */
+#define    CN23XX_INTR_PCIE_DATA                 \
+		(CN23XX_INTR_DMA_DATA | CN23XX_INTR_PKT_DAT)
+
+/* Sum of interrupts for error events */
+#define    CN23XX_INTR_ERR			\
+		(CN23XX_INTR_M0UPB0_ERR	|	\
+		 CN23XX_INTR_M0UPWI_ERR	|	\
+		 CN23XX_INTR_M0UNB0_ERR	|	\
+		 CN23XX_INTR_M0UNWI_ERR	|	\
+		 CN23XX_INTR_DMAVF_ERR	|	\
+		 CN23XX_INTR_DMAPF_ERR	|	\
+		 CN23XX_INTR_PKTPF_ERR	|	\
+		 CN23XX_INTR_PPPF_ERR	|	\
+		 CN23XX_INTR_PPVF_ERR)
+
+/* Programmed Mask for Interrupt Sum */
+#define    CN23XX_INTR_MASK			\
+		(CN23XX_INTR_DMA_DATA	|	\
+		 CN23XX_INTR_DMA0_FORCE	|	\
+		 CN23XX_INTR_DMA1_FORCE	|	\
+		 CN23XX_INTR_MIO_INT	|	\
+		 CN23XX_INTR_ERR)
+
+/* 4 Registers (64 - bit) */
+#define    CN23XX_SLI_S2M_PORT_CTL_START         0x23D80
+#define    CN23XX_SLI_S2M_PORTX_CTL(port)	\
+		(CN23XX_SLI_S2M_PORT_CTL_START + ((port) * 0x10))
+
+#define    CN23XX_SLI_MAC_NUMBER                 0x20050
+
+/** PEM(0..3)_BAR1_INDEX(0..15)address is defined as
+ *  addr = (0x00011800C0000100  |port <<24 |idx <<3 )
+ *  Here, port is PEM(0..3) & idx is INDEX(0..15)
+ */
+#define    CN23XX_PEM_BAR1_INDEX_START             0x00011800C0000100ULL
+#define    CN23XX_PEM_OFFSET                       24
+#define    CN23XX_BAR1_INDEX_OFFSET                3
+
+#define    CN23XX_PEM_BAR1_INDEX_REG(port, idx)		\
+		(CN23XX_PEM_BAR1_INDEX_START + (((u64)port) << CN23XX_PEM_OFFSET) + \
+		 ((idx) << CN23XX_BAR1_INDEX_OFFSET))
+
+/*############################ DPI #########################*/
+
+/* 1 register (64-bit) - provides DMA Enable */
+#define    CN23XX_DPI_CTL                 0x0001df0000000040ULL
+
+/* 1 register (64-bit) - Controls the DMA IO Operation */
+#define    CN23XX_DPI_DMA_CONTROL         0x0001df0000000048ULL
+
+/* 1 register (64-bit) - Provides DMA Instr'n Queue Enable  */
+#define    CN23XX_DPI_REQ_GBL_ENB         0x0001df0000000050ULL
+
+/* 1 register (64-bit) - DPI_REQ_ERR_RSP
+ * Indicates which Instr'n Queue received error response from the IO sub-system
+ */
+#define    CN23XX_DPI_REQ_ERR_RSP         0x0001df0000000058ULL
+
+/* 1 register (64-bit) - DPI_REQ_ERR_RST
+ * Indicates which Instr'n Queue dropped an Instr'n
+ */
+#define    CN23XX_DPI_REQ_ERR_RST         0x0001df0000000060ULL
+
+/* 6 register (64-bit) - DPI_DMA_ENG(0..5)_EN
+ * Provides DMA Engine Queue Enable
+ */
+#define    CN23XX_DPI_DMA_ENG0_ENB        0x0001df0000000080ULL
+#define    CN23XX_DPI_DMA_ENG_ENB(eng) (CN23XX_DPI_DMA_ENG0_ENB + ((eng) * 8))
+
+/* 8 register (64-bit) - DPI_DMA(0..7)_REQQ_CTL
+ * Provides control bits for transaction on 8 Queues
+ */
+#define    CN23XX_DPI_DMA_REQQ0_CTL       0x0001df0000000180ULL
+#define    CN23XX_DPI_DMA_REQQ_CTL(q_no)	\
+		(CN23XX_DPI_DMA_REQQ0_CTL + ((q_no) * 8))
+
+/* 6 register (64-bit) - DPI_ENG(0..5)_BUF
+ * Provides DMA Engine FIFO (Queue) Size
+ */
+#define    CN23XX_DPI_DMA_ENG0_BUF        0x0001df0000000880ULL
+#define    CN23XX_DPI_DMA_ENG_BUF(eng)   \
+		(CN23XX_DPI_DMA_ENG0_BUF + ((eng) * 8))
+
+/* 4 Registers (64-bit) */
+#define    CN23XX_DPI_SLI_PRT_CFG_START   0x0001df0000000900ULL
+#define    CN23XX_DPI_SLI_PRTX_CFG(port)        \
+		(CN23XX_DPI_SLI_PRT_CFG_START + ((port) * 0x8))
+
+/* Masks for DPI_DMA_CONTROL Register */
+#define    CN23XX_DPI_DMA_COMMIT_MODE     BIT_ULL(58)
+#define    CN23XX_DPI_DMA_PKT_EN          BIT_ULL(56)
+#define    CN23XX_DPI_DMA_ENB             (0x0FULL << 48)
+/* Set the DMA Control, to update packet count not byte count sent by DMA,
+ * when we use Interrupt Coalescing (CA mode)
+ */
+#define    CN23XX_DPI_DMA_O_ADD1          BIT(19)
+/*selecting 64-bit Byte Swap Mode */
+#define    CN23XX_DPI_DMA_O_ES            BIT(15)
+#define    CN23XX_DPI_DMA_O_MODE          BIT(14)
+
+#define    CN23XX_DPI_DMA_CTL_MASK			\
+		(CN23XX_DPI_DMA_COMMIT_MODE	|	\
+		 CN23XX_DPI_DMA_PKT_EN		|	\
+		 CN23XX_DPI_DMA_O_ES		|	\
+		 CN23XX_DPI_DMA_O_MODE)
+
+/*############################ RST #########################*/
+
+#define    CN23XX_RST_BOOT            0x0001180006001600ULL
+#define    CN23XX_RST_SOFT_RST        0x0001180006001680ULL
+
+#define    CN23XX_LMC0_RESET_CTL               0x0001180088000180ULL
+#define    CN23XX_LMC0_RESET_CTL_DDR3RST_MASK  0x0000000000000001ULL
+
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/cn23xx_vf_device.c b/drivers/net/ethernet/cavium/liquidio/cn23xx_vf_device.c
new file mode 100644
index 000000000..fda494049
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn23xx_vf_device.c
@@ -0,0 +1,682 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include <linux/vmalloc.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "cn23xx_vf_device.h"
+#include "octeon_main.h"
+#include "octeon_mailbox.h"
+
+u32 cn23xx_vf_get_oq_ticks(struct octeon_device *oct, u32 time_intr_in_us)
+{
+	/* This gives the SLI clock per microsec */
+	u32 oqticks_per_us = (u32)oct->pfvf_hsword.coproc_tics_per_us;
+
+	/* This gives the clock cycles per millisecond */
+	oqticks_per_us *= 1000;
+
+	/* This gives the oq ticks (1024 core clock cycles) per millisecond */
+	oqticks_per_us /= 1024;
+
+	/* time_intr is in microseconds. The next 2 steps gives the oq ticks
+	 * corressponding to time_intr.
+	 */
+	oqticks_per_us *= time_intr_in_us;
+	oqticks_per_us /= 1000;
+
+	return oqticks_per_us;
+}
+
+static int cn23xx_vf_reset_io_queues(struct octeon_device *oct, u32 num_queues)
+{
+	u32 loop = BUSY_READING_REG_VF_LOOP_COUNT;
+	int ret_val = 0;
+	u32 q_no;
+	u64 d64;
+
+	for (q_no = 0; q_no < num_queues; q_no++) {
+		/* set RST bit to 1. This bit applies to both IQ and OQ */
+		d64 = octeon_read_csr64(oct,
+					CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no));
+		d64 |= CN23XX_PKT_INPUT_CTL_RST;
+		octeon_write_csr64(oct, CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no),
+				   d64);
+	}
+
+	/* wait until the RST bit is clear or the RST and QUIET bits are set */
+	for (q_no = 0; q_no < num_queues; q_no++) {
+		u64 reg_val = octeon_read_csr64(oct,
+					CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no));
+		while ((READ_ONCE(reg_val) & CN23XX_PKT_INPUT_CTL_RST) &&
+		       !(READ_ONCE(reg_val) & CN23XX_PKT_INPUT_CTL_QUIET) &&
+		       loop) {
+			WRITE_ONCE(reg_val, octeon_read_csr64(
+			    oct, CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no)));
+			loop--;
+		}
+		if (!loop) {
+			dev_err(&oct->pci_dev->dev,
+				"clearing the reset reg failed or setting the quiet reg failed for qno: %u\n",
+				q_no);
+			return -1;
+		}
+		WRITE_ONCE(reg_val, READ_ONCE(reg_val) &
+			   ~CN23XX_PKT_INPUT_CTL_RST);
+		octeon_write_csr64(oct, CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no),
+				   READ_ONCE(reg_val));
+
+		WRITE_ONCE(reg_val, octeon_read_csr64(
+		    oct, CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no)));
+		if (READ_ONCE(reg_val) & CN23XX_PKT_INPUT_CTL_RST) {
+			dev_err(&oct->pci_dev->dev,
+				"clearing the reset failed for qno: %u\n",
+				q_no);
+			ret_val = -1;
+		}
+	}
+
+	return ret_val;
+}
+
+static int cn23xx_vf_setup_global_input_regs(struct octeon_device *oct)
+{
+	struct octeon_cn23xx_vf *cn23xx = (struct octeon_cn23xx_vf *)oct->chip;
+	struct octeon_instr_queue *iq;
+	u64 q_no, intr_threshold;
+	u64 d64;
+
+	if (cn23xx_vf_reset_io_queues(oct, oct->sriov_info.rings_per_vf))
+		return -1;
+
+	for (q_no = 0; q_no < (oct->sriov_info.rings_per_vf); q_no++) {
+		void __iomem *inst_cnt_reg;
+
+		octeon_write_csr64(oct, CN23XX_VF_SLI_IQ_DOORBELL(q_no),
+				   0xFFFFFFFF);
+		iq = oct->instr_queue[q_no];
+
+		if (iq)
+			inst_cnt_reg = iq->inst_cnt_reg;
+		else
+			inst_cnt_reg = (u8 *)oct->mmio[0].hw_addr +
+				       CN23XX_VF_SLI_IQ_INSTR_COUNT64(q_no);
+
+		d64 = octeon_read_csr64(oct,
+					CN23XX_VF_SLI_IQ_INSTR_COUNT64(q_no));
+
+		d64 &= 0xEFFFFFFFFFFFFFFFL;
+
+		octeon_write_csr64(oct, CN23XX_VF_SLI_IQ_INSTR_COUNT64(q_no),
+				   d64);
+
+		/* Select ES, RO, NS, RDSIZE,DPTR Fomat#0 for
+		 * the Input Queues
+		 */
+		octeon_write_csr64(oct, CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no),
+				   CN23XX_PKT_INPUT_CTL_MASK);
+
+		/* set the wmark level to trigger PI_INT */
+		intr_threshold = CFG_GET_IQ_INTR_PKT(cn23xx->conf) &
+				 CN23XX_PKT_IN_DONE_WMARK_MASK;
+
+		writeq((readq(inst_cnt_reg) &
+			~(CN23XX_PKT_IN_DONE_WMARK_MASK <<
+			  CN23XX_PKT_IN_DONE_WMARK_BIT_POS)) |
+		       (intr_threshold << CN23XX_PKT_IN_DONE_WMARK_BIT_POS),
+		       inst_cnt_reg);
+	}
+	return 0;
+}
+
+static void cn23xx_vf_setup_global_output_regs(struct octeon_device *oct)
+{
+	u32 reg_val;
+	u32 q_no;
+
+	for (q_no = 0; q_no < (oct->sriov_info.rings_per_vf); q_no++) {
+		octeon_write_csr(oct, CN23XX_VF_SLI_OQ_PKTS_CREDIT(q_no),
+				 0xFFFFFFFF);
+
+		reg_val =
+		    octeon_read_csr(oct, CN23XX_VF_SLI_OQ_PKTS_SENT(q_no));
+
+		reg_val &= 0xEFFFFFFFFFFFFFFFL;
+
+		reg_val =
+		    octeon_read_csr(oct, CN23XX_VF_SLI_OQ_PKT_CONTROL(q_no));
+
+		/* clear IPTR */
+		reg_val &= ~CN23XX_PKT_OUTPUT_CTL_IPTR;
+
+		/* set DPTR */
+		reg_val |= CN23XX_PKT_OUTPUT_CTL_DPTR;
+
+		/* reset BMODE */
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_BMODE);
+
+		/* No Relaxed Ordering, No Snoop, 64-bit Byte swap
+		 * for Output Queue ScatterList reset ROR_P, NSR_P
+		 */
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_ROR_P);
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_NSR_P);
+
+#ifdef __LITTLE_ENDIAN_BITFIELD
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_ES_P);
+#else
+		reg_val |= (CN23XX_PKT_OUTPUT_CTL_ES_P);
+#endif
+		/* No Relaxed Ordering, No Snoop, 64-bit Byte swap
+		 * for Output Queue Data reset ROR, NSR
+		 */
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_ROR);
+		reg_val &= ~(CN23XX_PKT_OUTPUT_CTL_NSR);
+		/* set the ES bit */
+		reg_val |= (CN23XX_PKT_OUTPUT_CTL_ES);
+
+		/* write all the selected settings */
+		octeon_write_csr(oct, CN23XX_VF_SLI_OQ_PKT_CONTROL(q_no),
+				 reg_val);
+	}
+}
+
+static int cn23xx_setup_vf_device_regs(struct octeon_device *oct)
+{
+	if (cn23xx_vf_setup_global_input_regs(oct))
+		return -1;
+
+	cn23xx_vf_setup_global_output_regs(oct);
+
+	return 0;
+}
+
+static void cn23xx_setup_vf_iq_regs(struct octeon_device *oct, u32 iq_no)
+{
+	struct octeon_instr_queue *iq = oct->instr_queue[iq_no];
+	u64 pkt_in_done;
+
+	/* Write the start of the input queue's ring and its size */
+	octeon_write_csr64(oct, CN23XX_VF_SLI_IQ_BASE_ADDR64(iq_no),
+			   iq->base_addr_dma);
+	octeon_write_csr(oct, CN23XX_VF_SLI_IQ_SIZE(iq_no), iq->max_count);
+
+	/* Remember the doorbell & instruction count register addr
+	 * for this queue
+	 */
+	iq->doorbell_reg =
+	    (u8 *)oct->mmio[0].hw_addr + CN23XX_VF_SLI_IQ_DOORBELL(iq_no);
+	iq->inst_cnt_reg =
+	    (u8 *)oct->mmio[0].hw_addr + CN23XX_VF_SLI_IQ_INSTR_COUNT64(iq_no);
+	dev_dbg(&oct->pci_dev->dev, "InstQ[%d]:dbell reg @ 0x%p instcnt_reg @ 0x%p\n",
+		iq_no, iq->doorbell_reg, iq->inst_cnt_reg);
+
+	/* Store the current instruction counter (used in flush_iq
+	 * calculation)
+	 */
+	pkt_in_done = readq(iq->inst_cnt_reg);
+
+	if (oct->msix_on) {
+		/* Set CINT_ENB to enable IQ interrupt */
+		writeq((pkt_in_done | CN23XX_INTR_CINT_ENB),
+		       iq->inst_cnt_reg);
+	}
+	iq->reset_instr_cnt = 0;
+}
+
+static void cn23xx_setup_vf_oq_regs(struct octeon_device *oct, u32 oq_no)
+{
+	struct octeon_droq *droq = oct->droq[oq_no];
+
+	octeon_write_csr64(oct, CN23XX_VF_SLI_OQ_BASE_ADDR64(oq_no),
+			   droq->desc_ring_dma);
+	octeon_write_csr(oct, CN23XX_VF_SLI_OQ_SIZE(oq_no), droq->max_count);
+
+	octeon_write_csr(oct, CN23XX_VF_SLI_OQ_BUFF_INFO_SIZE(oq_no),
+			 droq->buffer_size);
+
+	/* Get the mapped address of the pkt_sent and pkts_credit regs */
+	droq->pkts_sent_reg =
+	    (u8 *)oct->mmio[0].hw_addr + CN23XX_VF_SLI_OQ_PKTS_SENT(oq_no);
+	droq->pkts_credit_reg =
+	    (u8 *)oct->mmio[0].hw_addr + CN23XX_VF_SLI_OQ_PKTS_CREDIT(oq_no);
+}
+
+static void cn23xx_vf_mbox_thread(struct work_struct *work)
+{
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct octeon_mbox *mbox = (struct octeon_mbox *)wk->ctxptr;
+
+	octeon_mbox_process_message(mbox);
+}
+
+static int cn23xx_free_vf_mbox(struct octeon_device *oct)
+{
+	cancel_delayed_work_sync(&oct->mbox[0]->mbox_poll_wk.work);
+	vfree(oct->mbox[0]);
+	return 0;
+}
+
+static int cn23xx_setup_vf_mbox(struct octeon_device *oct)
+{
+	struct octeon_mbox *mbox = NULL;
+
+	mbox = vmalloc(sizeof(*mbox));
+	if (!mbox)
+		return 1;
+
+	memset(mbox, 0, sizeof(struct octeon_mbox));
+
+	spin_lock_init(&mbox->lock);
+
+	mbox->oct_dev = oct;
+
+	mbox->q_no = 0;
+
+	mbox->state = OCTEON_MBOX_STATE_IDLE;
+
+	/* VF mbox interrupt reg */
+	mbox->mbox_int_reg =
+	    (u8 *)oct->mmio[0].hw_addr + CN23XX_VF_SLI_PKT_MBOX_INT(0);
+	/* VF reads from SIG0 reg */
+	mbox->mbox_read_reg =
+	    (u8 *)oct->mmio[0].hw_addr + CN23XX_SLI_PKT_PF_VF_MBOX_SIG(0, 0);
+	/* VF writes into SIG1 reg */
+	mbox->mbox_write_reg =
+	    (u8 *)oct->mmio[0].hw_addr + CN23XX_SLI_PKT_PF_VF_MBOX_SIG(0, 1);
+
+	INIT_DELAYED_WORK(&mbox->mbox_poll_wk.work,
+			  cn23xx_vf_mbox_thread);
+
+	mbox->mbox_poll_wk.ctxptr = mbox;
+
+	oct->mbox[0] = mbox;
+
+	writeq(OCTEON_PFVFSIG, mbox->mbox_read_reg);
+
+	return 0;
+}
+
+static int cn23xx_enable_vf_io_queues(struct octeon_device *oct)
+{
+	u32 q_no;
+
+	for (q_no = 0; q_no < oct->num_iqs; q_no++) {
+		u64 reg_val;
+
+		/* set the corresponding IQ IS_64B bit */
+		if (oct->io_qmask.iq64B & BIT_ULL(q_no)) {
+			reg_val = octeon_read_csr64(
+			    oct, CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no));
+			reg_val |= CN23XX_PKT_INPUT_CTL_IS_64B;
+			octeon_write_csr64(
+			    oct, CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no), reg_val);
+		}
+
+		/* set the corresponding IQ ENB bit */
+		if (oct->io_qmask.iq & BIT_ULL(q_no)) {
+			reg_val = octeon_read_csr64(
+			    oct, CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no));
+			reg_val |= CN23XX_PKT_INPUT_CTL_RING_ENB;
+			octeon_write_csr64(
+			    oct, CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no), reg_val);
+		}
+	}
+	for (q_no = 0; q_no < oct->num_oqs; q_no++) {
+		u32 reg_val;
+
+		/* set the corresponding OQ ENB bit */
+		if (oct->io_qmask.oq & BIT_ULL(q_no)) {
+			reg_val = octeon_read_csr(
+			    oct, CN23XX_VF_SLI_OQ_PKT_CONTROL(q_no));
+			reg_val |= CN23XX_PKT_OUTPUT_CTL_RING_ENB;
+			octeon_write_csr(
+			    oct, CN23XX_VF_SLI_OQ_PKT_CONTROL(q_no), reg_val);
+		}
+	}
+
+	return 0;
+}
+
+static void cn23xx_disable_vf_io_queues(struct octeon_device *oct)
+{
+	u32 num_queues = oct->num_iqs;
+
+	/* per HRM, rings can only be disabled via reset operation,
+	 * NOT via SLI_PKT()_INPUT/OUTPUT_CONTROL[ENB]
+	 */
+	if (num_queues < oct->num_oqs)
+		num_queues = oct->num_oqs;
+
+	cn23xx_vf_reset_io_queues(oct, num_queues);
+}
+
+void cn23xx_vf_ask_pf_to_do_flr(struct octeon_device *oct)
+{
+	struct octeon_mbox_cmd mbox_cmd;
+
+	mbox_cmd.msg.u64 = 0;
+	mbox_cmd.msg.s.type = OCTEON_MBOX_REQUEST;
+	mbox_cmd.msg.s.resp_needed = 0;
+	mbox_cmd.msg.s.cmd = OCTEON_VF_FLR_REQUEST;
+	mbox_cmd.msg.s.len = 1;
+	mbox_cmd.q_no = 0;
+	mbox_cmd.recv_len = 0;
+	mbox_cmd.recv_status = 0;
+	mbox_cmd.fn = NULL;
+	mbox_cmd.fn_arg = NULL;
+
+	octeon_mbox_write(oct, &mbox_cmd);
+}
+
+static void octeon_pfvf_hs_callback(struct octeon_device *oct,
+				    struct octeon_mbox_cmd *cmd,
+				    void *arg)
+{
+	u32 major = 0;
+
+	memcpy((uint8_t *)&oct->pfvf_hsword, cmd->msg.s.params,
+	       CN23XX_MAILBOX_MSGPARAM_SIZE);
+	if (cmd->recv_len > 1)  {
+		major = ((struct lio_version *)(cmd->data))->major;
+		major = major << 16;
+	}
+
+	atomic_set((atomic_t *)arg, major | 1);
+}
+
+int cn23xx_octeon_pfvf_handshake(struct octeon_device *oct)
+{
+	struct octeon_mbox_cmd mbox_cmd;
+	u32 q_no, count = 0;
+	atomic_t status;
+	u32 pfmajor;
+	u32 vfmajor;
+	u32 ret;
+
+	/* Sending VF_ACTIVE indication to the PF driver */
+	dev_dbg(&oct->pci_dev->dev, "requesting info from pf\n");
+
+	mbox_cmd.msg.u64 = 0;
+	mbox_cmd.msg.s.type = OCTEON_MBOX_REQUEST;
+	mbox_cmd.msg.s.resp_needed = 1;
+	mbox_cmd.msg.s.cmd = OCTEON_VF_ACTIVE;
+	mbox_cmd.msg.s.len = 2;
+	mbox_cmd.data[0] = 0;
+	((struct lio_version *)&mbox_cmd.data[0])->major =
+						LIQUIDIO_BASE_MAJOR_VERSION;
+	((struct lio_version *)&mbox_cmd.data[0])->minor =
+						LIQUIDIO_BASE_MINOR_VERSION;
+	((struct lio_version *)&mbox_cmd.data[0])->micro =
+						LIQUIDIO_BASE_MICRO_VERSION;
+	mbox_cmd.q_no = 0;
+	mbox_cmd.recv_len = 0;
+	mbox_cmd.recv_status = 0;
+	mbox_cmd.fn = (octeon_mbox_callback_t)octeon_pfvf_hs_callback;
+	mbox_cmd.fn_arg = &status;
+
+	octeon_mbox_write(oct, &mbox_cmd);
+
+	atomic_set(&status, 0);
+
+	do {
+		schedule_timeout_uninterruptible(1);
+	} while ((!atomic_read(&status)) && (count++ < 100000));
+
+	ret = atomic_read(&status);
+	if (!ret) {
+		dev_err(&oct->pci_dev->dev, "octeon_pfvf_handshake timeout\n");
+		return 1;
+	}
+
+	for (q_no = 0 ; q_no < oct->num_iqs ; q_no++)
+		oct->instr_queue[q_no]->txpciq.s.pkind = oct->pfvf_hsword.pkind;
+
+	vfmajor = LIQUIDIO_BASE_MAJOR_VERSION;
+	pfmajor = ret >> 16;
+	if (pfmajor != vfmajor) {
+		dev_err(&oct->pci_dev->dev,
+			"VF Liquidio driver (major version %d) is not compatible with Liquidio PF driver (major version %d)\n",
+			vfmajor, pfmajor);
+		return 1;
+	}
+
+	dev_dbg(&oct->pci_dev->dev,
+		"VF Liquidio driver (major version %d), Liquidio PF driver (major version %d)\n",
+		vfmajor, pfmajor);
+
+	dev_dbg(&oct->pci_dev->dev, "got data from pf pkind is %d\n",
+		oct->pfvf_hsword.pkind);
+
+	return 0;
+}
+
+static void cn23xx_handle_vf_mbox_intr(struct octeon_ioq_vector *ioq_vector)
+{
+	struct octeon_device *oct = ioq_vector->oct_dev;
+	u64 mbox_int_val;
+
+	if (!ioq_vector->droq_index) {
+		/* read and clear by writing 1 */
+		mbox_int_val = readq(oct->mbox[0]->mbox_int_reg);
+		writeq(mbox_int_val, oct->mbox[0]->mbox_int_reg);
+		if (octeon_mbox_read(oct->mbox[0]))
+			schedule_delayed_work(&oct->mbox[0]->mbox_poll_wk.work,
+					      msecs_to_jiffies(0));
+	}
+}
+
+static u64 cn23xx_vf_msix_interrupt_handler(void *dev)
+{
+	struct octeon_ioq_vector *ioq_vector = (struct octeon_ioq_vector *)dev;
+	struct octeon_device *oct = ioq_vector->oct_dev;
+	struct octeon_droq *droq = oct->droq[ioq_vector->droq_index];
+	u64 pkts_sent;
+	u64 ret = 0;
+
+	dev_dbg(&oct->pci_dev->dev, "In %s octeon_dev @ %p\n", __func__, oct);
+	pkts_sent = readq(droq->pkts_sent_reg);
+
+	/* If our device has interrupted, then proceed. Also check
+	 * for all f's if interrupt was triggered on an error
+	 * and the PCI read fails.
+	 */
+	if (!pkts_sent || (pkts_sent == 0xFFFFFFFFFFFFFFFFULL))
+		return ret;
+
+	/* Write count reg in sli_pkt_cnts to clear these int. */
+	if ((pkts_sent & CN23XX_INTR_PO_INT) ||
+	    (pkts_sent & CN23XX_INTR_PI_INT)) {
+		if (pkts_sent & CN23XX_INTR_PO_INT)
+			ret |= MSIX_PO_INT;
+	}
+
+	if (pkts_sent & CN23XX_INTR_PI_INT)
+		/* We will clear the count when we update the read_index. */
+		ret |= MSIX_PI_INT;
+
+	if (pkts_sent & CN23XX_INTR_MBOX_INT) {
+		cn23xx_handle_vf_mbox_intr(ioq_vector);
+		ret |= MSIX_MBOX_INT;
+	}
+
+	return ret;
+}
+
+static u32 cn23xx_update_read_index(struct octeon_instr_queue *iq)
+{
+	u32 pkt_in_done = readl(iq->inst_cnt_reg);
+	u32 last_done;
+	u32 new_idx;
+
+	last_done = pkt_in_done - iq->pkt_in_done;
+	iq->pkt_in_done = pkt_in_done;
+
+	/* Modulo of the new index with the IQ size will give us
+	 * the new index.  The iq->reset_instr_cnt is always zero for
+	 * cn23xx, so no extra adjustments are needed.
+	 */
+	new_idx = (iq->octeon_read_index +
+		   (u32)(last_done & CN23XX_PKT_IN_DONE_CNT_MASK)) %
+		  iq->max_count;
+
+	return new_idx;
+}
+
+static void cn23xx_enable_vf_interrupt(struct octeon_device *oct, u8 intr_flag)
+{
+	struct octeon_cn23xx_vf *cn23xx = (struct octeon_cn23xx_vf *)oct->chip;
+	u32 q_no, time_threshold;
+
+	if (intr_flag & OCTEON_OUTPUT_INTR) {
+		for (q_no = 0; q_no < oct->num_oqs; q_no++) {
+			/* Set up interrupt packet and time thresholds
+			 * for all the OQs
+			 */
+			time_threshold = cn23xx_vf_get_oq_ticks(
+				oct, (u32)CFG_GET_OQ_INTR_TIME(cn23xx->conf));
+
+			octeon_write_csr64(
+			    oct, CN23XX_VF_SLI_OQ_PKT_INT_LEVELS(q_no),
+			    (CFG_GET_OQ_INTR_PKT(cn23xx->conf) |
+			     ((u64)time_threshold << 32)));
+		}
+	}
+
+	if (intr_flag & OCTEON_INPUT_INTR) {
+		for (q_no = 0; q_no < oct->num_oqs; q_no++) {
+			/* Set CINT_ENB to enable IQ interrupt */
+			octeon_write_csr64(
+			    oct, CN23XX_VF_SLI_IQ_INSTR_COUNT64(q_no),
+			    ((octeon_read_csr64(
+				  oct, CN23XX_VF_SLI_IQ_INSTR_COUNT64(q_no)) &
+			      ~CN23XX_PKT_IN_DONE_CNT_MASK) |
+			     CN23XX_INTR_CINT_ENB));
+		}
+	}
+
+	/* Set queue-0 MBOX_ENB to enable VF mailbox interrupt */
+	if (intr_flag & OCTEON_MBOX_INTR) {
+		octeon_write_csr64(
+		    oct, CN23XX_VF_SLI_PKT_MBOX_INT(0),
+		    (octeon_read_csr64(oct, CN23XX_VF_SLI_PKT_MBOX_INT(0)) |
+		     CN23XX_INTR_MBOX_ENB));
+	}
+}
+
+static void cn23xx_disable_vf_interrupt(struct octeon_device *oct, u8 intr_flag)
+{
+	u32 q_no;
+
+	if (intr_flag & OCTEON_OUTPUT_INTR) {
+		for (q_no = 0; q_no < oct->num_oqs; q_no++) {
+			/* Write all 1's in INT_LEVEL reg to disable PO_INT */
+			octeon_write_csr64(
+			    oct, CN23XX_VF_SLI_OQ_PKT_INT_LEVELS(q_no),
+			    0x3fffffffffffff);
+		}
+	}
+	if (intr_flag & OCTEON_INPUT_INTR) {
+		for (q_no = 0; q_no < oct->num_oqs; q_no++) {
+			octeon_write_csr64(
+			    oct, CN23XX_VF_SLI_IQ_INSTR_COUNT64(q_no),
+			    (octeon_read_csr64(
+				 oct, CN23XX_VF_SLI_IQ_INSTR_COUNT64(q_no)) &
+			     ~(CN23XX_INTR_CINT_ENB |
+			       CN23XX_PKT_IN_DONE_CNT_MASK)));
+		}
+	}
+
+	if (intr_flag & OCTEON_MBOX_INTR) {
+		octeon_write_csr64(
+		    oct, CN23XX_VF_SLI_PKT_MBOX_INT(0),
+		    (octeon_read_csr64(oct, CN23XX_VF_SLI_PKT_MBOX_INT(0)) &
+		     ~CN23XX_INTR_MBOX_ENB));
+	}
+}
+
+int cn23xx_setup_octeon_vf_device(struct octeon_device *oct)
+{
+	struct octeon_cn23xx_vf *cn23xx = (struct octeon_cn23xx_vf *)oct->chip;
+	u32 rings_per_vf;
+	u64 reg_val;
+
+	if (octeon_map_pci_barx(oct, 0, 0))
+		return 1;
+
+	/* INPUT_CONTROL[RPVF] gives the VF IOq count */
+	reg_val = octeon_read_csr64(oct, CN23XX_VF_SLI_IQ_PKT_CONTROL64(0));
+
+	oct->pf_num = (reg_val >> CN23XX_PKT_INPUT_CTL_PF_NUM_POS) &
+		      CN23XX_PKT_INPUT_CTL_PF_NUM_MASK;
+	oct->vf_num = (reg_val >> CN23XX_PKT_INPUT_CTL_VF_NUM_POS) &
+		      CN23XX_PKT_INPUT_CTL_VF_NUM_MASK;
+
+	reg_val = reg_val >> CN23XX_PKT_INPUT_CTL_RPVF_POS;
+
+	rings_per_vf = reg_val & CN23XX_PKT_INPUT_CTL_RPVF_MASK;
+
+	cn23xx->conf  = oct_get_config_info(oct, LIO_23XX);
+	if (!cn23xx->conf) {
+		dev_err(&oct->pci_dev->dev, "%s No Config found for CN23XX\n",
+			__func__);
+		octeon_unmap_pci_barx(oct, 0);
+		return 1;
+	}
+
+	if (oct->sriov_info.rings_per_vf > rings_per_vf) {
+		dev_warn(&oct->pci_dev->dev,
+			 "num_queues:%d greater than PF configured rings_per_vf:%d. Reducing to %d.\n",
+			 oct->sriov_info.rings_per_vf, rings_per_vf,
+			 rings_per_vf);
+		oct->sriov_info.rings_per_vf = rings_per_vf;
+	} else {
+		if (rings_per_vf > num_present_cpus()) {
+			dev_warn(&oct->pci_dev->dev,
+				 "PF configured rings_per_vf:%d greater than num_cpu:%d. Using rings_per_vf:%d equal to num cpus\n",
+				 rings_per_vf,
+				 num_present_cpus(),
+				 num_present_cpus());
+			oct->sriov_info.rings_per_vf =
+				num_present_cpus();
+		} else {
+			oct->sriov_info.rings_per_vf = rings_per_vf;
+		}
+	}
+
+	oct->fn_list.setup_iq_regs = cn23xx_setup_vf_iq_regs;
+	oct->fn_list.setup_oq_regs = cn23xx_setup_vf_oq_regs;
+	oct->fn_list.setup_mbox = cn23xx_setup_vf_mbox;
+	oct->fn_list.free_mbox = cn23xx_free_vf_mbox;
+
+	oct->fn_list.msix_interrupt_handler = cn23xx_vf_msix_interrupt_handler;
+
+	oct->fn_list.setup_device_regs = cn23xx_setup_vf_device_regs;
+	oct->fn_list.update_iq_read_idx = cn23xx_update_read_index;
+
+	oct->fn_list.enable_interrupt = cn23xx_enable_vf_interrupt;
+	oct->fn_list.disable_interrupt = cn23xx_disable_vf_interrupt;
+
+	oct->fn_list.enable_io_queues = cn23xx_enable_vf_io_queues;
+	oct->fn_list.disable_io_queues = cn23xx_disable_vf_io_queues;
+
+	return 0;
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/cn23xx_vf_device.h b/drivers/net/ethernet/cavium/liquidio/cn23xx_vf_device.h
new file mode 100644
index 000000000..2d06097d3
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn23xx_vf_device.h
@@ -0,0 +1,48 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*! \file  cn23xx_device.h
+ * \brief Host Driver: Routines that perform CN23XX specific operations.
+ */
+
+#ifndef __CN23XX_VF_DEVICE_H__
+#define __CN23XX_VF_DEVICE_H__
+
+#include "cn23xx_vf_regs.h"
+
+/* Register address and configuration for a CN23XX devices.
+ * If device specific changes need to be made then add a struct to include
+ * device specific fields as shown in the commented section
+ */
+struct octeon_cn23xx_vf {
+	struct octeon_config *conf;
+};
+
+#define BUSY_READING_REG_VF_LOOP_COUNT		10000
+
+#define CN23XX_MAILBOX_MSGPARAM_SIZE		6
+
+void cn23xx_vf_ask_pf_to_do_flr(struct octeon_device *oct);
+
+int cn23xx_octeon_pfvf_handshake(struct octeon_device *oct);
+
+int cn23xx_setup_octeon_vf_device(struct octeon_device *oct);
+
+u32 cn23xx_vf_get_oq_ticks(struct octeon_device *oct, u32 time_intr_in_us);
+
+void cn23xx_dump_vf_initialized_regs(struct octeon_device *oct);
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/cn23xx_vf_regs.h b/drivers/net/ethernet/cavium/liquidio/cn23xx_vf_regs.h
new file mode 100644
index 000000000..e95610941
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn23xx_vf_regs.h
@@ -0,0 +1,274 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*! \file cn23xx_vf_regs.h
+ * \brief Host Driver: Register Address and Register Mask values for
+ * Octeon CN23XX vf functions.
+ */
+
+#ifndef __CN23XX_VF_REGS_H__
+#define __CN23XX_VF_REGS_H__
+
+#define     CN23XX_CONFIG_XPANSION_BAR             0x38
+
+#define     CN23XX_CONFIG_PCIE_CAP                 0x70
+#define     CN23XX_CONFIG_PCIE_DEVCAP              0x74
+#define     CN23XX_CONFIG_PCIE_DEVCTL              0x78
+#define     CN23XX_CONFIG_PCIE_LINKCAP             0x7C
+#define     CN23XX_CONFIG_PCIE_LINKCTL             0x80
+#define     CN23XX_CONFIG_PCIE_SLOTCAP             0x84
+#define     CN23XX_CONFIG_PCIE_SLOTCTL             0x88
+
+#define     CN23XX_CONFIG_PCIE_FLTMSK              0x720
+
+/* The input jabber is used to determine the TSO max size.
+ * Due to H/W limitation, this needs to be reduced to 60000
+ * in order to use H/W TSO and avoid the WQE malformation
+ * PKO_BUG_24989_WQE_LEN
+ */
+#define    CN23XX_DEFAULT_INPUT_JABBER             0xEA60 /*60000*/
+
+/* ##############  BAR0 Registers ################ */
+
+/* Each Input Queue register is at a 16-byte Offset in BAR0 */
+#define    CN23XX_VF_IQ_OFFSET                     0x20000
+
+/*###################### REQUEST QUEUE #########################*/
+
+/* 64 registers for Input Queue Instr Count - SLI_PKT_IN_DONE0_CNTS */
+#define    CN23XX_VF_SLI_IQ_INSTR_COUNT_START64     0x10040
+
+/* 64 registers for Input Queues Start Addr - SLI_PKT0_INSTR_BADDR */
+#define    CN23XX_VF_SLI_IQ_BASE_ADDR_START64       0x10010
+
+/* 64 registers for Input Doorbell - SLI_PKT0_INSTR_BAOFF_DBELL */
+#define    CN23XX_VF_SLI_IQ_DOORBELL_START          0x10020
+
+/* 64 registers for Input Queue size - SLI_PKT0_INSTR_FIFO_RSIZE */
+#define    CN23XX_VF_SLI_IQ_SIZE_START              0x10030
+
+/* 64 registers (64-bit) - ES, RO, NS, Arbitration for Input Queue Data &
+ * gather list fetches. SLI_PKT(0..63)_INPUT_CONTROL.
+ */
+#define    CN23XX_VF_SLI_IQ_PKT_CONTROL_START64     0x10000
+
+/*------- Request Queue Macros ---------*/
+#define CN23XX_VF_SLI_IQ_PKT_CONTROL64(iq)		\
+	(CN23XX_VF_SLI_IQ_PKT_CONTROL_START64 + ((iq) * CN23XX_VF_IQ_OFFSET))
+
+#define CN23XX_VF_SLI_IQ_BASE_ADDR64(iq)		\
+	(CN23XX_VF_SLI_IQ_BASE_ADDR_START64 + ((iq) * CN23XX_VF_IQ_OFFSET))
+
+#define CN23XX_VF_SLI_IQ_SIZE(iq)			\
+	(CN23XX_VF_SLI_IQ_SIZE_START + ((iq) * CN23XX_VF_IQ_OFFSET))
+
+#define CN23XX_VF_SLI_IQ_DOORBELL(iq)			\
+	(CN23XX_VF_SLI_IQ_DOORBELL_START + ((iq) * CN23XX_VF_IQ_OFFSET))
+
+#define CN23XX_VF_SLI_IQ_INSTR_COUNT64(iq)		\
+	(CN23XX_VF_SLI_IQ_INSTR_COUNT_START64 + ((iq) * CN23XX_VF_IQ_OFFSET))
+
+/*------------------ Masks ----------------*/
+#define    CN23XX_PKT_INPUT_CTL_VF_NUM                  BIT_ULL(32)
+#define    CN23XX_PKT_INPUT_CTL_MAC_NUM                 BIT(29)
+/* Number of instructions to be read in one MAC read request.
+ * setting to Max value(4)
+ */
+#define    CN23XX_PKT_INPUT_CTL_RDSIZE                  (3 << 25)
+#define    CN23XX_PKT_INPUT_CTL_IS_64B                  BIT(24)
+#define    CN23XX_PKT_INPUT_CTL_RST                     BIT(23)
+#define    CN23XX_PKT_INPUT_CTL_QUIET                   BIT(28)
+#define    CN23XX_PKT_INPUT_CTL_RING_ENB                BIT(22)
+#define    CN23XX_PKT_INPUT_CTL_DATA_NS                 BIT(8)
+#define    CN23XX_PKT_INPUT_CTL_DATA_ES_64B_SWAP        BIT(6)
+#define    CN23XX_PKT_INPUT_CTL_DATA_RO                 BIT(5)
+#define    CN23XX_PKT_INPUT_CTL_USE_CSR                 BIT(4)
+#define    CN23XX_PKT_INPUT_CTL_GATHER_NS               BIT(3)
+#define    CN23XX_PKT_INPUT_CTL_GATHER_ES_64B_SWAP      (2)
+#define    CN23XX_PKT_INPUT_CTL_GATHER_RO               (1)
+
+/** Rings per Virtual Function [RO] **/
+#define    CN23XX_PKT_INPUT_CTL_RPVF_MASK               (0x3F)
+#define    CN23XX_PKT_INPUT_CTL_RPVF_POS                (48)
+/* These bits[47:44][RO] give the Physical function number info within the MAC*/
+#define    CN23XX_PKT_INPUT_CTL_PF_NUM_MASK             (0x7)
+#define    CN23XX_PKT_INPUT_CTL_PF_NUM_POS              (45)
+/** These bits[43:32][RO] give the virtual function number info within the PF*/
+#define    CN23XX_PKT_INPUT_CTL_VF_NUM_MASK             (0x1FFF)
+#define    CN23XX_PKT_INPUT_CTL_VF_NUM_POS              (32)
+#define    CN23XX_PKT_INPUT_CTL_MAC_NUM_MASK            (0x3)
+#define    CN23XX_PKT_INPUT_CTL_MAC_NUM_POS             (29)
+#define    CN23XX_PKT_IN_DONE_WMARK_MASK                (0xFFFFULL)
+#define    CN23XX_PKT_IN_DONE_WMARK_BIT_POS             (32)
+#define    CN23XX_PKT_IN_DONE_CNT_MASK                  (0x00000000FFFFFFFFULL)
+
+#ifdef __LITTLE_ENDIAN_BITFIELD
+#define CN23XX_PKT_INPUT_CTL_MASK			\
+	(CN23XX_PKT_INPUT_CTL_RDSIZE			\
+	 | CN23XX_PKT_INPUT_CTL_DATA_ES_64B_SWAP	\
+	 | CN23XX_PKT_INPUT_CTL_USE_CSR)
+#else
+#define CN23XX_PKT_INPUT_CTL_MASK			\
+	(CN23XX_PKT_INPUT_CTL_RDSIZE			\
+	 | CN23XX_PKT_INPUT_CTL_DATA_ES_64B_SWAP	\
+	 | CN23XX_PKT_INPUT_CTL_USE_CSR			\
+	 | CN23XX_PKT_INPUT_CTL_GATHER_ES_64B_SWAP)
+#endif
+
+/** Masks for SLI_PKT_IN_DONE(0..63)_CNTS Register */
+#define    CN23XX_IN_DONE_CNTS_PI_INT               BIT_ULL(62)
+#define    CN23XX_IN_DONE_CNTS_CINT_ENB             BIT_ULL(48)
+
+/*############################ OUTPUT QUEUE #########################*/
+
+/* 64 registers for Output queue control - SLI_PKT(0..63)_OUTPUT_CONTROL */
+#define    CN23XX_VF_SLI_OQ_PKT_CONTROL_START       0x10050
+
+/* 64 registers for Output queue buffer and info size - SLI_PKT0_OUT_SIZE */
+#define    CN23XX_VF_SLI_OQ0_BUFF_INFO_SIZE         0x10060
+
+/* 64 registers for Output Queue Start Addr - SLI_PKT0_SLIST_BADDR */
+#define    CN23XX_VF_SLI_OQ_BASE_ADDR_START64       0x10070
+
+/* 64 registers for Output Queue Packet Credits - SLI_PKT0_SLIST_BAOFF_DBELL */
+#define    CN23XX_VF_SLI_OQ_PKT_CREDITS_START       0x10080
+
+/* 64 registers for Output Queue size - SLI_PKT0_SLIST_FIFO_RSIZE */
+#define    CN23XX_VF_SLI_OQ_SIZE_START              0x10090
+
+/* 64 registers for Output Queue Packet Count - SLI_PKT0_CNTS */
+#define    CN23XX_VF_SLI_OQ_PKT_SENT_START          0x100B0
+
+/* 64 registers for Output Queue INT Levels - SLI_PKT0_INT_LEVELS */
+#define    CN23XX_VF_SLI_OQ_PKT_INT_LEVELS_START64  0x100A0
+
+/* Each Output Queue register is at a 16-byte Offset in BAR0 */
+#define    CN23XX_VF_OQ_OFFSET                      0x20000
+
+/*------- Output Queue Macros ---------*/
+
+#define CN23XX_VF_SLI_OQ_PKT_CONTROL(oq)		\
+	(CN23XX_VF_SLI_OQ_PKT_CONTROL_START + ((oq) * CN23XX_VF_OQ_OFFSET))
+
+#define CN23XX_VF_SLI_OQ_BASE_ADDR64(oq)		\
+	(CN23XX_VF_SLI_OQ_BASE_ADDR_START64 + ((oq) * CN23XX_VF_OQ_OFFSET))
+
+#define CN23XX_VF_SLI_OQ_SIZE(oq)			\
+	(CN23XX_VF_SLI_OQ_SIZE_START + ((oq) * CN23XX_VF_OQ_OFFSET))
+
+#define CN23XX_VF_SLI_OQ_BUFF_INFO_SIZE(oq)		\
+	(CN23XX_VF_SLI_OQ0_BUFF_INFO_SIZE + ((oq) * CN23XX_VF_OQ_OFFSET))
+
+#define CN23XX_VF_SLI_OQ_PKTS_SENT(oq)		\
+	(CN23XX_VF_SLI_OQ_PKT_SENT_START + ((oq) * CN23XX_VF_OQ_OFFSET))
+
+#define CN23XX_VF_SLI_OQ_PKTS_CREDIT(oq)		\
+	(CN23XX_VF_SLI_OQ_PKT_CREDITS_START + ((oq) * CN23XX_VF_OQ_OFFSET))
+
+#define CN23XX_VF_SLI_OQ_PKT_INT_LEVELS(oq)		\
+	(CN23XX_VF_SLI_OQ_PKT_INT_LEVELS_START64 + ((oq) * CN23XX_VF_OQ_OFFSET))
+
+/* Macro's for accessing CNT and TIME separately from INT_LEVELS */
+#define CN23XX_VF_SLI_OQ_PKT_INT_LEVELS_CNT(oq)	\
+	(CN23XX_VF_SLI_OQ_PKT_INT_LEVELS_START64 + ((oq) * CN23XX_VF_OQ_OFFSET))
+
+#define CN23XX_VF_SLI_OQ_PKT_INT_LEVELS_TIME(oq)	\
+	(CN23XX_VF_SLI_OQ_PKT_INT_LEVELS_START64 +	\
+	 ((oq) * CN23XX_VF_OQ_OFFSET) + 4)
+
+/*------------------ Masks ----------------*/
+#define    CN23XX_PKT_OUTPUT_CTL_TENB                  BIT(13)
+#define    CN23XX_PKT_OUTPUT_CTL_CENB                  BIT(12)
+#define    CN23XX_PKT_OUTPUT_CTL_IPTR                  BIT(11)
+#define    CN23XX_PKT_OUTPUT_CTL_ES                    BIT(9)
+#define    CN23XX_PKT_OUTPUT_CTL_NSR                   BIT(8)
+#define    CN23XX_PKT_OUTPUT_CTL_ROR                   BIT(7)
+#define    CN23XX_PKT_OUTPUT_CTL_DPTR                  BIT(6)
+#define    CN23XX_PKT_OUTPUT_CTL_BMODE                 BIT(5)
+#define    CN23XX_PKT_OUTPUT_CTL_ES_P                  BIT(3)
+#define    CN23XX_PKT_OUTPUT_CTL_NSR_P                 BIT(2)
+#define    CN23XX_PKT_OUTPUT_CTL_ROR_P                 BIT(1)
+#define    CN23XX_PKT_OUTPUT_CTL_RING_ENB              BIT(0)
+
+/*######################### Mailbox Reg Macros ########################*/
+#define    CN23XX_VF_SLI_PKT_MBOX_INT_START            0x10210
+#define    CN23XX_SLI_PKT_PF_VF_MBOX_SIG_START         0x10200
+
+#define    CN23XX_SLI_MBOX_OFFSET                      0x20000
+#define    CN23XX_SLI_MBOX_SIG_IDX_OFFSET              0x8
+
+#define CN23XX_VF_SLI_PKT_MBOX_INT(q)	\
+	(CN23XX_VF_SLI_PKT_MBOX_INT_START + ((q) * CN23XX_SLI_MBOX_OFFSET))
+
+#define CN23XX_SLI_PKT_PF_VF_MBOX_SIG(q, idx)		\
+	(CN23XX_SLI_PKT_PF_VF_MBOX_SIG_START +		\
+	 ((q) * CN23XX_SLI_MBOX_OFFSET +		\
+	  (idx) * CN23XX_SLI_MBOX_SIG_IDX_OFFSET))
+
+/*######################## INTERRUPTS #########################*/
+
+#define    CN23XX_VF_SLI_INT_SUM_START		  0x100D0
+
+#define CN23XX_VF_SLI_INT_SUM(q)			\
+	(CN23XX_VF_SLI_INT_SUM_START + ((q) * CN23XX_VF_IQ_OFFSET))
+
+/*------------------ Interrupt Masks ----------------*/
+
+#define    CN23XX_INTR_PO_INT                   BIT_ULL(63)
+#define    CN23XX_INTR_PI_INT                   BIT_ULL(62)
+#define    CN23XX_INTR_MBOX_INT                 BIT_ULL(61)
+#define    CN23XX_INTR_RESEND                   BIT_ULL(60)
+
+#define    CN23XX_INTR_CINT_ENB                 BIT_ULL(48)
+#define    CN23XX_INTR_MBOX_ENB                 BIT(0)
+
+/*############################ MIO #########################*/
+#define    CN23XX_MIO_PTP_CLOCK_CFG       0x0001070000000f00ULL
+#define    CN23XX_MIO_PTP_CLOCK_LO        0x0001070000000f08ULL
+#define    CN23XX_MIO_PTP_CLOCK_HI        0x0001070000000f10ULL
+#define    CN23XX_MIO_PTP_CLOCK_COMP      0x0001070000000f18ULL
+#define    CN23XX_MIO_PTP_TIMESTAMP       0x0001070000000f20ULL
+#define    CN23XX_MIO_PTP_EVT_CNT         0x0001070000000f28ULL
+#define    CN23XX_MIO_PTP_CKOUT_THRESH_LO 0x0001070000000f30ULL
+#define    CN23XX_MIO_PTP_CKOUT_THRESH_HI 0x0001070000000f38ULL
+#define    CN23XX_MIO_PTP_CKOUT_HI_INCR   0x0001070000000f40ULL
+#define    CN23XX_MIO_PTP_CKOUT_LO_INCR   0x0001070000000f48ULL
+#define    CN23XX_MIO_PTP_PPS_THRESH_LO   0x0001070000000f50ULL
+#define    CN23XX_MIO_PTP_PPS_THRESH_HI   0x0001070000000f58ULL
+#define    CN23XX_MIO_PTP_PPS_HI_INCR     0x0001070000000f60ULL
+#define    CN23XX_MIO_PTP_PPS_LO_INCR     0x0001070000000f68ULL
+
+/*############################ RST #########################*/
+#define    CN23XX_RST_BOOT                0x0001180006001600ULL
+
+/*######################## MSIX TABLE #########################*/
+
+#define    CN23XX_MSIX_TABLE_ADDR_START    0x0
+#define    CN23XX_MSIX_TABLE_DATA_START    0x8
+
+#define    CN23XX_MSIX_TABLE_SIZE          0x10
+#define    CN23XX_MSIX_TABLE_ENTRIES       0x41
+
+#define    CN23XX_MSIX_ENTRY_VECTOR_CTL    BIT_ULL(32)
+
+#define CN23XX_MSIX_TABLE_ADDR(idx)		\
+	(CN23XX_MSIX_TABLE_ADDR_START + ((idx) * CN23XX_MSIX_TABLE_SIZE))
+
+#define CN23XX_MSIX_TABLE_DATA(idx)		\
+	(CN23XX_MSIX_TABLE_DATA_START + ((idx) * CN23XX_MSIX_TABLE_SIZE))
+
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/cn66xx_device.c b/drivers/net/ethernet/cavium/liquidio/cn66xx_device.c
new file mode 100644
index 000000000..39643be8c
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn66xx_device.c
@@ -0,0 +1,737 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_main.h"
+#include "cn66xx_regs.h"
+#include "cn66xx_device.h"
+
+int lio_cn6xxx_soft_reset(struct octeon_device *oct)
+{
+	octeon_write_csr64(oct, CN6XXX_WIN_WR_MASK_REG, 0xFF);
+
+	dev_dbg(&oct->pci_dev->dev, "BIST enabled for soft reset\n");
+
+	lio_pci_writeq(oct, 1, CN6XXX_CIU_SOFT_BIST);
+	octeon_write_csr64(oct, CN6XXX_SLI_SCRATCH1, 0x1234ULL);
+
+	lio_pci_readq(oct, CN6XXX_CIU_SOFT_RST);
+	lio_pci_writeq(oct, 1, CN6XXX_CIU_SOFT_RST);
+
+	/* Wait for 10ms as Octeon resets. */
+	mdelay(100);
+
+	if (octeon_read_csr64(oct, CN6XXX_SLI_SCRATCH1)) {
+		dev_err(&oct->pci_dev->dev, "Soft reset failed\n");
+		return 1;
+	}
+
+	dev_dbg(&oct->pci_dev->dev, "Reset completed\n");
+	octeon_write_csr64(oct, CN6XXX_WIN_WR_MASK_REG, 0xFF);
+
+	return 0;
+}
+
+void lio_cn6xxx_enable_error_reporting(struct octeon_device *oct)
+{
+	u32 val;
+
+	pci_read_config_dword(oct->pci_dev, CN6XXX_PCIE_DEVCTL, &val);
+	if (val & 0x000c0000) {
+		dev_err(&oct->pci_dev->dev, "PCI-E Link error detected: 0x%08x\n",
+			val & 0x000c0000);
+	}
+
+	val |= 0xf;          /* Enable Link error reporting */
+
+	dev_dbg(&oct->pci_dev->dev, "Enabling PCI-E error reporting..\n");
+	pci_write_config_dword(oct->pci_dev, CN6XXX_PCIE_DEVCTL, val);
+}
+
+void lio_cn6xxx_setup_pcie_mps(struct octeon_device *oct,
+			       enum octeon_pcie_mps mps)
+{
+	u32 val;
+	u64 r64;
+
+	/* Read config register for MPS */
+	pci_read_config_dword(oct->pci_dev, CN6XXX_PCIE_DEVCTL, &val);
+
+	if (mps == PCIE_MPS_DEFAULT) {
+		mps = ((val & (0x7 << 5)) >> 5);
+	} else {
+		val &= ~(0x7 << 5);  /* Turn off any MPS bits */
+		val |= (mps << 5);   /* Set MPS */
+		pci_write_config_dword(oct->pci_dev, CN6XXX_PCIE_DEVCTL, val);
+	}
+
+	/* Set MPS in DPI_SLI_PRT0_CFG to the same value. */
+	r64 = lio_pci_readq(oct, CN6XXX_DPI_SLI_PRTX_CFG(oct->pcie_port));
+	r64 |= (mps << 4);
+	lio_pci_writeq(oct, r64, CN6XXX_DPI_SLI_PRTX_CFG(oct->pcie_port));
+}
+
+void lio_cn6xxx_setup_pcie_mrrs(struct octeon_device *oct,
+				enum octeon_pcie_mrrs mrrs)
+{
+	u32 val;
+	u64 r64;
+
+	/* Read config register for MRRS */
+	pci_read_config_dword(oct->pci_dev, CN6XXX_PCIE_DEVCTL, &val);
+
+	if (mrrs == PCIE_MRRS_DEFAULT) {
+		mrrs = ((val & (0x7 << 12)) >> 12);
+	} else {
+		val &= ~(0x7 << 12); /* Turn off any MRRS bits */
+		val |= (mrrs << 12); /* Set MRRS */
+		pci_write_config_dword(oct->pci_dev, CN6XXX_PCIE_DEVCTL, val);
+	}
+
+	/* Set MRRS in SLI_S2M_PORT0_CTL to the same value. */
+	r64 = octeon_read_csr64(oct, CN6XXX_SLI_S2M_PORTX_CTL(oct->pcie_port));
+	r64 |= mrrs;
+	octeon_write_csr64(oct, CN6XXX_SLI_S2M_PORTX_CTL(oct->pcie_port), r64);
+
+	/* Set MRRS in DPI_SLI_PRT0_CFG to the same value. */
+	r64 = lio_pci_readq(oct, CN6XXX_DPI_SLI_PRTX_CFG(oct->pcie_port));
+	r64 |= mrrs;
+	lio_pci_writeq(oct, r64, CN6XXX_DPI_SLI_PRTX_CFG(oct->pcie_port));
+}
+
+u32 lio_cn6xxx_coprocessor_clock(struct octeon_device *oct)
+{
+	/* Bits 29:24 of MIO_RST_BOOT holds the ref. clock multiplier
+	 * for SLI.
+	 */
+	return ((lio_pci_readq(oct, CN6XXX_MIO_RST_BOOT) >> 24) & 0x3f) * 50;
+}
+
+u32 lio_cn6xxx_get_oq_ticks(struct octeon_device *oct,
+			    u32 time_intr_in_us)
+{
+	/* This gives the SLI clock per microsec */
+	u32 oqticks_per_us = lio_cn6xxx_coprocessor_clock(oct);
+
+	/* core clock per us / oq ticks will be fractional. TO avoid that
+	 * we use the method below.
+	 */
+
+	/* This gives the clock cycles per millisecond */
+	oqticks_per_us *= 1000;
+
+	/* This gives the oq ticks (1024 core clock cycles) per millisecond */
+	oqticks_per_us /= 1024;
+
+	/* time_intr is in microseconds. The next 2 steps gives the oq ticks
+	 * corressponding to time_intr.
+	 */
+	oqticks_per_us *= time_intr_in_us;
+	oqticks_per_us /= 1000;
+
+	return oqticks_per_us;
+}
+
+void lio_cn6xxx_setup_global_input_regs(struct octeon_device *oct)
+{
+	/* Select Round-Robin Arb, ES, RO, NS for Input Queues */
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_INPUT_CONTROL,
+			 CN6XXX_INPUT_CTL_MASK);
+
+	/* Instruction Read Size - Max 4 instructions per PCIE Read */
+	octeon_write_csr64(oct, CN6XXX_SLI_PKT_INSTR_RD_SIZE,
+			   0xFFFFFFFFFFFFFFFFULL);
+
+	/* Select PCIE Port for all Input rings. */
+	octeon_write_csr64(oct, CN6XXX_SLI_IN_PCIE_PORT,
+			   (oct->pcie_port * 0x5555555555555555ULL));
+}
+
+static void lio_cn66xx_setup_pkt_ctl_regs(struct octeon_device *oct)
+{
+	u64 pktctl;
+
+	struct octeon_cn6xxx *cn6xxx = (struct octeon_cn6xxx *)oct->chip;
+
+	pktctl = octeon_read_csr64(oct, CN6XXX_SLI_PKT_CTL);
+
+	/* 66XX SPECIFIC */
+	if (CFG_GET_OQ_MAX_Q(cn6xxx->conf) <= 4)
+		/* Disable RING_EN if only upto 4 rings are used. */
+		pktctl &= ~(1 << 4);
+	else
+		pktctl |= (1 << 4);
+
+	if (CFG_GET_IS_SLI_BP_ON(cn6xxx->conf))
+		pktctl |= 0xF;
+	else
+		/* Disable per-port backpressure. */
+		pktctl &= ~0xF;
+	octeon_write_csr64(oct, CN6XXX_SLI_PKT_CTL, pktctl);
+}
+
+void lio_cn6xxx_setup_global_output_regs(struct octeon_device *oct)
+{
+	u32 time_threshold;
+	struct octeon_cn6xxx *cn6xxx = (struct octeon_cn6xxx *)oct->chip;
+
+	/* / Select PCI-E Port for all Output queues */
+	octeon_write_csr64(oct, CN6XXX_SLI_PKT_PCIE_PORT64,
+			   (oct->pcie_port * 0x5555555555555555ULL));
+
+	if (CFG_GET_IS_SLI_BP_ON(cn6xxx->conf)) {
+		octeon_write_csr64(oct, CN6XXX_SLI_OQ_WMARK, 32);
+	} else {
+		/* / Set Output queue watermark to 0 to disable backpressure */
+		octeon_write_csr64(oct, CN6XXX_SLI_OQ_WMARK, 0);
+	}
+
+	/* / Select Packet count instead of bytes for SLI_PKTi_CNTS[CNT] */
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_OUT_BMODE, 0);
+
+	/* Select ES, RO, NS setting from register for Output Queue Packet
+	 * Address
+	 */
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_DPADDR, 0xFFFFFFFF);
+
+	/* No Relaxed Ordering, No Snoop, 64-bit swap for Output
+	 * Queue ScatterList
+	 */
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_SLIST_ROR, 0);
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_SLIST_NS, 0);
+
+	/* / ENDIAN_SPECIFIC CHANGES - 0 works for LE. */
+#ifdef __BIG_ENDIAN_BITFIELD
+	octeon_write_csr64(oct, CN6XXX_SLI_PKT_SLIST_ES64,
+			   0x5555555555555555ULL);
+#else
+	octeon_write_csr64(oct, CN6XXX_SLI_PKT_SLIST_ES64, 0ULL);
+#endif
+
+	/* / No Relaxed Ordering, No Snoop, 64-bit swap for Output Queue Data */
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_DATA_OUT_ROR, 0);
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_DATA_OUT_NS, 0);
+	octeon_write_csr64(oct, CN6XXX_SLI_PKT_DATA_OUT_ES64,
+			   0x5555555555555555ULL);
+
+	/* / Set up interrupt packet and time threshold */
+	octeon_write_csr(oct, CN6XXX_SLI_OQ_INT_LEVEL_PKTS,
+			 (u32)CFG_GET_OQ_INTR_PKT(cn6xxx->conf));
+	time_threshold =
+		lio_cn6xxx_get_oq_ticks(oct, (u32)
+					CFG_GET_OQ_INTR_TIME(cn6xxx->conf));
+
+	octeon_write_csr(oct, CN6XXX_SLI_OQ_INT_LEVEL_TIME, time_threshold);
+}
+
+static int lio_cn6xxx_setup_device_regs(struct octeon_device *oct)
+{
+	lio_cn6xxx_setup_pcie_mps(oct, PCIE_MPS_DEFAULT);
+	lio_cn6xxx_setup_pcie_mrrs(oct, PCIE_MRRS_512B);
+	lio_cn6xxx_enable_error_reporting(oct);
+
+	lio_cn6xxx_setup_global_input_regs(oct);
+	lio_cn66xx_setup_pkt_ctl_regs(oct);
+	lio_cn6xxx_setup_global_output_regs(oct);
+
+	/* Default error timeout value should be 0x200000 to avoid host hang
+	 * when reads invalid register
+	 */
+	octeon_write_csr64(oct, CN6XXX_SLI_WINDOW_CTL, 0x200000ULL);
+	return 0;
+}
+
+void lio_cn6xxx_setup_iq_regs(struct octeon_device *oct, u32 iq_no)
+{
+	struct octeon_instr_queue *iq = oct->instr_queue[iq_no];
+
+	octeon_write_csr64(oct, CN6XXX_SLI_IQ_PKT_INSTR_HDR64(iq_no), 0);
+
+	/* Write the start of the input queue's ring and its size  */
+	octeon_write_csr64(oct, CN6XXX_SLI_IQ_BASE_ADDR64(iq_no),
+			   iq->base_addr_dma);
+	octeon_write_csr(oct, CN6XXX_SLI_IQ_SIZE(iq_no), iq->max_count);
+
+	/* Remember the doorbell & instruction count register addr for this
+	 * queue
+	 */
+	iq->doorbell_reg = oct->mmio[0].hw_addr + CN6XXX_SLI_IQ_DOORBELL(iq_no);
+	iq->inst_cnt_reg = oct->mmio[0].hw_addr
+			   + CN6XXX_SLI_IQ_INSTR_COUNT(iq_no);
+	dev_dbg(&oct->pci_dev->dev, "InstQ[%d]:dbell reg @ 0x%p instcnt_reg @ 0x%p\n",
+		iq_no, iq->doorbell_reg, iq->inst_cnt_reg);
+
+	/* Store the current instruction counter
+	 * (used in flush_iq calculation)
+	 */
+	iq->reset_instr_cnt = readl(iq->inst_cnt_reg);
+}
+
+static void lio_cn66xx_setup_iq_regs(struct octeon_device *oct, u32 iq_no)
+{
+	lio_cn6xxx_setup_iq_regs(oct, iq_no);
+
+	/* Backpressure for this queue - WMARK set to all F's. This effectively
+	 * disables the backpressure mechanism.
+	 */
+	octeon_write_csr64(oct, CN66XX_SLI_IQ_BP64(iq_no),
+			   (0xFFFFFFFFULL << 32));
+}
+
+void lio_cn6xxx_setup_oq_regs(struct octeon_device *oct, u32 oq_no)
+{
+	u32 intr;
+	struct octeon_droq *droq = oct->droq[oq_no];
+
+	octeon_write_csr64(oct, CN6XXX_SLI_OQ_BASE_ADDR64(oq_no),
+			   droq->desc_ring_dma);
+	octeon_write_csr(oct, CN6XXX_SLI_OQ_SIZE(oq_no), droq->max_count);
+
+	octeon_write_csr(oct, CN6XXX_SLI_OQ_BUFF_INFO_SIZE(oq_no),
+			 droq->buffer_size);
+
+	/* Get the mapped address of the pkt_sent and pkts_credit regs */
+	droq->pkts_sent_reg =
+		oct->mmio[0].hw_addr + CN6XXX_SLI_OQ_PKTS_SENT(oq_no);
+	droq->pkts_credit_reg =
+		oct->mmio[0].hw_addr + CN6XXX_SLI_OQ_PKTS_CREDIT(oq_no);
+
+	/* Enable this output queue to generate Packet Timer Interrupt */
+	intr = octeon_read_csr(oct, CN6XXX_SLI_PKT_TIME_INT_ENB);
+	intr |= (1 << oq_no);
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_TIME_INT_ENB, intr);
+
+	/* Enable this output queue to generate Packet Timer Interrupt */
+	intr = octeon_read_csr(oct, CN6XXX_SLI_PKT_CNT_INT_ENB);
+	intr |= (1 << oq_no);
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_CNT_INT_ENB, intr);
+}
+
+int lio_cn6xxx_enable_io_queues(struct octeon_device *oct)
+{
+	u32 mask;
+
+	mask = octeon_read_csr(oct, CN6XXX_SLI_PKT_INSTR_SIZE);
+	mask |= oct->io_qmask.iq64B;
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_INSTR_SIZE, mask);
+
+	mask = octeon_read_csr(oct, CN6XXX_SLI_PKT_INSTR_ENB);
+	mask |= oct->io_qmask.iq;
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_INSTR_ENB, mask);
+
+	mask = octeon_read_csr(oct, CN6XXX_SLI_PKT_OUT_ENB);
+	mask |= oct->io_qmask.oq;
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_OUT_ENB, mask);
+
+	return 0;
+}
+
+void lio_cn6xxx_disable_io_queues(struct octeon_device *oct)
+{
+	int i;
+	u32 mask, loop = HZ;
+	u32 d32;
+
+	/* Reset the Enable bits for Input Queues. */
+	mask = octeon_read_csr(oct, CN6XXX_SLI_PKT_INSTR_ENB);
+	mask ^= oct->io_qmask.iq;
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_INSTR_ENB, mask);
+
+	/* Wait until hardware indicates that the queues are out of reset. */
+	mask = (u32)oct->io_qmask.iq;
+	d32 = octeon_read_csr(oct, CN6XXX_SLI_PORT_IN_RST_IQ);
+	while (((d32 & mask) != mask) && loop--) {
+		d32 = octeon_read_csr(oct, CN6XXX_SLI_PORT_IN_RST_IQ);
+		schedule_timeout_uninterruptible(1);
+	}
+
+	/* Reset the doorbell register for each Input queue. */
+	for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct); i++) {
+		if (!(oct->io_qmask.iq & BIT_ULL(i)))
+			continue;
+		octeon_write_csr(oct, CN6XXX_SLI_IQ_DOORBELL(i), 0xFFFFFFFF);
+		d32 = octeon_read_csr(oct, CN6XXX_SLI_IQ_DOORBELL(i));
+	}
+
+	/* Reset the Enable bits for Output Queues. */
+	mask = octeon_read_csr(oct, CN6XXX_SLI_PKT_OUT_ENB);
+	mask ^= oct->io_qmask.oq;
+	octeon_write_csr(oct, CN6XXX_SLI_PKT_OUT_ENB, mask);
+
+	/* Wait until hardware indicates that the queues are out of reset. */
+	loop = HZ;
+	mask = (u32)oct->io_qmask.oq;
+	d32 = octeon_read_csr(oct, CN6XXX_SLI_PORT_IN_RST_OQ);
+	while (((d32 & mask) != mask) && loop--) {
+		d32 = octeon_read_csr(oct, CN6XXX_SLI_PORT_IN_RST_OQ);
+		schedule_timeout_uninterruptible(1);
+	}
+	;
+
+	/* Reset the doorbell register for each Output queue. */
+	for (i = 0; i < MAX_OCTEON_OUTPUT_QUEUES(oct); i++) {
+		if (!(oct->io_qmask.oq & BIT_ULL(i)))
+			continue;
+		octeon_write_csr(oct, CN6XXX_SLI_OQ_PKTS_CREDIT(i), 0xFFFFFFFF);
+		d32 = octeon_read_csr(oct, CN6XXX_SLI_OQ_PKTS_CREDIT(i));
+
+		d32 = octeon_read_csr(oct, CN6XXX_SLI_OQ_PKTS_SENT(i));
+		octeon_write_csr(oct, CN6XXX_SLI_OQ_PKTS_SENT(i), d32);
+	}
+
+	d32 = octeon_read_csr(oct, CN6XXX_SLI_PKT_CNT_INT);
+	if (d32)
+		octeon_write_csr(oct, CN6XXX_SLI_PKT_CNT_INT, d32);
+
+	d32 = octeon_read_csr(oct, CN6XXX_SLI_PKT_TIME_INT);
+	if (d32)
+		octeon_write_csr(oct, CN6XXX_SLI_PKT_TIME_INT, d32);
+}
+
+void
+lio_cn6xxx_bar1_idx_setup(struct octeon_device *oct,
+			  u64 core_addr,
+			  u32 idx,
+			  int valid)
+{
+	u64 bar1;
+
+	if (valid == 0) {
+		bar1 = lio_pci_readq(oct, CN6XXX_BAR1_REG(idx, oct->pcie_port));
+		lio_pci_writeq(oct, (bar1 & 0xFFFFFFFEULL),
+			       CN6XXX_BAR1_REG(idx, oct->pcie_port));
+		bar1 = lio_pci_readq(oct, CN6XXX_BAR1_REG(idx, oct->pcie_port));
+		return;
+	}
+
+	/* Bits 17:4 of the PCI_BAR1_INDEXx stores bits 35:22 of
+	 * the Core Addr
+	 */
+	lio_pci_writeq(oct, (((core_addr >> 22) << 4) | PCI_BAR1_MASK),
+		       CN6XXX_BAR1_REG(idx, oct->pcie_port));
+
+	bar1 = lio_pci_readq(oct, CN6XXX_BAR1_REG(idx, oct->pcie_port));
+}
+
+void lio_cn6xxx_bar1_idx_write(struct octeon_device *oct,
+			       u32 idx,
+			       u32 mask)
+{
+	lio_pci_writeq(oct, mask, CN6XXX_BAR1_REG(idx, oct->pcie_port));
+}
+
+u32 lio_cn6xxx_bar1_idx_read(struct octeon_device *oct, u32 idx)
+{
+	return (u32)lio_pci_readq(oct, CN6XXX_BAR1_REG(idx, oct->pcie_port));
+}
+
+u32
+lio_cn6xxx_update_read_index(struct octeon_instr_queue *iq)
+{
+	u32 new_idx = readl(iq->inst_cnt_reg);
+
+	/* The new instr cnt reg is a 32-bit counter that can roll over. We have
+	 * noted the counter's initial value at init time into
+	 * reset_instr_cnt
+	 */
+	if (iq->reset_instr_cnt < new_idx)
+		new_idx -= iq->reset_instr_cnt;
+	else
+		new_idx += (0xffffffff - iq->reset_instr_cnt) + 1;
+
+	/* Modulo of the new index with the IQ size will give us
+	 * the new index.
+	 */
+	new_idx %= iq->max_count;
+
+	return new_idx;
+}
+
+void lio_cn6xxx_enable_interrupt(struct octeon_device *oct,
+				 u8 unused __attribute__((unused)))
+{
+	struct octeon_cn6xxx *cn6xxx = (struct octeon_cn6xxx *)oct->chip;
+	u64 mask = cn6xxx->intr_mask64 | CN6XXX_INTR_DMA0_FORCE;
+
+	/* Enable Interrupt */
+	writeq(mask, cn6xxx->intr_enb_reg64);
+}
+
+void lio_cn6xxx_disable_interrupt(struct octeon_device *oct,
+				  u8 unused __attribute__((unused)))
+{
+	struct octeon_cn6xxx *cn6xxx = (struct octeon_cn6xxx *)oct->chip;
+
+	/* Disable Interrupts */
+	writeq(0, cn6xxx->intr_enb_reg64);
+}
+
+static void lio_cn6xxx_get_pcie_qlmport(struct octeon_device *oct)
+{
+	/* CN63xx Pass2 and newer parts implements the SLI_MAC_NUMBER register
+	 * to determine the PCIE port #
+	 */
+	oct->pcie_port = octeon_read_csr(oct, CN6XXX_SLI_MAC_NUMBER) & 0xff;
+
+	dev_dbg(&oct->pci_dev->dev, "Using PCIE Port %d\n", oct->pcie_port);
+}
+
+static void
+lio_cn6xxx_process_pcie_error_intr(struct octeon_device *oct, u64 intr64)
+{
+	dev_err(&oct->pci_dev->dev, "Error Intr: 0x%016llx\n",
+		CVM_CAST64(intr64));
+}
+
+static int lio_cn6xxx_process_droq_intr_regs(struct octeon_device *oct)
+{
+	struct octeon_droq *droq;
+	int oq_no;
+	u32 pkt_count, droq_time_mask, droq_mask, droq_int_enb;
+	u32 droq_cnt_enb, droq_cnt_mask;
+
+	droq_cnt_enb = octeon_read_csr(oct, CN6XXX_SLI_PKT_CNT_INT_ENB);
+	droq_cnt_mask = octeon_read_csr(oct, CN6XXX_SLI_PKT_CNT_INT);
+	droq_mask = droq_cnt_mask & droq_cnt_enb;
+
+	droq_time_mask = octeon_read_csr(oct, CN6XXX_SLI_PKT_TIME_INT);
+	droq_int_enb = octeon_read_csr(oct, CN6XXX_SLI_PKT_TIME_INT_ENB);
+	droq_mask |= (droq_time_mask & droq_int_enb);
+
+	droq_mask &= oct->io_qmask.oq;
+
+	oct->droq_intr = 0;
+
+	for (oq_no = 0; oq_no < MAX_OCTEON_OUTPUT_QUEUES(oct); oq_no++) {
+		if (!(droq_mask & BIT_ULL(oq_no)))
+			continue;
+
+		droq = oct->droq[oq_no];
+		pkt_count = octeon_droq_check_hw_for_pkts(droq);
+		if (pkt_count) {
+			oct->droq_intr |= BIT_ULL(oq_no);
+			if (droq->ops.poll_mode) {
+				u32 value;
+				u32 reg;
+
+				struct octeon_cn6xxx *cn6xxx =
+					(struct octeon_cn6xxx *)oct->chip;
+
+				/* disable interrupts for this droq */
+				spin_lock
+					(&cn6xxx->lock_for_droq_int_enb_reg);
+				reg = CN6XXX_SLI_PKT_TIME_INT_ENB;
+				value = octeon_read_csr(oct, reg);
+				value &= ~(1 << oq_no);
+				octeon_write_csr(oct, reg, value);
+				reg = CN6XXX_SLI_PKT_CNT_INT_ENB;
+				value = octeon_read_csr(oct, reg);
+				value &= ~(1 << oq_no);
+				octeon_write_csr(oct, reg, value);
+
+				spin_unlock(&cn6xxx->lock_for_droq_int_enb_reg);
+			}
+		}
+	}
+
+	droq_time_mask &= oct->io_qmask.oq;
+	droq_cnt_mask &= oct->io_qmask.oq;
+
+	/* Reset the PKT_CNT/TIME_INT registers. */
+	if (droq_time_mask)
+		octeon_write_csr(oct, CN6XXX_SLI_PKT_TIME_INT, droq_time_mask);
+
+	if (droq_cnt_mask)      /* reset PKT_CNT register:66xx */
+		octeon_write_csr(oct, CN6XXX_SLI_PKT_CNT_INT, droq_cnt_mask);
+
+	return 0;
+}
+
+irqreturn_t lio_cn6xxx_process_interrupt_regs(void *dev)
+{
+	struct octeon_device *oct = (struct octeon_device *)dev;
+	struct octeon_cn6xxx *cn6xxx = (struct octeon_cn6xxx *)oct->chip;
+	u64 intr64;
+
+	intr64 = readq(cn6xxx->intr_sum_reg64);
+
+	/* If our device has interrupted, then proceed.
+	 * Also check for all f's if interrupt was triggered on an error
+	 * and the PCI read fails.
+	 */
+	if (!intr64 || (intr64 == 0xFFFFFFFFFFFFFFFFULL))
+		return IRQ_NONE;
+
+	oct->int_status = 0;
+
+	if (intr64 & CN6XXX_INTR_ERR)
+		lio_cn6xxx_process_pcie_error_intr(oct, intr64);
+
+	if (intr64 & CN6XXX_INTR_PKT_DATA) {
+		lio_cn6xxx_process_droq_intr_regs(oct);
+		oct->int_status |= OCT_DEV_INTR_PKT_DATA;
+	}
+
+	if (intr64 & CN6XXX_INTR_DMA0_FORCE)
+		oct->int_status |= OCT_DEV_INTR_DMA0_FORCE;
+
+	if (intr64 & CN6XXX_INTR_DMA1_FORCE)
+		oct->int_status |= OCT_DEV_INTR_DMA1_FORCE;
+
+	/* Clear the current interrupts */
+	writeq(intr64, cn6xxx->intr_sum_reg64);
+
+	return IRQ_HANDLED;
+}
+
+void lio_cn6xxx_setup_reg_address(struct octeon_device *oct,
+				  void *chip,
+				  struct octeon_reg_list *reg_list)
+{
+	u8 __iomem *bar0_pciaddr = oct->mmio[0].hw_addr;
+	struct octeon_cn6xxx *cn6xxx = (struct octeon_cn6xxx *)chip;
+
+	reg_list->pci_win_wr_addr_hi =
+		(u32 __iomem *)(bar0_pciaddr + CN6XXX_WIN_WR_ADDR_HI);
+	reg_list->pci_win_wr_addr_lo =
+		(u32 __iomem *)(bar0_pciaddr + CN6XXX_WIN_WR_ADDR_LO);
+	reg_list->pci_win_wr_addr =
+		(u64 __iomem *)(bar0_pciaddr + CN6XXX_WIN_WR_ADDR64);
+
+	reg_list->pci_win_rd_addr_hi =
+		(u32 __iomem *)(bar0_pciaddr + CN6XXX_WIN_RD_ADDR_HI);
+	reg_list->pci_win_rd_addr_lo =
+		(u32 __iomem *)(bar0_pciaddr + CN6XXX_WIN_RD_ADDR_LO);
+	reg_list->pci_win_rd_addr =
+		(u64 __iomem *)(bar0_pciaddr + CN6XXX_WIN_RD_ADDR64);
+
+	reg_list->pci_win_wr_data_hi =
+		(u32 __iomem *)(bar0_pciaddr + CN6XXX_WIN_WR_DATA_HI);
+	reg_list->pci_win_wr_data_lo =
+		(u32 __iomem *)(bar0_pciaddr + CN6XXX_WIN_WR_DATA_LO);
+	reg_list->pci_win_wr_data =
+		(u64 __iomem *)(bar0_pciaddr + CN6XXX_WIN_WR_DATA64);
+
+	reg_list->pci_win_rd_data_hi =
+		(u32 __iomem *)(bar0_pciaddr + CN6XXX_WIN_RD_DATA_HI);
+	reg_list->pci_win_rd_data_lo =
+		(u32 __iomem *)(bar0_pciaddr + CN6XXX_WIN_RD_DATA_LO);
+	reg_list->pci_win_rd_data =
+		(u64 __iomem *)(bar0_pciaddr + CN6XXX_WIN_RD_DATA64);
+
+	lio_cn6xxx_get_pcie_qlmport(oct);
+
+	cn6xxx->intr_sum_reg64 = bar0_pciaddr + CN6XXX_SLI_INT_SUM64;
+	cn6xxx->intr_mask64 = CN6XXX_INTR_MASK;
+	cn6xxx->intr_enb_reg64 =
+		bar0_pciaddr + CN6XXX_SLI_INT_ENB64(oct->pcie_port);
+}
+
+int lio_setup_cn66xx_octeon_device(struct octeon_device *oct)
+{
+	struct octeon_cn6xxx *cn6xxx = (struct octeon_cn6xxx *)oct->chip;
+
+	if (octeon_map_pci_barx(oct, 0, 0))
+		return 1;
+
+	if (octeon_map_pci_barx(oct, 1, MAX_BAR1_IOREMAP_SIZE)) {
+		dev_err(&oct->pci_dev->dev, "%s CN66XX BAR1 map failed\n",
+			__func__);
+		octeon_unmap_pci_barx(oct, 0);
+		return 1;
+	}
+
+	spin_lock_init(&cn6xxx->lock_for_droq_int_enb_reg);
+
+	oct->fn_list.setup_iq_regs = lio_cn66xx_setup_iq_regs;
+	oct->fn_list.setup_oq_regs = lio_cn6xxx_setup_oq_regs;
+
+	oct->fn_list.soft_reset = lio_cn6xxx_soft_reset;
+	oct->fn_list.setup_device_regs = lio_cn6xxx_setup_device_regs;
+	oct->fn_list.update_iq_read_idx = lio_cn6xxx_update_read_index;
+
+	oct->fn_list.bar1_idx_setup = lio_cn6xxx_bar1_idx_setup;
+	oct->fn_list.bar1_idx_write = lio_cn6xxx_bar1_idx_write;
+	oct->fn_list.bar1_idx_read = lio_cn6xxx_bar1_idx_read;
+
+	oct->fn_list.process_interrupt_regs = lio_cn6xxx_process_interrupt_regs;
+	oct->fn_list.enable_interrupt = lio_cn6xxx_enable_interrupt;
+	oct->fn_list.disable_interrupt = lio_cn6xxx_disable_interrupt;
+
+	oct->fn_list.enable_io_queues = lio_cn6xxx_enable_io_queues;
+	oct->fn_list.disable_io_queues = lio_cn6xxx_disable_io_queues;
+
+	lio_cn6xxx_setup_reg_address(oct, oct->chip, &oct->reg_list);
+
+	cn6xxx->conf = (struct octeon_config *)
+		       oct_get_config_info(oct, LIO_210SV);
+	if (!cn6xxx->conf) {
+		dev_err(&oct->pci_dev->dev, "%s No Config found for CN66XX\n",
+			__func__);
+		octeon_unmap_pci_barx(oct, 0);
+		octeon_unmap_pci_barx(oct, 1);
+		return 1;
+	}
+
+	oct->coproc_clock_rate = 1000000ULL * lio_cn6xxx_coprocessor_clock(oct);
+
+	return 0;
+}
+
+int lio_validate_cn6xxx_config_info(struct octeon_device *oct,
+				    struct octeon_config *conf6xxx)
+{
+	if (CFG_GET_IQ_MAX_Q(conf6xxx) > CN6XXX_MAX_INPUT_QUEUES) {
+		dev_err(&oct->pci_dev->dev, "%s: Num IQ (%d) exceeds Max (%d)\n",
+			__func__, CFG_GET_IQ_MAX_Q(conf6xxx),
+			CN6XXX_MAX_INPUT_QUEUES);
+		return 1;
+	}
+
+	if (CFG_GET_OQ_MAX_Q(conf6xxx) > CN6XXX_MAX_OUTPUT_QUEUES) {
+		dev_err(&oct->pci_dev->dev, "%s: Num OQ (%d) exceeds Max (%d)\n",
+			__func__, CFG_GET_OQ_MAX_Q(conf6xxx),
+			CN6XXX_MAX_OUTPUT_QUEUES);
+		return 1;
+	}
+
+	if (CFG_GET_IQ_INSTR_TYPE(conf6xxx) != OCTEON_32BYTE_INSTR &&
+	    CFG_GET_IQ_INSTR_TYPE(conf6xxx) != OCTEON_64BYTE_INSTR) {
+		dev_err(&oct->pci_dev->dev, "%s: Invalid instr type for IQ\n",
+			__func__);
+		return 1;
+	}
+	if (!CFG_GET_OQ_REFILL_THRESHOLD(conf6xxx)) {
+		dev_err(&oct->pci_dev->dev, "%s: Invalid parameter for OQ\n",
+			__func__);
+		return 1;
+	}
+
+	if (!(CFG_GET_OQ_INTR_TIME(conf6xxx))) {
+		dev_err(&oct->pci_dev->dev, "%s: No Time Interrupt for OQ\n",
+			__func__);
+		return 1;
+	}
+
+	return 0;
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/cn66xx_device.h b/drivers/net/ethernet/cavium/liquidio/cn66xx_device.h
new file mode 100644
index 000000000..8ed57134e
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn66xx_device.h
@@ -0,0 +1,98 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*! \file  cn66xx_device.h
+ *  \brief Host Driver: Routines that perform CN66XX specific operations.
+ */
+
+#ifndef __CN66XX_DEVICE_H__
+#define  __CN66XX_DEVICE_H__
+
+/* Register address and configuration for a CN6XXX devices.
+ * If device specific changes need to be made then add a struct to include
+ * device specific fields as shown in the commented section
+ */
+struct octeon_cn6xxx {
+	/** PCI interrupt summary register */
+	u8 __iomem *intr_sum_reg64;
+
+	/** PCI interrupt enable register */
+	u8 __iomem *intr_enb_reg64;
+
+	/** The PCI interrupt mask used by interrupt handler */
+	u64 intr_mask64;
+
+	struct octeon_config *conf;
+
+	/* Example additional fields - not used currently
+	 *  struct {
+	 *  }cn6xyz;
+	 */
+
+	/* For the purpose of atomic access to interrupt enable reg */
+	spinlock_t lock_for_droq_int_enb_reg;
+
+};
+
+enum octeon_pcie_mps {
+	PCIE_MPS_DEFAULT = -1,	/* Use the default setup by BIOS */
+	PCIE_MPS_128B = 0,
+	PCIE_MPS_256B = 1
+};
+
+enum octeon_pcie_mrrs {
+	PCIE_MRRS_DEFAULT = -1,	/* Use the default setup by BIOS */
+	PCIE_MRRS_128B = 0,
+	PCIE_MRRS_256B = 1,
+	PCIE_MRRS_512B = 2,
+	PCIE_MRRS_1024B = 3,
+	PCIE_MRRS_2048B = 4,
+	PCIE_MRRS_4096B = 5
+};
+
+/* Common functions for 66xx and 68xx */
+int lio_cn6xxx_soft_reset(struct octeon_device *oct);
+void lio_cn6xxx_enable_error_reporting(struct octeon_device *oct);
+void lio_cn6xxx_setup_pcie_mps(struct octeon_device *oct,
+			       enum octeon_pcie_mps mps);
+void lio_cn6xxx_setup_pcie_mrrs(struct octeon_device *oct,
+				enum octeon_pcie_mrrs mrrs);
+void lio_cn6xxx_setup_global_input_regs(struct octeon_device *oct);
+void lio_cn6xxx_setup_global_output_regs(struct octeon_device *oct);
+void lio_cn6xxx_setup_iq_regs(struct octeon_device *oct, u32 iq_no);
+void lio_cn6xxx_setup_oq_regs(struct octeon_device *oct, u32 oq_no);
+int lio_cn6xxx_enable_io_queues(struct octeon_device *oct);
+void lio_cn6xxx_disable_io_queues(struct octeon_device *oct);
+irqreturn_t lio_cn6xxx_process_interrupt_regs(void *dev);
+void lio_cn6xxx_bar1_idx_setup(struct octeon_device *oct, u64 core_addr,
+			       u32 idx, int valid);
+void lio_cn6xxx_bar1_idx_write(struct octeon_device *oct, u32 idx, u32 mask);
+u32 lio_cn6xxx_bar1_idx_read(struct octeon_device *oct, u32 idx);
+u32
+lio_cn6xxx_update_read_index(struct octeon_instr_queue *iq);
+void lio_cn6xxx_enable_interrupt(struct octeon_device *oct, u8 unused);
+void lio_cn6xxx_disable_interrupt(struct octeon_device *oct, u8 unused);
+void cn6xxx_get_pcie_qlmport(struct octeon_device *oct);
+void lio_cn6xxx_setup_reg_address(struct octeon_device *oct, void *chip,
+				  struct octeon_reg_list *reg_list);
+u32 lio_cn6xxx_coprocessor_clock(struct octeon_device *oct);
+u32 lio_cn6xxx_get_oq_ticks(struct octeon_device *oct, u32 time_intr_in_us);
+int lio_setup_cn66xx_octeon_device(struct octeon_device *oct);
+int lio_validate_cn6xxx_config_info(struct octeon_device *oct,
+				    struct octeon_config *conf6xxx);
+
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/cn66xx_regs.h b/drivers/net/ethernet/cavium/liquidio/cn66xx_regs.h
new file mode 100644
index 000000000..7aad40b2a
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn66xx_regs.h
@@ -0,0 +1,530 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*! \file cn66xx_regs.h
+ *  \brief Host Driver: Register Address and Register Mask values for
+ *  Octeon CN66XX devices.
+ */
+
+#ifndef __CN66XX_REGS_H__
+#define __CN66XX_REGS_H__
+
+#define     CN6XXX_XPANSION_BAR             0x30
+
+#define     CN6XXX_MSI_CAP                  0x50
+#define     CN6XXX_MSI_ADDR_LO              0x54
+#define     CN6XXX_MSI_ADDR_HI              0x58
+#define     CN6XXX_MSI_DATA                 0x5C
+
+#define     CN6XXX_PCIE_CAP                 0x70
+#define     CN6XXX_PCIE_DEVCAP              0x74
+#define     CN6XXX_PCIE_DEVCTL              0x78
+#define     CN6XXX_PCIE_LINKCAP             0x7C
+#define     CN6XXX_PCIE_LINKCTL             0x80
+#define     CN6XXX_PCIE_SLOTCAP             0x84
+#define     CN6XXX_PCIE_SLOTCTL             0x88
+
+#define     CN6XXX_PCIE_ENH_CAP             0x100
+#define     CN6XXX_PCIE_UNCORR_ERR_STATUS   0x104
+#define     CN6XXX_PCIE_UNCORR_ERR_MASK     0x108
+#define     CN6XXX_PCIE_UNCORR_ERR          0x10C
+#define     CN6XXX_PCIE_CORR_ERR_STATUS     0x110
+#define     CN6XXX_PCIE_CORR_ERR_MASK       0x114
+#define     CN6XXX_PCIE_ADV_ERR_CAP         0x118
+
+#define     CN6XXX_PCIE_ACK_REPLAY_TIMER    0x700
+#define     CN6XXX_PCIE_OTHER_MSG           0x704
+#define     CN6XXX_PCIE_PORT_FORCE_LINK     0x708
+#define     CN6XXX_PCIE_ACK_FREQ            0x70C
+#define     CN6XXX_PCIE_PORT_LINK_CTL       0x710
+#define     CN6XXX_PCIE_LANE_SKEW           0x714
+#define     CN6XXX_PCIE_SYM_NUM             0x718
+#define     CN6XXX_PCIE_FLTMSK              0x720
+
+/* ##############  BAR0 Registers ################  */
+
+#define    CN6XXX_SLI_CTL_PORT0                    0x0050
+#define    CN6XXX_SLI_CTL_PORT1                    0x0060
+
+#define    CN6XXX_SLI_WINDOW_CTL                   0x02E0
+#define    CN6XXX_SLI_DBG_DATA                     0x0310
+#define    CN6XXX_SLI_SCRATCH1                     0x03C0
+#define    CN6XXX_SLI_SCRATCH2                     0x03D0
+#define    CN6XXX_SLI_CTL_STATUS                   0x0570
+
+#define    CN6XXX_WIN_WR_ADDR_LO                   0x0000
+#define    CN6XXX_WIN_WR_ADDR_HI                   0x0004
+#define    CN6XXX_WIN_WR_ADDR64                    CN6XXX_WIN_WR_ADDR_LO
+
+#define    CN6XXX_WIN_RD_ADDR_LO                   0x0010
+#define    CN6XXX_WIN_RD_ADDR_HI                   0x0014
+#define    CN6XXX_WIN_RD_ADDR64                    CN6XXX_WIN_RD_ADDR_LO
+
+#define    CN6XXX_WIN_WR_DATA_LO                   0x0020
+#define    CN6XXX_WIN_WR_DATA_HI                   0x0024
+#define    CN6XXX_WIN_WR_DATA64                    CN6XXX_WIN_WR_DATA_LO
+
+#define    CN6XXX_WIN_RD_DATA_LO                   0x0040
+#define    CN6XXX_WIN_RD_DATA_HI                   0x0044
+#define    CN6XXX_WIN_RD_DATA64                    CN6XXX_WIN_RD_DATA_LO
+
+#define    CN6XXX_WIN_WR_MASK_LO                   0x0030
+#define    CN6XXX_WIN_WR_MASK_HI                   0x0034
+#define    CN6XXX_WIN_WR_MASK_REG                  CN6XXX_WIN_WR_MASK_LO
+
+/* 1 register (32-bit) to enable Input queues */
+#define    CN6XXX_SLI_PKT_INSTR_ENB               0x1000
+
+/* 1 register (32-bit) to enable Output queues */
+#define    CN6XXX_SLI_PKT_OUT_ENB                 0x1010
+
+/* 1 register (32-bit) to determine whether Output queues are in reset. */
+#define    CN6XXX_SLI_PORT_IN_RST_OQ              0x11F0
+
+/* 1 register (32-bit) to determine whether Input queues are in reset. */
+#define    CN6XXX_SLI_PORT_IN_RST_IQ              0x11F4
+
+/*###################### REQUEST QUEUE #########################*/
+
+/* 1 register (32-bit) - instr. size of each input queue. */
+#define    CN6XXX_SLI_PKT_INSTR_SIZE             0x1020
+
+/* 32 registers for Input Queue Instr Count - SLI_PKT_IN_DONE0_CNTS */
+#define    CN6XXX_SLI_IQ_INSTR_COUNT_START       0x2000
+
+/* 32 registers for Input Queue Start Addr - SLI_PKT0_INSTR_BADDR */
+#define    CN6XXX_SLI_IQ_BASE_ADDR_START64       0x2800
+
+/* 32 registers for Input Doorbell - SLI_PKT0_INSTR_BAOFF_DBELL */
+#define    CN6XXX_SLI_IQ_DOORBELL_START          0x2C00
+
+/* 32 registers for Input Queue size - SLI_PKT0_INSTR_FIFO_RSIZE */
+#define    CN6XXX_SLI_IQ_SIZE_START              0x3000
+
+/* 32 registers for Instruction Header Options - SLI_PKT0_INSTR_HEADER */
+#define    CN6XXX_SLI_IQ_PKT_INSTR_HDR_START64   0x3400
+
+/* 1 register (64-bit) - Back Pressure for each input queue - SLI_PKT0_IN_BP */
+#define    CN66XX_SLI_INPUT_BP_START64           0x3800
+
+/* Each Input Queue register is at a 16-byte Offset in BAR0 */
+#define    CN6XXX_IQ_OFFSET                      0x10
+
+/* 1 register (32-bit) - ES, RO, NS, Arbitration for Input Queue Data &
+ * gather list fetches. SLI_PKT_INPUT_CONTROL.
+ */
+#define    CN6XXX_SLI_PKT_INPUT_CONTROL          0x1170
+
+/* 1 register (64-bit) - Number of instructions to read at one time
+ * - 2 bits for each input ring. SLI_PKT_INSTR_RD_SIZE.
+ */
+#define    CN6XXX_SLI_PKT_INSTR_RD_SIZE          0x11A0
+
+/* 1 register (64-bit) - Assign Input ring to MAC port
+ * - 2 bits for each input ring. SLI_PKT_IN_PCIE_PORT.
+ */
+#define    CN6XXX_SLI_IN_PCIE_PORT               0x11B0
+
+/*------- Request Queue Macros ---------*/
+#define    CN6XXX_SLI_IQ_BASE_ADDR64(iq)          \
+	(CN6XXX_SLI_IQ_BASE_ADDR_START64 + ((iq) * CN6XXX_IQ_OFFSET))
+
+#define    CN6XXX_SLI_IQ_SIZE(iq)                 \
+	(CN6XXX_SLI_IQ_SIZE_START + ((iq) * CN6XXX_IQ_OFFSET))
+
+#define    CN6XXX_SLI_IQ_PKT_INSTR_HDR64(iq)      \
+	(CN6XXX_SLI_IQ_PKT_INSTR_HDR_START64 + ((iq) * CN6XXX_IQ_OFFSET))
+
+#define    CN6XXX_SLI_IQ_DOORBELL(iq)             \
+	(CN6XXX_SLI_IQ_DOORBELL_START + ((iq) * CN6XXX_IQ_OFFSET))
+
+#define    CN6XXX_SLI_IQ_INSTR_COUNT(iq)          \
+	(CN6XXX_SLI_IQ_INSTR_COUNT_START + ((iq) * CN6XXX_IQ_OFFSET))
+
+#define    CN66XX_SLI_IQ_BP64(iq)                 \
+	(CN66XX_SLI_INPUT_BP_START64 + ((iq) * CN6XXX_IQ_OFFSET))
+
+/*------------------ Masks ----------------*/
+#define    CN6XXX_INPUT_CTL_ROUND_ROBIN_ARB         BIT(22)
+#define    CN6XXX_INPUT_CTL_DATA_NS                 BIT(8)
+#define    CN6XXX_INPUT_CTL_DATA_ES_64B_SWAP        BIT(6)
+#define    CN6XXX_INPUT_CTL_DATA_RO                 BIT(5)
+#define    CN6XXX_INPUT_CTL_USE_CSR                 BIT(4)
+#define    CN6XXX_INPUT_CTL_GATHER_NS               BIT(3)
+#define    CN6XXX_INPUT_CTL_GATHER_ES_64B_SWAP      BIT(2)
+#define    CN6XXX_INPUT_CTL_GATHER_RO               BIT(1)
+
+#ifdef __BIG_ENDIAN_BITFIELD
+#define    CN6XXX_INPUT_CTL_MASK                    \
+	(CN6XXX_INPUT_CTL_DATA_ES_64B_SWAP      \
+	  | CN6XXX_INPUT_CTL_USE_CSR              \
+	  | CN6XXX_INPUT_CTL_GATHER_ES_64B_SWAP)
+#else
+#define    CN6XXX_INPUT_CTL_MASK                    \
+	(CN6XXX_INPUT_CTL_DATA_ES_64B_SWAP     \
+	  | CN6XXX_INPUT_CTL_USE_CSR)
+#endif
+
+/*############################ OUTPUT QUEUE #########################*/
+
+/* 32 registers for Output queue buffer and info size - SLI_PKT0_OUT_SIZE */
+#define    CN6XXX_SLI_OQ0_BUFF_INFO_SIZE         0x0C00
+
+/* 32 registers for Output Queue Start Addr - SLI_PKT0_SLIST_BADDR */
+#define    CN6XXX_SLI_OQ_BASE_ADDR_START64       0x1400
+
+/* 32 registers for Output Queue Packet Credits - SLI_PKT0_SLIST_BAOFF_DBELL */
+#define    CN6XXX_SLI_OQ_PKT_CREDITS_START       0x1800
+
+/* 32 registers for Output Queue size - SLI_PKT0_SLIST_FIFO_RSIZE */
+#define    CN6XXX_SLI_OQ_SIZE_START              0x1C00
+
+/* 32 registers for Output Queue Packet Count - SLI_PKT0_CNTS */
+#define    CN6XXX_SLI_OQ_PKT_SENT_START          0x2400
+
+/* Each Output Queue register is at a 16-byte Offset in BAR0 */
+#define    CN6XXX_OQ_OFFSET                      0x10
+
+/* 1 register (32-bit) - 1 bit for each output queue
+ * - Relaxed Ordering setting for reading Output Queues descriptors
+ * - SLI_PKT_SLIST_ROR
+ */
+#define    CN6XXX_SLI_PKT_SLIST_ROR              0x1030
+
+/* 1 register (32-bit) - 1 bit for each output queue
+ * - No Snoop mode for reading Output Queues descriptors
+ * - SLI_PKT_SLIST_NS
+ */
+#define    CN6XXX_SLI_PKT_SLIST_NS               0x1040
+
+/* 1 register (64-bit) - 2 bits for each output queue
+ * - Endian-Swap mode for reading Output Queue descriptors
+ * - SLI_PKT_SLIST_ES
+ */
+#define    CN6XXX_SLI_PKT_SLIST_ES64             0x1050
+
+/* 1 register (32-bit) - 1 bit for each output queue
+ * - InfoPtr mode for Output Queues.
+ * - SLI_PKT_IPTR
+ */
+#define    CN6XXX_SLI_PKT_IPTR                   0x1070
+
+/* 1 register (32-bit) - 1 bit for each output queue
+ * - DPTR format selector for Output queues.
+ * - SLI_PKT_DPADDR
+ */
+#define    CN6XXX_SLI_PKT_DPADDR                 0x1080
+
+/* 1 register (32-bit) - 1 bit for each output queue
+ * - Relaxed Ordering setting for reading Output Queues data
+ * - SLI_PKT_DATA_OUT_ROR
+ */
+#define    CN6XXX_SLI_PKT_DATA_OUT_ROR           0x1090
+
+/* 1 register (32-bit) - 1 bit for each output queue
+ * - No Snoop mode for reading Output Queues data
+ * - SLI_PKT_DATA_OUT_NS
+ */
+#define    CN6XXX_SLI_PKT_DATA_OUT_NS            0x10A0
+
+/* 1 register (64-bit)  - 2 bits for each output queue
+ * - Endian-Swap mode for reading Output Queue data
+ * - SLI_PKT_DATA_OUT_ES
+ */
+#define    CN6XXX_SLI_PKT_DATA_OUT_ES64          0x10B0
+
+/* 1 register (32-bit) - 1 bit for each output queue
+ * - Controls whether SLI_PKTn_CNTS is incremented for bytes or for packets.
+ * - SLI_PKT_OUT_BMODE
+ */
+#define    CN6XXX_SLI_PKT_OUT_BMODE              0x10D0
+
+/* 1 register (64-bit) - 2 bits for each output queue
+ * - Assign PCIE port for Output queues
+ * - SLI_PKT_PCIE_PORT.
+ */
+#define    CN6XXX_SLI_PKT_PCIE_PORT64            0x10E0
+
+/* 1 (64-bit) register for Output Queue Packet Count Interrupt Threshold
+ * & Time Threshold. The same setting applies to all 32 queues.
+ * The register is defined as a 64-bit registers, but we use the
+ * 32-bit offsets to define distinct addresses.
+ */
+#define    CN6XXX_SLI_OQ_INT_LEVEL_PKTS          0x1120
+#define    CN6XXX_SLI_OQ_INT_LEVEL_TIME          0x1124
+
+/* 1 (64-bit register) for Output Queue backpressure across all rings. */
+#define    CN6XXX_SLI_OQ_WMARK                   0x1180
+
+/* 1 register to control output queue global backpressure & ring enable. */
+#define    CN6XXX_SLI_PKT_CTL                    0x1220
+
+/*------- Output Queue Macros ---------*/
+#define    CN6XXX_SLI_OQ_BASE_ADDR64(oq)          \
+	(CN6XXX_SLI_OQ_BASE_ADDR_START64 + ((oq) * CN6XXX_OQ_OFFSET))
+
+#define    CN6XXX_SLI_OQ_SIZE(oq)                 \
+	(CN6XXX_SLI_OQ_SIZE_START + ((oq) * CN6XXX_OQ_OFFSET))
+
+#define    CN6XXX_SLI_OQ_BUFF_INFO_SIZE(oq)                 \
+	(CN6XXX_SLI_OQ0_BUFF_INFO_SIZE + ((oq) * CN6XXX_OQ_OFFSET))
+
+#define    CN6XXX_SLI_OQ_PKTS_SENT(oq)            \
+	(CN6XXX_SLI_OQ_PKT_SENT_START + ((oq) * CN6XXX_OQ_OFFSET))
+
+#define    CN6XXX_SLI_OQ_PKTS_CREDIT(oq)          \
+	(CN6XXX_SLI_OQ_PKT_CREDITS_START + ((oq) * CN6XXX_OQ_OFFSET))
+
+/*######################### DMA Counters #########################*/
+
+/* 2 registers (64-bit) - DMA Count - 1 for each DMA counter 0/1. */
+#define    CN6XXX_DMA_CNT_START                   0x0400
+
+/* 2 registers (64-bit) - DMA Timer 0/1, contains DMA timer values
+ * SLI_DMA_0_TIM
+ */
+#define    CN6XXX_DMA_TIM_START                   0x0420
+
+/* 2 registers (64-bit) - DMA count & Time Interrupt threshold -
+ * SLI_DMA_0_INT_LEVEL
+ */
+#define    CN6XXX_DMA_INT_LEVEL_START             0x03E0
+
+/* Each DMA register is at a 16-byte Offset in BAR0 */
+#define    CN6XXX_DMA_OFFSET                      0x10
+
+/*---------- DMA Counter Macros ---------*/
+#define    CN6XXX_DMA_CNT(dq)                      \
+	(CN6XXX_DMA_CNT_START + ((dq) * CN6XXX_DMA_OFFSET))
+
+#define    CN6XXX_DMA_INT_LEVEL(dq)                \
+	(CN6XXX_DMA_INT_LEVEL_START + ((dq) * CN6XXX_DMA_OFFSET))
+
+#define    CN6XXX_DMA_PKT_INT_LEVEL(dq)            \
+	(CN6XXX_DMA_INT_LEVEL_START + ((dq) * CN6XXX_DMA_OFFSET))
+
+#define    CN6XXX_DMA_TIME_INT_LEVEL(dq)           \
+	(CN6XXX_DMA_INT_LEVEL_START + 4 + ((dq) * CN6XXX_DMA_OFFSET))
+
+#define    CN6XXX_DMA_TIM(dq)                      \
+	(CN6XXX_DMA_TIM_START + ((dq) * CN6XXX_DMA_OFFSET))
+
+/*######################## INTERRUPTS #########################*/
+
+/* 1 register (64-bit) for Interrupt Summary */
+#define    CN6XXX_SLI_INT_SUM64                  0x0330
+
+/* 1 register (64-bit) for Interrupt Enable */
+#define    CN6XXX_SLI_INT_ENB64_PORT0            0x0340
+#define    CN6XXX_SLI_INT_ENB64_PORT1            0x0350
+
+/* 1 register (32-bit) to enable Output Queue Packet/Byte Count Interrupt */
+#define    CN6XXX_SLI_PKT_CNT_INT_ENB            0x1150
+
+/* 1 register (32-bit) to enable Output Queue Packet Timer Interrupt */
+#define    CN6XXX_SLI_PKT_TIME_INT_ENB           0x1160
+
+/* 1 register (32-bit) to indicate which Output Queue reached pkt threshold */
+#define    CN6XXX_SLI_PKT_CNT_INT                0x1130
+
+/* 1 register (32-bit) to indicate which Output Queue reached time threshold */
+#define    CN6XXX_SLI_PKT_TIME_INT               0x1140
+
+/*------------------ Interrupt Masks ----------------*/
+
+#define    CN6XXX_INTR_RML_TIMEOUT_ERR           BIT(1)
+#define    CN6XXX_INTR_BAR0_RW_TIMEOUT_ERR       BIT(2)
+#define    CN6XXX_INTR_IO2BIG_ERR                BIT(3)
+#define    CN6XXX_INTR_PKT_COUNT                 BIT(4)
+#define    CN6XXX_INTR_PKT_TIME                  BIT(5)
+#define    CN6XXX_INTR_M0UPB0_ERR                BIT(8)
+#define    CN6XXX_INTR_M0UPWI_ERR                BIT(9)
+#define    CN6XXX_INTR_M0UNB0_ERR                BIT(10)
+#define    CN6XXX_INTR_M0UNWI_ERR                BIT(11)
+#define    CN6XXX_INTR_M1UPB0_ERR                BIT(12)
+#define    CN6XXX_INTR_M1UPWI_ERR                BIT(13)
+#define    CN6XXX_INTR_M1UNB0_ERR                BIT(14)
+#define    CN6XXX_INTR_M1UNWI_ERR                BIT(15)
+#define    CN6XXX_INTR_MIO_INT0                  BIT(16)
+#define    CN6XXX_INTR_MIO_INT1                  BIT(17)
+#define    CN6XXX_INTR_MAC_INT0                  BIT(18)
+#define    CN6XXX_INTR_MAC_INT1                  BIT(19)
+
+#define    CN6XXX_INTR_DMA0_FORCE                BIT_ULL(32)
+#define    CN6XXX_INTR_DMA1_FORCE                BIT_ULL(33)
+#define    CN6XXX_INTR_DMA0_COUNT                BIT_ULL(34)
+#define    CN6XXX_INTR_DMA1_COUNT                BIT_ULL(35)
+#define    CN6XXX_INTR_DMA0_TIME                 BIT_ULL(36)
+#define    CN6XXX_INTR_DMA1_TIME                 BIT_ULL(37)
+#define    CN6XXX_INTR_INSTR_DB_OF_ERR           BIT_ULL(48)
+#define    CN6XXX_INTR_SLIST_DB_OF_ERR           BIT_ULL(49)
+#define    CN6XXX_INTR_POUT_ERR                  BIT_ULL(50)
+#define    CN6XXX_INTR_PIN_BP_ERR                BIT_ULL(51)
+#define    CN6XXX_INTR_PGL_ERR                   BIT_ULL(52)
+#define    CN6XXX_INTR_PDI_ERR                   BIT_ULL(53)
+#define    CN6XXX_INTR_POP_ERR                   BIT_ULL(54)
+#define    CN6XXX_INTR_PINS_ERR                  BIT_ULL(55)
+#define    CN6XXX_INTR_SPRT0_ERR                 BIT_ULL(56)
+#define    CN6XXX_INTR_SPRT1_ERR                 BIT_ULL(57)
+#define    CN6XXX_INTR_ILL_PAD_ERR               BIT_ULL(60)
+
+#define    CN6XXX_INTR_DMA0_DATA                 (CN6XXX_INTR_DMA0_TIME)
+
+#define    CN6XXX_INTR_DMA1_DATA                 (CN6XXX_INTR_DMA1_TIME)
+
+#define    CN6XXX_INTR_DMA_DATA                  \
+	(CN6XXX_INTR_DMA0_DATA | CN6XXX_INTR_DMA1_DATA)
+
+#define    CN6XXX_INTR_PKT_DATA                  (CN6XXX_INTR_PKT_TIME | \
+						  CN6XXX_INTR_PKT_COUNT)
+
+/* Sum of interrupts for all PCI-Express Data Interrupts */
+#define    CN6XXX_INTR_PCIE_DATA                 \
+	(CN6XXX_INTR_DMA_DATA | CN6XXX_INTR_PKT_DATA)
+
+#define    CN6XXX_INTR_MIO                       \
+	(CN6XXX_INTR_MIO_INT0 | CN6XXX_INTR_MIO_INT1)
+
+#define    CN6XXX_INTR_MAC                       \
+	(CN6XXX_INTR_MAC_INT0 | CN6XXX_INTR_MAC_INT1)
+
+/* Sum of interrupts for error events */
+#define    CN6XXX_INTR_ERR                       \
+	(CN6XXX_INTR_BAR0_RW_TIMEOUT_ERR    \
+	   | CN6XXX_INTR_IO2BIG_ERR             \
+	   | CN6XXX_INTR_M0UPB0_ERR             \
+	   | CN6XXX_INTR_M0UPWI_ERR             \
+	   | CN6XXX_INTR_M0UNB0_ERR             \
+	   | CN6XXX_INTR_M0UNWI_ERR             \
+	   | CN6XXX_INTR_M1UPB0_ERR             \
+	   | CN6XXX_INTR_M1UPWI_ERR             \
+	   | CN6XXX_INTR_M1UNB0_ERR             \
+	   | CN6XXX_INTR_M1UNWI_ERR             \
+	   | CN6XXX_INTR_INSTR_DB_OF_ERR        \
+	   | CN6XXX_INTR_SLIST_DB_OF_ERR        \
+	   | CN6XXX_INTR_POUT_ERR               \
+	   | CN6XXX_INTR_PIN_BP_ERR             \
+	   | CN6XXX_INTR_PGL_ERR                \
+	   | CN6XXX_INTR_PDI_ERR                \
+	   | CN6XXX_INTR_POP_ERR                \
+	   | CN6XXX_INTR_PINS_ERR               \
+	   | CN6XXX_INTR_SPRT0_ERR              \
+	   | CN6XXX_INTR_SPRT1_ERR              \
+	   | CN6XXX_INTR_ILL_PAD_ERR)
+
+/* Programmed Mask for Interrupt Sum */
+#define    CN6XXX_INTR_MASK                      \
+	(CN6XXX_INTR_PCIE_DATA              \
+	   | CN6XXX_INTR_DMA0_FORCE             \
+	   | CN6XXX_INTR_DMA1_FORCE             \
+	   | CN6XXX_INTR_MIO                    \
+	   | CN6XXX_INTR_MAC                    \
+	   | CN6XXX_INTR_ERR)
+
+#define    CN6XXX_SLI_S2M_PORT0_CTL              0x3D80
+#define    CN6XXX_SLI_S2M_PORT1_CTL              0x3D90
+#define    CN6XXX_SLI_S2M_PORTX_CTL(port)        \
+	(CN6XXX_SLI_S2M_PORT0_CTL + ((port) * 0x10))
+
+#define    CN6XXX_SLI_INT_ENB64(port)            \
+	(CN6XXX_SLI_INT_ENB64_PORT0 + ((port) * 0x10))
+
+#define    CN6XXX_SLI_MAC_NUMBER                 0x3E00
+
+/* CN6XXX BAR1 Index registers. */
+#define    CN6XXX_PEM_BAR1_INDEX000                0x00011800C00000A8ULL
+#define    CN6XXX_PEM_OFFSET                       0x0000000001000000ULL
+
+#define    CN6XXX_BAR1_INDEX_START                 CN6XXX_PEM_BAR1_INDEX000
+#define    CN6XXX_PCI_BAR1_OFFSET                  0x8
+
+#define    CN6XXX_BAR1_REG(idx, port) \
+		(CN6XXX_BAR1_INDEX_START + ((port) * CN6XXX_PEM_OFFSET) + \
+		(CN6XXX_PCI_BAR1_OFFSET * (idx)))
+
+/*############################ DPI #########################*/
+
+#define    CN6XXX_DPI_CTL                 0x0001df0000000040ULL
+
+#define    CN6XXX_DPI_DMA_CONTROL         0x0001df0000000048ULL
+
+#define    CN6XXX_DPI_REQ_GBL_ENB         0x0001df0000000050ULL
+
+#define    CN6XXX_DPI_REQ_ERR_RSP         0x0001df0000000058ULL
+
+#define    CN6XXX_DPI_REQ_ERR_RST         0x0001df0000000060ULL
+
+#define    CN6XXX_DPI_DMA_ENG0_ENB        0x0001df0000000080ULL
+
+#define    CN6XXX_DPI_DMA_ENG_ENB(q_no)   \
+	(CN6XXX_DPI_DMA_ENG0_ENB + ((q_no) * 8))
+
+#define    CN6XXX_DPI_DMA_ENG0_BUF        0x0001df0000000880ULL
+
+#define    CN6XXX_DPI_DMA_ENG_BUF(q_no)   \
+	(CN6XXX_DPI_DMA_ENG0_BUF + ((q_no) * 8))
+
+#define    CN6XXX_DPI_SLI_PRT0_CFG        0x0001df0000000900ULL
+#define    CN6XXX_DPI_SLI_PRT1_CFG        0x0001df0000000908ULL
+#define    CN6XXX_DPI_SLI_PRTX_CFG(port)        \
+	(CN6XXX_DPI_SLI_PRT0_CFG + ((port) * 0x10))
+
+#define    CN6XXX_DPI_DMA_COMMIT_MODE     BIT_ULL(58)
+#define    CN6XXX_DPI_DMA_PKT_HP          BIT_ULL(57)
+#define    CN6XXX_DPI_DMA_PKT_EN          BIT_ULL(56)
+#define    CN6XXX_DPI_DMA_O_ES            BIT_ULL(15)
+#define    CN6XXX_DPI_DMA_O_MODE          BIT_ULL(14)
+
+#define    CN6XXX_DPI_DMA_CTL_MASK             \
+	(CN6XXX_DPI_DMA_COMMIT_MODE    |    \
+	 CN6XXX_DPI_DMA_PKT_HP         |    \
+	 CN6XXX_DPI_DMA_PKT_EN         |    \
+	 CN6XXX_DPI_DMA_O_ES           |    \
+	 CN6XXX_DPI_DMA_O_MODE)
+
+/*############################ CIU #########################*/
+
+#define    CN6XXX_CIU_SOFT_BIST           0x0001070000000738ULL
+#define    CN6XXX_CIU_SOFT_RST            0x0001070000000740ULL
+
+/*############################ MIO #########################*/
+#define    CN6XXX_MIO_PTP_CLOCK_CFG       0x0001070000000f00ULL
+#define    CN6XXX_MIO_PTP_CLOCK_LO        0x0001070000000f08ULL
+#define    CN6XXX_MIO_PTP_CLOCK_HI        0x0001070000000f10ULL
+#define    CN6XXX_MIO_PTP_CLOCK_COMP      0x0001070000000f18ULL
+#define    CN6XXX_MIO_PTP_TIMESTAMP       0x0001070000000f20ULL
+#define    CN6XXX_MIO_PTP_EVT_CNT         0x0001070000000f28ULL
+#define    CN6XXX_MIO_PTP_CKOUT_THRESH_LO 0x0001070000000f30ULL
+#define    CN6XXX_MIO_PTP_CKOUT_THRESH_HI 0x0001070000000f38ULL
+#define    CN6XXX_MIO_PTP_CKOUT_HI_INCR   0x0001070000000f40ULL
+#define    CN6XXX_MIO_PTP_CKOUT_LO_INCR   0x0001070000000f48ULL
+#define    CN6XXX_MIO_PTP_PPS_THRESH_LO   0x0001070000000f50ULL
+#define    CN6XXX_MIO_PTP_PPS_THRESH_HI   0x0001070000000f58ULL
+#define    CN6XXX_MIO_PTP_PPS_HI_INCR     0x0001070000000f60ULL
+#define    CN6XXX_MIO_PTP_PPS_LO_INCR     0x0001070000000f68ULL
+
+#define    CN6XXX_MIO_QLM4_CFG            0x00011800000015B0ULL
+#define    CN6XXX_MIO_RST_BOOT            0x0001180000001600ULL
+
+#define    CN6XXX_MIO_QLM_CFG_MASK        0x7
+
+/*############################ LMC #########################*/
+
+#define    CN6XXX_LMC0_RESET_CTL               0x0001180088000180ULL
+#define    CN6XXX_LMC0_RESET_CTL_DDR3RST_MASK  0x0000000000000001ULL
+
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/cn68xx_device.c b/drivers/net/ethernet/cavium/liquidio/cn68xx_device.c
new file mode 100644
index 000000000..30254e4cf
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn68xx_device.c
@@ -0,0 +1,183 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_main.h"
+#include "cn66xx_regs.h"
+#include "cn66xx_device.h"
+#include "cn68xx_device.h"
+#include "cn68xx_regs.h"
+
+static void lio_cn68xx_set_dpi_regs(struct octeon_device *oct)
+{
+	u32 i;
+	u32 fifo_sizes[6] = { 3, 3, 1, 1, 1, 8 };
+
+	lio_pci_writeq(oct, CN6XXX_DPI_DMA_CTL_MASK, CN6XXX_DPI_DMA_CONTROL);
+	dev_dbg(&oct->pci_dev->dev, "DPI_DMA_CONTROL: 0x%016llx\n",
+		lio_pci_readq(oct, CN6XXX_DPI_DMA_CONTROL));
+
+	for (i = 0; i < 6; i++) {
+		/* Prevent service of instruction queue for all DMA engines
+		 * Engine 5 will remain 0. Engines 0 - 4 will be setup by
+		 * core.
+		 */
+		lio_pci_writeq(oct, 0, CN6XXX_DPI_DMA_ENG_ENB(i));
+		lio_pci_writeq(oct, fifo_sizes[i], CN6XXX_DPI_DMA_ENG_BUF(i));
+		dev_dbg(&oct->pci_dev->dev, "DPI_ENG_BUF%d: 0x%016llx\n", i,
+			lio_pci_readq(oct, CN6XXX_DPI_DMA_ENG_BUF(i)));
+	}
+
+	/* DPI_SLI_PRT_CFG has MPS and MRRS settings that will be set
+	 * separately.
+	 */
+
+	lio_pci_writeq(oct, 1, CN6XXX_DPI_CTL);
+	dev_dbg(&oct->pci_dev->dev, "DPI_CTL: 0x%016llx\n",
+		lio_pci_readq(oct, CN6XXX_DPI_CTL));
+}
+
+static int lio_cn68xx_soft_reset(struct octeon_device *oct)
+{
+	lio_cn6xxx_soft_reset(oct);
+	lio_cn68xx_set_dpi_regs(oct);
+
+	return 0;
+}
+
+static void lio_cn68xx_setup_pkt_ctl_regs(struct octeon_device *oct)
+{
+	struct octeon_cn6xxx *cn68xx = (struct octeon_cn6xxx *)oct->chip;
+	u64 pktctl, tx_pipe, max_oqs;
+
+	pktctl = octeon_read_csr64(oct, CN6XXX_SLI_PKT_CTL);
+
+	/* 68XX specific */
+	max_oqs = CFG_GET_OQ_MAX_Q(CHIP_CONF(oct, cn6xxx));
+	tx_pipe  = octeon_read_csr64(oct, CN68XX_SLI_TX_PIPE);
+	tx_pipe &= 0xffffffffff00ffffULL; /* clear out NUMP field */
+	tx_pipe |= max_oqs << 16; /* put max_oqs in NUMP field */
+	octeon_write_csr64(oct, CN68XX_SLI_TX_PIPE, tx_pipe);
+
+	if (CFG_GET_IS_SLI_BP_ON(cn68xx->conf))
+		pktctl |= 0xF;
+	else
+		/* Disable per-port backpressure. */
+		pktctl &= ~0xF;
+	octeon_write_csr64(oct, CN6XXX_SLI_PKT_CTL, pktctl);
+}
+
+static int lio_cn68xx_setup_device_regs(struct octeon_device *oct)
+{
+	lio_cn6xxx_setup_pcie_mps(oct, PCIE_MPS_DEFAULT);
+	lio_cn6xxx_setup_pcie_mrrs(oct, PCIE_MRRS_256B);
+	lio_cn6xxx_enable_error_reporting(oct);
+
+	lio_cn6xxx_setup_global_input_regs(oct);
+	lio_cn68xx_setup_pkt_ctl_regs(oct);
+	lio_cn6xxx_setup_global_output_regs(oct);
+
+	/* Default error timeout value should be 0x200000 to avoid host hang
+	 * when reads invalid register
+	 */
+	octeon_write_csr64(oct, CN6XXX_SLI_WINDOW_CTL, 0x200000ULL);
+
+	return 0;
+}
+
+static inline void lio_cn68xx_vendor_message_fix(struct octeon_device *oct)
+{
+	u32 val = 0;
+
+	/* Set M_VEND1_DRP and M_VEND0_DRP bits */
+	pci_read_config_dword(oct->pci_dev, CN6XXX_PCIE_FLTMSK, &val);
+	val |= 0x3;
+	pci_write_config_dword(oct->pci_dev, CN6XXX_PCIE_FLTMSK, val);
+}
+
+static int lio_is_210nv(struct octeon_device *oct)
+{
+	u64 mio_qlm4_cfg = lio_pci_readq(oct, CN6XXX_MIO_QLM4_CFG);
+
+	return ((mio_qlm4_cfg & CN6XXX_MIO_QLM_CFG_MASK) == 0);
+}
+
+int lio_setup_cn68xx_octeon_device(struct octeon_device *oct)
+{
+	struct octeon_cn6xxx *cn68xx = (struct octeon_cn6xxx *)oct->chip;
+	u16 card_type = LIO_410NV;
+
+	if (octeon_map_pci_barx(oct, 0, 0))
+		return 1;
+
+	if (octeon_map_pci_barx(oct, 1, MAX_BAR1_IOREMAP_SIZE)) {
+		dev_err(&oct->pci_dev->dev, "%s CN68XX BAR1 map failed\n",
+			__func__);
+		octeon_unmap_pci_barx(oct, 0);
+		return 1;
+	}
+
+	spin_lock_init(&cn68xx->lock_for_droq_int_enb_reg);
+
+	oct->fn_list.setup_iq_regs = lio_cn6xxx_setup_iq_regs;
+	oct->fn_list.setup_oq_regs = lio_cn6xxx_setup_oq_regs;
+
+	oct->fn_list.process_interrupt_regs = lio_cn6xxx_process_interrupt_regs;
+	oct->fn_list.soft_reset = lio_cn68xx_soft_reset;
+	oct->fn_list.setup_device_regs = lio_cn68xx_setup_device_regs;
+	oct->fn_list.update_iq_read_idx = lio_cn6xxx_update_read_index;
+
+	oct->fn_list.bar1_idx_setup = lio_cn6xxx_bar1_idx_setup;
+	oct->fn_list.bar1_idx_write = lio_cn6xxx_bar1_idx_write;
+	oct->fn_list.bar1_idx_read = lio_cn6xxx_bar1_idx_read;
+
+	oct->fn_list.enable_interrupt = lio_cn6xxx_enable_interrupt;
+	oct->fn_list.disable_interrupt = lio_cn6xxx_disable_interrupt;
+
+	oct->fn_list.enable_io_queues = lio_cn6xxx_enable_io_queues;
+	oct->fn_list.disable_io_queues = lio_cn6xxx_disable_io_queues;
+
+	lio_cn6xxx_setup_reg_address(oct, oct->chip, &oct->reg_list);
+
+	/* Determine variant of card */
+	if (lio_is_210nv(oct))
+		card_type = LIO_210NV;
+
+	cn68xx->conf = (struct octeon_config *)
+		       oct_get_config_info(oct, card_type);
+	if (!cn68xx->conf) {
+		dev_err(&oct->pci_dev->dev, "%s No Config found for CN68XX %s\n",
+			__func__,
+			(card_type == LIO_410NV) ? LIO_410NV_NAME :
+			LIO_210NV_NAME);
+		octeon_unmap_pci_barx(oct, 0);
+		octeon_unmap_pci_barx(oct, 1);
+		return 1;
+	}
+
+	oct->coproc_clock_rate = 1000000ULL * lio_cn6xxx_coprocessor_clock(oct);
+
+	lio_cn68xx_vendor_message_fix(oct);
+
+	return 0;
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/cn68xx_device.h b/drivers/net/ethernet/cavium/liquidio/cn68xx_device.h
new file mode 100644
index 000000000..66b8d6bf5
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn68xx_device.h
@@ -0,0 +1,27 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*! \file  cn68xx_device.h
+ *  \brief Host Driver: Routines that perform CN68XX specific operations.
+ */
+
+#ifndef __CN68XX_DEVICE_H__
+#define  __CN68XX_DEVICE_H__
+
+int lio_setup_cn68xx_octeon_device(struct octeon_device *oct);
+
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/cn68xx_regs.h b/drivers/net/ethernet/cavium/liquidio/cn68xx_regs.h
new file mode 100644
index 000000000..0b742f09e
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/cn68xx_regs.h
@@ -0,0 +1,45 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*! \file cn68xx_regs.h
+ *  \brief Host Driver: Register Address and Register Mask values for
+ *  Octeon CN68XX devices. The register map for CN66XX is the same
+ *  for most registers. This file has the other registers that are
+ *  68XX-specific.
+ */
+
+#ifndef __CN68XX_REGS_H__
+#define __CN68XX_REGS_H__
+
+/*###################### REQUEST QUEUE #########################*/
+
+#define    CN68XX_SLI_IQ_PORT0_PKIND             0x0800
+
+#define    CN68XX_SLI_IQ_PORT_PKIND(iq)           \
+	(CN68XX_SLI_IQ_PORT0_PKIND + ((iq) * CN6XXX_IQ_OFFSET))
+
+/*############################ OUTPUT QUEUE #########################*/
+
+/* Starting pipe number and number of pipes used by the SLI packet output. */
+#define    CN68XX_SLI_TX_PIPE                    0x1230
+
+/*######################## INTERRUPTS #########################*/
+
+/*------------------ Interrupt Masks ----------------*/
+#define    CN68XX_INTR_PIPE_ERR                  BIT_ULL(61)
+
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/lio_core.c b/drivers/net/ethernet/cavium/liquidio/lio_core.c
new file mode 100644
index 000000000..882b2be06
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/lio_core.c
@@ -0,0 +1,1814 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/pci.h>
+#include <linux/if_vlan.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_nic.h"
+#include "octeon_main.h"
+#include "octeon_network.h"
+
+/* OOM task polling interval */
+#define LIO_OOM_POLL_INTERVAL_MS 250
+
+#define OCTNIC_MAX_SG  MAX_SKB_FRAGS
+
+/**
+ * lio_delete_glists - Delete gather lists
+ * @lio: per-network private data
+ */
+void lio_delete_glists(struct lio *lio)
+{
+	struct octnic_gather *g;
+	int i;
+
+	kfree(lio->glist_lock);
+	lio->glist_lock = NULL;
+
+	if (!lio->glist)
+		return;
+
+	for (i = 0; i < lio->oct_dev->num_iqs; i++) {
+		do {
+			g = (struct octnic_gather *)
+			    lio_list_delete_head(&lio->glist[i]);
+			kfree(g);
+		} while (g);
+
+		if (lio->glists_virt_base && lio->glists_virt_base[i] &&
+		    lio->glists_dma_base && lio->glists_dma_base[i]) {
+			lio_dma_free(lio->oct_dev,
+				     lio->glist_entry_size * lio->tx_qsize,
+				     lio->glists_virt_base[i],
+				     lio->glists_dma_base[i]);
+		}
+	}
+
+	kfree(lio->glists_virt_base);
+	lio->glists_virt_base = NULL;
+
+	kfree(lio->glists_dma_base);
+	lio->glists_dma_base = NULL;
+
+	kfree(lio->glist);
+	lio->glist = NULL;
+}
+
+/**
+ * lio_setup_glists - Setup gather lists
+ * @oct: octeon_device
+ * @lio: per-network private data
+ * @num_iqs: count of iqs to allocate
+ */
+int lio_setup_glists(struct octeon_device *oct, struct lio *lio, int num_iqs)
+{
+	struct octnic_gather *g;
+	int i, j;
+
+	lio->glist_lock =
+	    kcalloc(num_iqs, sizeof(*lio->glist_lock), GFP_KERNEL);
+	if (!lio->glist_lock)
+		return -ENOMEM;
+
+	lio->glist =
+	    kcalloc(num_iqs, sizeof(*lio->glist), GFP_KERNEL);
+	if (!lio->glist) {
+		kfree(lio->glist_lock);
+		lio->glist_lock = NULL;
+		return -ENOMEM;
+	}
+
+	lio->glist_entry_size =
+		ROUNDUP8((ROUNDUP4(OCTNIC_MAX_SG) >> 2) * OCT_SG_ENTRY_SIZE);
+
+	/* allocate memory to store virtual and dma base address of
+	 * per glist consistent memory
+	 */
+	lio->glists_virt_base = kcalloc(num_iqs, sizeof(*lio->glists_virt_base),
+					GFP_KERNEL);
+	lio->glists_dma_base = kcalloc(num_iqs, sizeof(*lio->glists_dma_base),
+				       GFP_KERNEL);
+
+	if (!lio->glists_virt_base || !lio->glists_dma_base) {
+		lio_delete_glists(lio);
+		return -ENOMEM;
+	}
+
+	for (i = 0; i < num_iqs; i++) {
+		int numa_node = dev_to_node(&oct->pci_dev->dev);
+
+		spin_lock_init(&lio->glist_lock[i]);
+
+		INIT_LIST_HEAD(&lio->glist[i]);
+
+		lio->glists_virt_base[i] =
+			lio_dma_alloc(oct,
+				      lio->glist_entry_size * lio->tx_qsize,
+				      &lio->glists_dma_base[i]);
+
+		if (!lio->glists_virt_base[i]) {
+			lio_delete_glists(lio);
+			return -ENOMEM;
+		}
+
+		for (j = 0; j < lio->tx_qsize; j++) {
+			g = kzalloc_node(sizeof(*g), GFP_KERNEL,
+					 numa_node);
+			if (!g)
+				g = kzalloc(sizeof(*g), GFP_KERNEL);
+			if (!g)
+				break;
+
+			g->sg = lio->glists_virt_base[i] +
+				(j * lio->glist_entry_size);
+
+			g->sg_dma_ptr = lio->glists_dma_base[i] +
+					(j * lio->glist_entry_size);
+
+			list_add_tail(&g->list, &lio->glist[i]);
+		}
+
+		if (j != lio->tx_qsize) {
+			lio_delete_glists(lio);
+			return -ENOMEM;
+		}
+	}
+
+	return 0;
+}
+
+int liquidio_set_feature(struct net_device *netdev, int cmd, u16 param1)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = cmd;
+	nctrl.ncmd.s.param1 = param1;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "Feature change failed in core (ret: 0x%x)\n",
+			ret);
+		if (ret > 0)
+			ret = -EIO;
+	}
+	return ret;
+}
+
+void octeon_report_tx_completion_to_bql(void *txq, unsigned int pkts_compl,
+					unsigned int bytes_compl)
+{
+	struct netdev_queue *netdev_queue = txq;
+
+	netdev_tx_completed_queue(netdev_queue, pkts_compl, bytes_compl);
+}
+
+void octeon_update_tx_completion_counters(void *buf, int reqtype,
+					  unsigned int *pkts_compl,
+					  unsigned int *bytes_compl)
+{
+	struct octnet_buf_free_info *finfo;
+	struct sk_buff *skb = NULL;
+	struct octeon_soft_command *sc;
+
+	switch (reqtype) {
+	case REQTYPE_NORESP_NET:
+	case REQTYPE_NORESP_NET_SG:
+		finfo = buf;
+		skb = finfo->skb;
+		break;
+
+	case REQTYPE_RESP_NET_SG:
+	case REQTYPE_RESP_NET:
+		sc = buf;
+		skb = sc->callback_arg;
+		break;
+
+	default:
+		return;
+	}
+
+	(*pkts_compl)++;
+	*bytes_compl += skb->len;
+}
+
+int octeon_report_sent_bytes_to_bql(void *buf, int reqtype)
+{
+	struct octnet_buf_free_info *finfo;
+	struct sk_buff *skb;
+	struct octeon_soft_command *sc;
+	struct netdev_queue *txq;
+
+	switch (reqtype) {
+	case REQTYPE_NORESP_NET:
+	case REQTYPE_NORESP_NET_SG:
+		finfo = buf;
+		skb = finfo->skb;
+		break;
+
+	case REQTYPE_RESP_NET_SG:
+	case REQTYPE_RESP_NET:
+		sc = buf;
+		skb = sc->callback_arg;
+		break;
+
+	default:
+		return 0;
+	}
+
+	txq = netdev_get_tx_queue(skb->dev, skb_get_queue_mapping(skb));
+	netdev_tx_sent_queue(txq, skb->len);
+
+	return netif_xmit_stopped(txq);
+}
+
+void liquidio_link_ctrl_cmd_completion(void *nctrl_ptr)
+{
+	struct octnic_ctrl_pkt *nctrl = (struct octnic_ctrl_pkt *)nctrl_ptr;
+	struct net_device *netdev = (struct net_device *)nctrl->netpndev;
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	u8 *mac;
+
+	if (nctrl->sc_status)
+		return;
+
+	switch (nctrl->ncmd.s.cmd) {
+	case OCTNET_CMD_CHANGE_DEVFLAGS:
+	case OCTNET_CMD_SET_MULTI_LIST:
+	case OCTNET_CMD_SET_UC_LIST:
+		break;
+
+	case OCTNET_CMD_CHANGE_MACADDR:
+		mac = ((u8 *)&nctrl->udd[0]) + 2;
+		if (nctrl->ncmd.s.param1) {
+			/* vfidx is 0 based, but vf_num (param1) is 1 based */
+			int vfidx = nctrl->ncmd.s.param1 - 1;
+			bool mac_is_admin_assigned = nctrl->ncmd.s.param2;
+
+			if (mac_is_admin_assigned)
+				netif_info(lio, probe, lio->netdev,
+					   "MAC Address %pM is configured for VF %d\n",
+					   mac, vfidx);
+		} else {
+			netif_info(lio, probe, lio->netdev,
+				   " MACAddr changed to %pM\n",
+				   mac);
+		}
+		break;
+
+	case OCTNET_CMD_GPIO_ACCESS:
+		netif_info(lio, probe, lio->netdev, "LED Flashing visual identification\n");
+
+		break;
+
+	case OCTNET_CMD_ID_ACTIVE:
+		netif_info(lio, probe, lio->netdev, "LED Flashing visual identification\n");
+
+		break;
+
+	case OCTNET_CMD_LRO_ENABLE:
+		dev_info(&oct->pci_dev->dev, "%s LRO Enabled\n", netdev->name);
+		break;
+
+	case OCTNET_CMD_LRO_DISABLE:
+		dev_info(&oct->pci_dev->dev, "%s LRO Disabled\n",
+			 netdev->name);
+		break;
+
+	case OCTNET_CMD_VERBOSE_ENABLE:
+		dev_info(&oct->pci_dev->dev, "%s Firmware debug enabled\n",
+			 netdev->name);
+		break;
+
+	case OCTNET_CMD_VERBOSE_DISABLE:
+		dev_info(&oct->pci_dev->dev, "%s Firmware debug disabled\n",
+			 netdev->name);
+		break;
+
+	case OCTNET_CMD_VLAN_FILTER_CTL:
+		if (nctrl->ncmd.s.param1)
+			dev_info(&oct->pci_dev->dev,
+				 "%s VLAN filter enabled\n", netdev->name);
+		else
+			dev_info(&oct->pci_dev->dev,
+				 "%s VLAN filter disabled\n", netdev->name);
+		break;
+
+	case OCTNET_CMD_ADD_VLAN_FILTER:
+		dev_info(&oct->pci_dev->dev, "%s VLAN filter %d added\n",
+			 netdev->name, nctrl->ncmd.s.param1);
+		break;
+
+	case OCTNET_CMD_DEL_VLAN_FILTER:
+		dev_info(&oct->pci_dev->dev, "%s VLAN filter %d removed\n",
+			 netdev->name, nctrl->ncmd.s.param1);
+		break;
+
+	case OCTNET_CMD_SET_SETTINGS:
+		dev_info(&oct->pci_dev->dev, "%s settings changed\n",
+			 netdev->name);
+
+		break;
+
+	/* Case to handle "OCTNET_CMD_TNL_RX_CSUM_CTL"
+	 * Command passed by NIC driver
+	 */
+	case OCTNET_CMD_TNL_RX_CSUM_CTL:
+		if (nctrl->ncmd.s.param1 == OCTNET_CMD_RXCSUM_ENABLE) {
+			netif_info(lio, probe, lio->netdev,
+				   "RX Checksum Offload Enabled\n");
+		} else if (nctrl->ncmd.s.param1 ==
+			   OCTNET_CMD_RXCSUM_DISABLE) {
+			netif_info(lio, probe, lio->netdev,
+				   "RX Checksum Offload Disabled\n");
+		}
+		break;
+
+		/* Case to handle "OCTNET_CMD_TNL_TX_CSUM_CTL"
+		 * Command passed by NIC driver
+		 */
+	case OCTNET_CMD_TNL_TX_CSUM_CTL:
+		if (nctrl->ncmd.s.param1 == OCTNET_CMD_TXCSUM_ENABLE) {
+			netif_info(lio, probe, lio->netdev,
+				   "TX Checksum Offload Enabled\n");
+		} else if (nctrl->ncmd.s.param1 ==
+			   OCTNET_CMD_TXCSUM_DISABLE) {
+			netif_info(lio, probe, lio->netdev,
+				   "TX Checksum Offload Disabled\n");
+		}
+		break;
+
+		/* Case to handle "OCTNET_CMD_VXLAN_PORT_CONFIG"
+		 * Command passed by NIC driver
+		 */
+	case OCTNET_CMD_VXLAN_PORT_CONFIG:
+		if (nctrl->ncmd.s.more == OCTNET_CMD_VXLAN_PORT_ADD) {
+			netif_info(lio, probe, lio->netdev,
+				   "VxLAN Destination UDP PORT:%d ADDED\n",
+				   nctrl->ncmd.s.param1);
+		} else if (nctrl->ncmd.s.more ==
+			   OCTNET_CMD_VXLAN_PORT_DEL) {
+			netif_info(lio, probe, lio->netdev,
+				   "VxLAN Destination UDP PORT:%d DELETED\n",
+				   nctrl->ncmd.s.param1);
+		}
+		break;
+
+	case OCTNET_CMD_SET_FLOW_CTL:
+		netif_info(lio, probe, lio->netdev, "Set RX/TX flow control parameters\n");
+		break;
+
+	case OCTNET_CMD_QUEUE_COUNT_CTL:
+		netif_info(lio, probe, lio->netdev, "Queue count updated to %d\n",
+			   nctrl->ncmd.s.param1);
+		break;
+
+	default:
+		dev_err(&oct->pci_dev->dev, "%s Unknown cmd %d\n", __func__,
+			nctrl->ncmd.s.cmd);
+	}
+}
+
+void octeon_pf_changed_vf_macaddr(struct octeon_device *oct, u8 *mac)
+{
+	bool macaddr_changed = false;
+	struct net_device *netdev;
+	struct lio *lio;
+
+	rtnl_lock();
+
+	netdev = oct->props[0].netdev;
+	lio = GET_LIO(netdev);
+
+	lio->linfo.macaddr_is_admin_asgnd = true;
+
+	if (!ether_addr_equal(netdev->dev_addr, mac)) {
+		macaddr_changed = true;
+		eth_hw_addr_set(netdev, mac);
+		ether_addr_copy(((u8 *)&lio->linfo.hw_addr) + 2, mac);
+		call_netdevice_notifiers(NETDEV_CHANGEADDR, netdev);
+	}
+
+	rtnl_unlock();
+
+	if (macaddr_changed)
+		dev_info(&oct->pci_dev->dev,
+			 "PF changed VF's MAC address to %pM\n", mac);
+
+	/* no need to notify the firmware of the macaddr change because
+	 * the PF did that already
+	 */
+}
+
+void octeon_schedule_rxq_oom_work(struct octeon_device *oct,
+				  struct octeon_droq *droq)
+{
+	struct net_device *netdev = oct->props[0].netdev;
+	struct lio *lio = GET_LIO(netdev);
+	struct cavium_wq *wq = &lio->rxq_status_wq[droq->q_no];
+
+	queue_delayed_work(wq->wq, &wq->wk.work,
+			   msecs_to_jiffies(LIO_OOM_POLL_INTERVAL_MS));
+}
+
+static void octnet_poll_check_rxq_oom_status(struct work_struct *work)
+{
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct lio *lio = (struct lio *)wk->ctxptr;
+	struct octeon_device *oct = lio->oct_dev;
+	int q_no = wk->ctxul;
+	struct octeon_droq *droq = oct->droq[q_no];
+
+	if (!ifstate_check(lio, LIO_IFSTATE_RUNNING) || !droq)
+		return;
+
+	if (octeon_retry_droq_refill(droq))
+		octeon_schedule_rxq_oom_work(oct, droq);
+}
+
+int setup_rx_oom_poll_fn(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct cavium_wq *wq;
+	int q, q_no;
+
+	for (q = 0; q < oct->num_oqs; q++) {
+		q_no = lio->linfo.rxpciq[q].s.q_no;
+		wq = &lio->rxq_status_wq[q_no];
+		wq->wq = alloc_workqueue("rxq-oom-status",
+					 WQ_MEM_RECLAIM, 0);
+		if (!wq->wq) {
+			dev_err(&oct->pci_dev->dev, "unable to create cavium rxq oom status wq\n");
+			return -ENOMEM;
+		}
+
+		INIT_DELAYED_WORK(&wq->wk.work,
+				  octnet_poll_check_rxq_oom_status);
+		wq->wk.ctxptr = lio;
+		wq->wk.ctxul = q_no;
+	}
+
+	return 0;
+}
+
+void cleanup_rx_oom_poll_fn(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct cavium_wq *wq;
+	int q_no;
+
+	for (q_no = 0; q_no < oct->num_oqs; q_no++) {
+		wq = &lio->rxq_status_wq[q_no];
+		if (wq->wq) {
+			cancel_delayed_work_sync(&wq->wk.work);
+			destroy_workqueue(wq->wq);
+			wq->wq = NULL;
+		}
+	}
+}
+
+/* Runs in interrupt context. */
+static void lio_update_txq_status(struct octeon_device *oct, int iq_num)
+{
+	struct octeon_instr_queue *iq = oct->instr_queue[iq_num];
+	struct net_device *netdev;
+	struct lio *lio;
+
+	netdev = oct->props[iq->ifidx].netdev;
+
+	/* This is needed because the first IQ does not have
+	 * a netdev associated with it.
+	 */
+	if (!netdev)
+		return;
+
+	lio = GET_LIO(netdev);
+	if (__netif_subqueue_stopped(netdev, iq->q_index) &&
+	    lio->linfo.link.s.link_up &&
+	    (!octnet_iq_is_full(oct, iq_num))) {
+		netif_wake_subqueue(netdev, iq->q_index);
+		INCR_INSTRQUEUE_PKT_COUNT(lio->oct_dev, iq_num,
+					  tx_restart, 1);
+	}
+}
+
+/**
+ * octeon_setup_droq - Setup output queue
+ * @oct: octeon device
+ * @q_no: which queue
+ * @num_descs: how many descriptors
+ * @desc_size: size of each descriptor
+ * @app_ctx: application context
+ */
+static int octeon_setup_droq(struct octeon_device *oct, int q_no, int num_descs,
+			     int desc_size, void *app_ctx)
+{
+	int ret_val;
+
+	dev_dbg(&oct->pci_dev->dev, "Creating Droq: %d\n", q_no);
+	/* droq creation and local register settings. */
+	ret_val = octeon_create_droq(oct, q_no, num_descs, desc_size, app_ctx);
+	if (ret_val < 0)
+		return ret_val;
+
+	if (ret_val == 1) {
+		dev_dbg(&oct->pci_dev->dev, "Using default droq %d\n", q_no);
+		return 0;
+	}
+
+	/* Enable the droq queues */
+	octeon_set_droq_pkt_op(oct, q_no, 1);
+
+	/* Send Credit for Octeon Output queues. Credits are always
+	 * sent after the output queue is enabled.
+	 */
+	writel(oct->droq[q_no]->max_count, oct->droq[q_no]->pkts_credit_reg);
+
+	return ret_val;
+}
+
+/**
+ * liquidio_push_packet - Routine to push packets arriving on Octeon interface upto network layer.
+ * @octeon_id:octeon device id.
+ * @skbuff:   skbuff struct to be passed to network layer.
+ * @len:      size of total data received.
+ * @rh:       Control header associated with the packet
+ * @param:    additional control data with the packet
+ * @arg:      farg registered in droq_ops
+ */
+static void
+liquidio_push_packet(u32 __maybe_unused octeon_id,
+		     void *skbuff,
+		     u32 len,
+		     union octeon_rh *rh,
+		     void *param,
+		     void *arg)
+{
+	struct net_device *netdev = (struct net_device *)arg;
+	struct octeon_droq *droq =
+	    container_of(param, struct octeon_droq, napi);
+	struct sk_buff *skb = (struct sk_buff *)skbuff;
+	struct skb_shared_hwtstamps *shhwtstamps;
+	struct napi_struct *napi = param;
+	u16 vtag = 0;
+	u32 r_dh_off;
+	u64 ns;
+
+	if (netdev) {
+		struct lio *lio = GET_LIO(netdev);
+		struct octeon_device *oct = lio->oct_dev;
+
+		/* Do not proceed if the interface is not in RUNNING state. */
+		if (!ifstate_check(lio, LIO_IFSTATE_RUNNING)) {
+			recv_buffer_free(skb);
+			droq->stats.rx_dropped++;
+			return;
+		}
+
+		skb->dev = netdev;
+
+		skb_record_rx_queue(skb, droq->q_no);
+		if (likely(len > MIN_SKB_SIZE)) {
+			struct octeon_skb_page_info *pg_info;
+			unsigned char *va;
+
+			pg_info = ((struct octeon_skb_page_info *)(skb->cb));
+			if (pg_info->page) {
+				/* For Paged allocation use the frags */
+				va = page_address(pg_info->page) +
+					pg_info->page_offset;
+				memcpy(skb->data, va, MIN_SKB_SIZE);
+				skb_put(skb, MIN_SKB_SIZE);
+				skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags,
+						pg_info->page,
+						pg_info->page_offset +
+						MIN_SKB_SIZE,
+						len - MIN_SKB_SIZE,
+						LIO_RXBUFFER_SZ);
+			}
+		} else {
+			struct octeon_skb_page_info *pg_info =
+				((struct octeon_skb_page_info *)(skb->cb));
+			skb_copy_to_linear_data(skb, page_address(pg_info->page)
+						+ pg_info->page_offset, len);
+			skb_put(skb, len);
+			put_page(pg_info->page);
+		}
+
+		r_dh_off = (rh->r_dh.len - 1) * BYTES_PER_DHLEN_UNIT;
+
+		if (oct->ptp_enable) {
+			if (rh->r_dh.has_hwtstamp) {
+				/* timestamp is included from the hardware at
+				 * the beginning of the packet.
+				 */
+				if (ifstate_check
+					(lio,
+					 LIO_IFSTATE_RX_TIMESTAMP_ENABLED)) {
+					/* Nanoseconds are in the first 64-bits
+					 * of the packet.
+					 */
+					memcpy(&ns, (skb->data + r_dh_off),
+					       sizeof(ns));
+					r_dh_off -= BYTES_PER_DHLEN_UNIT;
+					shhwtstamps = skb_hwtstamps(skb);
+					shhwtstamps->hwtstamp =
+						ns_to_ktime(ns +
+							    lio->ptp_adjust);
+				}
+			}
+		}
+
+		if (rh->r_dh.has_hash) {
+			__be32 *hash_be = (__be32 *)(skb->data + r_dh_off);
+			u32 hash = be32_to_cpu(*hash_be);
+
+			skb_set_hash(skb, hash, PKT_HASH_TYPE_L4);
+			r_dh_off -= BYTES_PER_DHLEN_UNIT;
+		}
+
+		skb_pull(skb, rh->r_dh.len * BYTES_PER_DHLEN_UNIT);
+		skb->protocol = eth_type_trans(skb, skb->dev);
+
+		if ((netdev->features & NETIF_F_RXCSUM) &&
+		    (((rh->r_dh.encap_on) &&
+		      (rh->r_dh.csum_verified & CNNIC_TUN_CSUM_VERIFIED)) ||
+		     (!(rh->r_dh.encap_on) &&
+		      ((rh->r_dh.csum_verified & CNNIC_CSUM_VERIFIED) ==
+			CNNIC_CSUM_VERIFIED))))
+			/* checksum has already been verified */
+			skb->ip_summed = CHECKSUM_UNNECESSARY;
+		else
+			skb->ip_summed = CHECKSUM_NONE;
+
+		/* Setting Encapsulation field on basis of status received
+		 * from the firmware
+		 */
+		if (rh->r_dh.encap_on) {
+			skb->encapsulation = 1;
+			skb->csum_level = 1;
+			droq->stats.rx_vxlan++;
+		}
+
+		/* inbound VLAN tag */
+		if ((netdev->features & NETIF_F_HW_VLAN_CTAG_RX) &&
+		    rh->r_dh.vlan) {
+			u16 priority = rh->r_dh.priority;
+			u16 vid = rh->r_dh.vlan;
+
+			vtag = (priority << VLAN_PRIO_SHIFT) | vid;
+			__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), vtag);
+		}
+
+		napi_gro_receive(napi, skb);
+
+		droq->stats.rx_bytes_received += len -
+			rh->r_dh.len * BYTES_PER_DHLEN_UNIT;
+		droq->stats.rx_pkts_received++;
+	} else {
+		recv_buffer_free(skb);
+	}
+}
+
+/**
+ * napi_schedule_wrapper - wrapper for calling napi_schedule
+ * @param: parameters to pass to napi_schedule
+ *
+ * Used when scheduling on different CPUs
+ */
+static void napi_schedule_wrapper(void *param)
+{
+	struct napi_struct *napi = param;
+
+	napi_schedule(napi);
+}
+
+/**
+ * liquidio_napi_drv_callback - callback when receive interrupt occurs and we are in NAPI mode
+ * @arg: pointer to octeon output queue
+ */
+static void liquidio_napi_drv_callback(void *arg)
+{
+	struct octeon_device *oct;
+	struct octeon_droq *droq = arg;
+	int this_cpu = smp_processor_id();
+
+	oct = droq->oct_dev;
+
+	if (OCTEON_CN23XX_PF(oct) || OCTEON_CN23XX_VF(oct) ||
+	    droq->cpu_id == this_cpu) {
+		napi_schedule_irqoff(&droq->napi);
+	} else {
+		INIT_CSD(&droq->csd, napi_schedule_wrapper, &droq->napi);
+		smp_call_function_single_async(droq->cpu_id, &droq->csd);
+	}
+}
+
+/**
+ * liquidio_napi_poll - Entry point for NAPI polling
+ * @napi: NAPI structure
+ * @budget: maximum number of items to process
+ */
+static int liquidio_napi_poll(struct napi_struct *napi, int budget)
+{
+	struct octeon_instr_queue *iq;
+	struct octeon_device *oct;
+	struct octeon_droq *droq;
+	int tx_done = 0, iq_no;
+	int work_done;
+
+	droq = container_of(napi, struct octeon_droq, napi);
+	oct = droq->oct_dev;
+	iq_no = droq->q_no;
+
+	/* Handle Droq descriptors */
+	work_done = octeon_droq_process_poll_pkts(oct, droq, budget);
+
+	/* Flush the instruction queue */
+	iq = oct->instr_queue[iq_no];
+	if (iq) {
+		/* TODO: move this check to inside octeon_flush_iq,
+		 * once check_db_timeout is removed
+		 */
+		if (atomic_read(&iq->instr_pending))
+			/* Process iq buffers with in the budget limits */
+			tx_done = octeon_flush_iq(oct, iq, budget);
+		else
+			tx_done = 1;
+		/* Update iq read-index rather than waiting for next interrupt.
+		 * Return back if tx_done is false.
+		 */
+		/* sub-queue status update */
+		lio_update_txq_status(oct, iq_no);
+	} else {
+		dev_err(&oct->pci_dev->dev, "%s:  iq (%d) num invalid\n",
+			__func__, iq_no);
+	}
+
+#define MAX_REG_CNT  2000000U
+	/* force enable interrupt if reg cnts are high to avoid wraparound */
+	if ((work_done < budget && tx_done) ||
+	    (iq && iq->pkt_in_done >= MAX_REG_CNT) ||
+	    (droq->pkt_count >= MAX_REG_CNT)) {
+		napi_complete_done(napi, work_done);
+
+		octeon_enable_irq(droq->oct_dev, droq->q_no);
+		return 0;
+	}
+
+	return (!tx_done) ? (budget) : (work_done);
+}
+
+/**
+ * liquidio_setup_io_queues - Setup input and output queues
+ * @octeon_dev: octeon device
+ * @ifidx: Interface index
+ * @num_iqs: input io queue count
+ * @num_oqs: output io queue count
+ *
+ * Note: Queues are with respect to the octeon device. Thus
+ * an input queue is for egress packets, and output queues
+ * are for ingress packets.
+ */
+int liquidio_setup_io_queues(struct octeon_device *octeon_dev, int ifidx,
+			     u32 num_iqs, u32 num_oqs)
+{
+	struct octeon_droq_ops droq_ops;
+	struct net_device *netdev;
+	struct octeon_droq *droq;
+	struct napi_struct *napi;
+	int cpu_id_modulus;
+	int num_tx_descs;
+	struct lio *lio;
+	int retval = 0;
+	int q, q_no;
+	int cpu_id;
+
+	netdev = octeon_dev->props[ifidx].netdev;
+
+	lio = GET_LIO(netdev);
+
+	memset(&droq_ops, 0, sizeof(struct octeon_droq_ops));
+
+	droq_ops.fptr = liquidio_push_packet;
+	droq_ops.farg = netdev;
+
+	droq_ops.poll_mode = 1;
+	droq_ops.napi_fn = liquidio_napi_drv_callback;
+	cpu_id = 0;
+	cpu_id_modulus = num_present_cpus();
+
+	/* set up DROQs. */
+	for (q = 0; q < num_oqs; q++) {
+		q_no = lio->linfo.rxpciq[q].s.q_no;
+		dev_dbg(&octeon_dev->pci_dev->dev,
+			"%s index:%d linfo.rxpciq.s.q_no:%d\n",
+			__func__, q, q_no);
+		retval = octeon_setup_droq(
+		    octeon_dev, q_no,
+		    CFG_GET_NUM_RX_DESCS_NIC_IF(octeon_get_conf(octeon_dev),
+						lio->ifidx),
+		    CFG_GET_NUM_RX_BUF_SIZE_NIC_IF(octeon_get_conf(octeon_dev),
+						   lio->ifidx),
+		    NULL);
+		if (retval) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"%s : Runtime DROQ(RxQ) creation failed.\n",
+				__func__);
+			return 1;
+		}
+
+		droq = octeon_dev->droq[q_no];
+		napi = &droq->napi;
+		dev_dbg(&octeon_dev->pci_dev->dev, "netif_napi_add netdev:%llx oct:%llx\n",
+			(u64)netdev, (u64)octeon_dev);
+		netif_napi_add(netdev, napi, liquidio_napi_poll);
+
+		/* designate a CPU for this droq */
+		droq->cpu_id = cpu_id;
+		cpu_id++;
+		if (cpu_id >= cpu_id_modulus)
+			cpu_id = 0;
+
+		octeon_register_droq_ops(octeon_dev, q_no, &droq_ops);
+	}
+
+	if (OCTEON_CN23XX_PF(octeon_dev) || OCTEON_CN23XX_VF(octeon_dev)) {
+		/* 23XX PF/VF can send/recv control messages (via the first
+		 * PF/VF-owned droq) from the firmware even if the ethX
+		 * interface is down, so that's why poll_mode must be off
+		 * for the first droq.
+		 */
+		octeon_dev->droq[0]->ops.poll_mode = 0;
+	}
+
+	/* set up IQs. */
+	for (q = 0; q < num_iqs; q++) {
+		num_tx_descs = CFG_GET_NUM_TX_DESCS_NIC_IF(
+		    octeon_get_conf(octeon_dev), lio->ifidx);
+		retval = octeon_setup_iq(octeon_dev, ifidx, q,
+					 lio->linfo.txpciq[q], num_tx_descs,
+					 netdev_get_tx_queue(netdev, q));
+		if (retval) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				" %s : Runtime IQ(TxQ) creation failed.\n",
+				__func__);
+			return 1;
+		}
+
+		/* XPS */
+		if (!OCTEON_CN23XX_VF(octeon_dev) && octeon_dev->msix_on &&
+		    octeon_dev->ioq_vector) {
+			struct octeon_ioq_vector    *ioq_vector;
+
+			ioq_vector = &octeon_dev->ioq_vector[q];
+			netif_set_xps_queue(netdev,
+					    &ioq_vector->affinity_mask,
+					    ioq_vector->iq_index);
+		}
+	}
+
+	return 0;
+}
+
+static
+int liquidio_schedule_msix_droq_pkt_handler(struct octeon_droq *droq, u64 ret)
+{
+	struct octeon_device *oct = droq->oct_dev;
+	struct octeon_device_priv *oct_priv =
+	    (struct octeon_device_priv *)oct->priv;
+
+	if (droq->ops.poll_mode) {
+		droq->ops.napi_fn(droq);
+	} else {
+		if (ret & MSIX_PO_INT) {
+			if (OCTEON_CN23XX_VF(oct))
+				dev_err(&oct->pci_dev->dev,
+					"should not come here should not get rx when poll mode = 0 for vf\n");
+			tasklet_schedule(&oct_priv->droq_tasklet);
+			return 1;
+		}
+		/* this will be flushed periodically by check iq db */
+		if (ret & MSIX_PI_INT)
+			return 0;
+	}
+
+	return 0;
+}
+
+irqreturn_t
+liquidio_msix_intr_handler(int __maybe_unused irq, void *dev)
+{
+	struct octeon_ioq_vector *ioq_vector = (struct octeon_ioq_vector *)dev;
+	struct octeon_device *oct = ioq_vector->oct_dev;
+	struct octeon_droq *droq = oct->droq[ioq_vector->droq_index];
+	u64 ret;
+
+	ret = oct->fn_list.msix_interrupt_handler(ioq_vector);
+
+	if (ret & MSIX_PO_INT || ret & MSIX_PI_INT)
+		liquidio_schedule_msix_droq_pkt_handler(droq, ret);
+
+	return IRQ_HANDLED;
+}
+
+/**
+ * liquidio_schedule_droq_pkt_handlers - Droq packet processor sceduler
+ * @oct: octeon device
+ */
+static void liquidio_schedule_droq_pkt_handlers(struct octeon_device *oct)
+{
+	struct octeon_device_priv *oct_priv =
+		(struct octeon_device_priv *)oct->priv;
+	struct octeon_droq *droq;
+	u64 oq_no;
+
+	if (oct->int_status & OCT_DEV_INTR_PKT_DATA) {
+		for (oq_no = 0; oq_no < MAX_OCTEON_OUTPUT_QUEUES(oct);
+		     oq_no++) {
+			if (!(oct->droq_intr & BIT_ULL(oq_no)))
+				continue;
+
+			droq = oct->droq[oq_no];
+
+			if (droq->ops.poll_mode) {
+				droq->ops.napi_fn(droq);
+				oct_priv->napi_mask |= BIT_ULL(oq_no);
+			} else {
+				tasklet_schedule(&oct_priv->droq_tasklet);
+			}
+		}
+	}
+}
+
+/**
+ * liquidio_legacy_intr_handler - Interrupt handler for octeon
+ * @irq: unused
+ * @dev: octeon device
+ */
+static
+irqreturn_t liquidio_legacy_intr_handler(int __maybe_unused irq, void *dev)
+{
+	struct octeon_device *oct = (struct octeon_device *)dev;
+	irqreturn_t ret;
+
+	/* Disable our interrupts for the duration of ISR */
+	oct->fn_list.disable_interrupt(oct, OCTEON_ALL_INTR);
+
+	ret = oct->fn_list.process_interrupt_regs(oct);
+
+	if (ret == IRQ_HANDLED)
+		liquidio_schedule_droq_pkt_handlers(oct);
+
+	/* Re-enable our interrupts  */
+	if (!(atomic_read(&oct->status) == OCT_DEV_IN_RESET))
+		oct->fn_list.enable_interrupt(oct, OCTEON_ALL_INTR);
+
+	return ret;
+}
+
+/**
+ * octeon_setup_interrupt - Setup interrupt for octeon device
+ * @oct: octeon device
+ * @num_ioqs: number of queues
+ *
+ *  Enable interrupt in Octeon device as given in the PCI interrupt mask.
+ */
+int octeon_setup_interrupt(struct octeon_device *oct, u32 num_ioqs)
+{
+	struct msix_entry *msix_entries;
+	char *queue_irq_names = NULL;
+	int i, num_interrupts = 0;
+	int num_alloc_ioq_vectors;
+	char *aux_irq_name = NULL;
+	int num_ioq_vectors;
+	int irqret, err;
+
+	if (oct->msix_on) {
+		oct->num_msix_irqs = num_ioqs;
+		if (OCTEON_CN23XX_PF(oct)) {
+			num_interrupts = MAX_IOQ_INTERRUPTS_PER_PF + 1;
+
+			/* one non ioq interrupt for handling
+			 * sli_mac_pf_int_sum
+			 */
+			oct->num_msix_irqs += 1;
+		} else if (OCTEON_CN23XX_VF(oct)) {
+			num_interrupts = MAX_IOQ_INTERRUPTS_PER_VF;
+		}
+
+		/* allocate storage for the names assigned to each irq */
+		oct->irq_name_storage =
+			kcalloc(num_interrupts, INTRNAMSIZ, GFP_KERNEL);
+		if (!oct->irq_name_storage) {
+			dev_err(&oct->pci_dev->dev, "Irq name storage alloc failed...\n");
+			return -ENOMEM;
+		}
+
+		queue_irq_names = oct->irq_name_storage;
+
+		if (OCTEON_CN23XX_PF(oct))
+			aux_irq_name = &queue_irq_names
+				[IRQ_NAME_OFF(MAX_IOQ_INTERRUPTS_PER_PF)];
+
+		oct->msix_entries = kcalloc(oct->num_msix_irqs,
+					    sizeof(struct msix_entry),
+					    GFP_KERNEL);
+		if (!oct->msix_entries) {
+			dev_err(&oct->pci_dev->dev, "Memory Alloc failed...\n");
+			kfree(oct->irq_name_storage);
+			oct->irq_name_storage = NULL;
+			return -ENOMEM;
+		}
+
+		msix_entries = (struct msix_entry *)oct->msix_entries;
+
+		/*Assumption is that pf msix vectors start from pf srn to pf to
+		 * trs and not from 0. if not change this code
+		 */
+		if (OCTEON_CN23XX_PF(oct)) {
+			for (i = 0; i < oct->num_msix_irqs - 1; i++)
+				msix_entries[i].entry =
+					oct->sriov_info.pf_srn + i;
+
+			msix_entries[oct->num_msix_irqs - 1].entry =
+				oct->sriov_info.trs;
+		} else if (OCTEON_CN23XX_VF(oct)) {
+			for (i = 0; i < oct->num_msix_irqs; i++)
+				msix_entries[i].entry = i;
+		}
+		num_alloc_ioq_vectors = pci_enable_msix_range(
+						oct->pci_dev, msix_entries,
+						oct->num_msix_irqs,
+						oct->num_msix_irqs);
+		if (num_alloc_ioq_vectors < 0) {
+			dev_err(&oct->pci_dev->dev, "unable to Allocate MSI-X interrupts\n");
+			kfree(oct->msix_entries);
+			oct->msix_entries = NULL;
+			kfree(oct->irq_name_storage);
+			oct->irq_name_storage = NULL;
+			return num_alloc_ioq_vectors;
+		}
+
+		dev_dbg(&oct->pci_dev->dev, "OCTEON: Enough MSI-X interrupts are allocated...\n");
+
+		num_ioq_vectors = oct->num_msix_irqs;
+		/* For PF, there is one non-ioq interrupt handler */
+		if (OCTEON_CN23XX_PF(oct)) {
+			num_ioq_vectors -= 1;
+
+			snprintf(aux_irq_name, INTRNAMSIZ,
+				 "LiquidIO%u-pf%u-aux", oct->octeon_id,
+				 oct->pf_num);
+			irqret = request_irq(
+					msix_entries[num_ioq_vectors].vector,
+					liquidio_legacy_intr_handler, 0,
+					aux_irq_name, oct);
+			if (irqret) {
+				dev_err(&oct->pci_dev->dev,
+					"Request_irq failed for MSIX interrupt Error: %d\n",
+					irqret);
+				pci_disable_msix(oct->pci_dev);
+				kfree(oct->msix_entries);
+				kfree(oct->irq_name_storage);
+				oct->irq_name_storage = NULL;
+				oct->msix_entries = NULL;
+				return irqret;
+			}
+		}
+		for (i = 0 ; i < num_ioq_vectors ; i++) {
+			if (OCTEON_CN23XX_PF(oct))
+				snprintf(&queue_irq_names[IRQ_NAME_OFF(i)],
+					 INTRNAMSIZ, "LiquidIO%u-pf%u-rxtx-%u",
+					 oct->octeon_id, oct->pf_num, i);
+
+			if (OCTEON_CN23XX_VF(oct))
+				snprintf(&queue_irq_names[IRQ_NAME_OFF(i)],
+					 INTRNAMSIZ, "LiquidIO%u-vf%u-rxtx-%u",
+					 oct->octeon_id, oct->vf_num, i);
+
+			irqret = request_irq(msix_entries[i].vector,
+					     liquidio_msix_intr_handler, 0,
+					     &queue_irq_names[IRQ_NAME_OFF(i)],
+					     &oct->ioq_vector[i]);
+
+			if (irqret) {
+				dev_err(&oct->pci_dev->dev,
+					"Request_irq failed for MSIX interrupt Error: %d\n",
+					irqret);
+				/* Freeing the non-ioq irq vector here . */
+				free_irq(msix_entries[num_ioq_vectors].vector,
+					 oct);
+
+				while (i) {
+					i--;
+					/* clearing affinity mask. */
+					irq_set_affinity_hint(
+						      msix_entries[i].vector,
+						      NULL);
+					free_irq(msix_entries[i].vector,
+						 &oct->ioq_vector[i]);
+				}
+				pci_disable_msix(oct->pci_dev);
+				kfree(oct->msix_entries);
+				kfree(oct->irq_name_storage);
+				oct->irq_name_storage = NULL;
+				oct->msix_entries = NULL;
+				return irqret;
+			}
+			oct->ioq_vector[i].vector = msix_entries[i].vector;
+			/* assign the cpu mask for this msix interrupt vector */
+			irq_set_affinity_hint(msix_entries[i].vector,
+					      &oct->ioq_vector[i].affinity_mask
+					      );
+		}
+		dev_dbg(&oct->pci_dev->dev, "OCTEON[%d]: MSI-X enabled\n",
+			oct->octeon_id);
+	} else {
+		err = pci_enable_msi(oct->pci_dev);
+		if (err)
+			dev_warn(&oct->pci_dev->dev, "Reverting to legacy interrupts. Error: %d\n",
+				 err);
+		else
+			oct->flags |= LIO_FLAG_MSI_ENABLED;
+
+		/* allocate storage for the names assigned to the irq */
+		oct->irq_name_storage = kzalloc(INTRNAMSIZ, GFP_KERNEL);
+		if (!oct->irq_name_storage)
+			return -ENOMEM;
+
+		queue_irq_names = oct->irq_name_storage;
+
+		if (OCTEON_CN23XX_PF(oct))
+			snprintf(&queue_irq_names[IRQ_NAME_OFF(0)], INTRNAMSIZ,
+				 "LiquidIO%u-pf%u-rxtx-%u",
+				 oct->octeon_id, oct->pf_num, 0);
+
+		if (OCTEON_CN23XX_VF(oct))
+			snprintf(&queue_irq_names[IRQ_NAME_OFF(0)], INTRNAMSIZ,
+				 "LiquidIO%u-vf%u-rxtx-%u",
+				 oct->octeon_id, oct->vf_num, 0);
+
+		irqret = request_irq(oct->pci_dev->irq,
+				     liquidio_legacy_intr_handler,
+				     IRQF_SHARED,
+				     &queue_irq_names[IRQ_NAME_OFF(0)], oct);
+		if (irqret) {
+			if (oct->flags & LIO_FLAG_MSI_ENABLED)
+				pci_disable_msi(oct->pci_dev);
+			dev_err(&oct->pci_dev->dev, "Request IRQ failed with code: %d\n",
+				irqret);
+			kfree(oct->irq_name_storage);
+			oct->irq_name_storage = NULL;
+			return irqret;
+		}
+	}
+	return 0;
+}
+
+/**
+ * liquidio_change_mtu - Net device change_mtu
+ * @netdev: network device
+ * @new_mtu: the new max transmit unit size
+ */
+int liquidio_change_mtu(struct net_device *netdev, int new_mtu)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octeon_soft_command *sc;
+	union octnet_cmd *ncmd;
+	int ret = 0;
+
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct, OCTNET_CMD_SIZE, 16, 0);
+	if (!sc) {
+		netif_info(lio, rx_err, lio->netdev,
+			   "Failed to allocate soft command\n");
+		return -ENOMEM;
+	}
+
+	ncmd = (union octnet_cmd *)sc->virtdptr;
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	ncmd->u64 = 0;
+	ncmd->s.cmd = OCTNET_CMD_CHANGE_MTU;
+	ncmd->s.param1 = new_mtu;
+
+	octeon_swap_8B_data((u64 *)ncmd, (OCTNET_CMD_SIZE >> 3));
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC,
+				    OPCODE_NIC_CMD, 0, 0, 0);
+
+	ret = octeon_send_soft_command(oct, sc);
+	if (ret == IQ_SEND_FAILED) {
+		netif_info(lio, rx_err, lio->netdev, "Failed to change MTU\n");
+		octeon_free_soft_command(oct, sc);
+		return -EINVAL;
+	}
+	/* Sleep on a wait queue till the cond flag indicates that the
+	 * response arrived or timed-out.
+	 */
+	ret = wait_for_sc_completion_timeout(oct, sc, 0);
+	if (ret)
+		return ret;
+
+	if (sc->sc_status) {
+		WRITE_ONCE(sc->caller_is_done, true);
+		return -EINVAL;
+	}
+
+	netdev->mtu = new_mtu;
+	lio->mtu = new_mtu;
+
+	WRITE_ONCE(sc->caller_is_done, true);
+	return 0;
+}
+
+int lio_wait_for_clean_oq(struct octeon_device *oct)
+{
+	int retry = 100, pending_pkts = 0;
+	int idx;
+
+	do {
+		pending_pkts = 0;
+
+		for (idx = 0; idx < MAX_OCTEON_OUTPUT_QUEUES(oct); idx++) {
+			if (!(oct->io_qmask.oq & BIT_ULL(idx)))
+				continue;
+			pending_pkts +=
+				atomic_read(&oct->droq[idx]->pkts_pending);
+		}
+
+		if (pending_pkts > 0)
+			schedule_timeout_uninterruptible(1);
+
+	} while (retry-- && pending_pkts);
+
+	return pending_pkts;
+}
+
+static void
+octnet_nic_stats_callback(struct octeon_device *oct_dev,
+			  u32 status, void *ptr)
+{
+	struct octeon_soft_command *sc = (struct octeon_soft_command *)ptr;
+	struct oct_nic_stats_resp *resp =
+	    (struct oct_nic_stats_resp *)sc->virtrptr;
+	struct nic_rx_stats *rsp_rstats = &resp->stats.fromwire;
+	struct nic_tx_stats *rsp_tstats = &resp->stats.fromhost;
+	struct nic_rx_stats *rstats = &oct_dev->link_stats.fromwire;
+	struct nic_tx_stats *tstats = &oct_dev->link_stats.fromhost;
+
+	if (status != OCTEON_REQUEST_TIMEOUT && !resp->status) {
+		octeon_swap_8B_data((u64 *)&resp->stats,
+				    (sizeof(struct oct_link_stats)) >> 3);
+
+		/* RX link-level stats */
+		rstats->total_rcvd = rsp_rstats->total_rcvd;
+		rstats->bytes_rcvd = rsp_rstats->bytes_rcvd;
+		rstats->total_bcst = rsp_rstats->total_bcst;
+		rstats->total_mcst = rsp_rstats->total_mcst;
+		rstats->runts      = rsp_rstats->runts;
+		rstats->ctl_rcvd   = rsp_rstats->ctl_rcvd;
+		/* Accounts for over/under-run of buffers */
+		rstats->fifo_err  = rsp_rstats->fifo_err;
+		rstats->dmac_drop = rsp_rstats->dmac_drop;
+		rstats->fcs_err   = rsp_rstats->fcs_err;
+		rstats->jabber_err = rsp_rstats->jabber_err;
+		rstats->l2_err    = rsp_rstats->l2_err;
+		rstats->frame_err = rsp_rstats->frame_err;
+		rstats->red_drops = rsp_rstats->red_drops;
+
+		/* RX firmware stats */
+		rstats->fw_total_rcvd = rsp_rstats->fw_total_rcvd;
+		rstats->fw_total_fwd = rsp_rstats->fw_total_fwd;
+		rstats->fw_total_mcast = rsp_rstats->fw_total_mcast;
+		rstats->fw_total_bcast = rsp_rstats->fw_total_bcast;
+		rstats->fw_err_pko = rsp_rstats->fw_err_pko;
+		rstats->fw_err_link = rsp_rstats->fw_err_link;
+		rstats->fw_err_drop = rsp_rstats->fw_err_drop;
+		rstats->fw_rx_vxlan = rsp_rstats->fw_rx_vxlan;
+		rstats->fw_rx_vxlan_err = rsp_rstats->fw_rx_vxlan_err;
+
+		/* Number of packets that are LROed      */
+		rstats->fw_lro_pkts = rsp_rstats->fw_lro_pkts;
+		/* Number of octets that are LROed       */
+		rstats->fw_lro_octs = rsp_rstats->fw_lro_octs;
+		/* Number of LRO packets formed          */
+		rstats->fw_total_lro = rsp_rstats->fw_total_lro;
+		/* Number of times lRO of packet aborted */
+		rstats->fw_lro_aborts = rsp_rstats->fw_lro_aborts;
+		rstats->fw_lro_aborts_port = rsp_rstats->fw_lro_aborts_port;
+		rstats->fw_lro_aborts_seq = rsp_rstats->fw_lro_aborts_seq;
+		rstats->fw_lro_aborts_tsval = rsp_rstats->fw_lro_aborts_tsval;
+		rstats->fw_lro_aborts_timer = rsp_rstats->fw_lro_aborts_timer;
+		/* intrmod: packet forward rate */
+		rstats->fwd_rate = rsp_rstats->fwd_rate;
+
+		/* TX link-level stats */
+		tstats->total_pkts_sent = rsp_tstats->total_pkts_sent;
+		tstats->total_bytes_sent = rsp_tstats->total_bytes_sent;
+		tstats->mcast_pkts_sent = rsp_tstats->mcast_pkts_sent;
+		tstats->bcast_pkts_sent = rsp_tstats->bcast_pkts_sent;
+		tstats->ctl_sent = rsp_tstats->ctl_sent;
+		/* Packets sent after one collision*/
+		tstats->one_collision_sent = rsp_tstats->one_collision_sent;
+		/* Packets sent after multiple collision*/
+		tstats->multi_collision_sent = rsp_tstats->multi_collision_sent;
+		/* Packets not sent due to max collisions */
+		tstats->max_collision_fail = rsp_tstats->max_collision_fail;
+		/* Packets not sent due to max deferrals */
+		tstats->max_deferral_fail = rsp_tstats->max_deferral_fail;
+		/* Accounts for over/under-run of buffers */
+		tstats->fifo_err = rsp_tstats->fifo_err;
+		tstats->runts = rsp_tstats->runts;
+		/* Total number of collisions detected */
+		tstats->total_collisions = rsp_tstats->total_collisions;
+
+		/* firmware stats */
+		tstats->fw_total_sent = rsp_tstats->fw_total_sent;
+		tstats->fw_total_fwd = rsp_tstats->fw_total_fwd;
+		tstats->fw_total_mcast_sent = rsp_tstats->fw_total_mcast_sent;
+		tstats->fw_total_bcast_sent = rsp_tstats->fw_total_bcast_sent;
+		tstats->fw_err_pko = rsp_tstats->fw_err_pko;
+		tstats->fw_err_pki = rsp_tstats->fw_err_pki;
+		tstats->fw_err_link = rsp_tstats->fw_err_link;
+		tstats->fw_err_drop = rsp_tstats->fw_err_drop;
+		tstats->fw_tso = rsp_tstats->fw_tso;
+		tstats->fw_tso_fwd = rsp_tstats->fw_tso_fwd;
+		tstats->fw_err_tso = rsp_tstats->fw_err_tso;
+		tstats->fw_tx_vxlan = rsp_tstats->fw_tx_vxlan;
+
+		resp->status = 1;
+	} else {
+		dev_err(&oct_dev->pci_dev->dev, "sc OPCODE_NIC_PORT_STATS command failed\n");
+		resp->status = -1;
+	}
+}
+
+static int lio_fetch_vf_stats(struct lio *lio)
+{
+	struct octeon_device *oct_dev = lio->oct_dev;
+	struct octeon_soft_command *sc;
+	struct oct_nic_vf_stats_resp *resp;
+
+	int retval;
+
+	/* Alloc soft command */
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct_dev,
+					  0,
+					  sizeof(struct oct_nic_vf_stats_resp),
+					  0);
+
+	if (!sc) {
+		dev_err(&oct_dev->pci_dev->dev, "Soft command allocation failed\n");
+		retval = -ENOMEM;
+		goto lio_fetch_vf_stats_exit;
+	}
+
+	resp = (struct oct_nic_vf_stats_resp *)sc->virtrptr;
+	memset(resp, 0, sizeof(struct oct_nic_vf_stats_resp));
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct_dev, sc, OPCODE_NIC,
+				    OPCODE_NIC_VF_PORT_STATS, 0, 0, 0);
+
+	retval = octeon_send_soft_command(oct_dev, sc);
+	if (retval == IQ_SEND_FAILED) {
+		octeon_free_soft_command(oct_dev, sc);
+		goto lio_fetch_vf_stats_exit;
+	}
+
+	retval =
+		wait_for_sc_completion_timeout(oct_dev, sc,
+					       (2 * LIO_SC_MAX_TMO_MS));
+	if (retval)  {
+		dev_err(&oct_dev->pci_dev->dev,
+			"sc OPCODE_NIC_VF_PORT_STATS command failed\n");
+		goto lio_fetch_vf_stats_exit;
+	}
+
+	if (sc->sc_status != OCTEON_REQUEST_TIMEOUT && !resp->status) {
+		octeon_swap_8B_data((u64 *)&resp->spoofmac_cnt,
+				    (sizeof(u64)) >> 3);
+
+		if (resp->spoofmac_cnt != 0) {
+			dev_warn(&oct_dev->pci_dev->dev,
+				 "%llu Spoofed packets detected\n",
+				 resp->spoofmac_cnt);
+		}
+	}
+	WRITE_ONCE(sc->caller_is_done, 1);
+
+lio_fetch_vf_stats_exit:
+	return retval;
+}
+
+void lio_fetch_stats(struct work_struct *work)
+{
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct lio *lio = wk->ctxptr;
+	struct octeon_device *oct_dev = lio->oct_dev;
+	struct octeon_soft_command *sc;
+	struct oct_nic_stats_resp *resp;
+	unsigned long time_in_jiffies;
+	int retval;
+
+	if (OCTEON_CN23XX_PF(oct_dev)) {
+		/* report spoofchk every 2 seconds */
+		if (!(oct_dev->vfstats_poll % LIO_VFSTATS_POLL) &&
+		    (oct_dev->fw_info.app_cap_flags & LIQUIDIO_SPOOFCHK_CAP) &&
+		    oct_dev->sriov_info.num_vfs_alloced) {
+			lio_fetch_vf_stats(lio);
+		}
+
+		oct_dev->vfstats_poll++;
+	}
+
+	/* Alloc soft command */
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct_dev,
+					  0,
+					  sizeof(struct oct_nic_stats_resp),
+					  0);
+
+	if (!sc) {
+		dev_err(&oct_dev->pci_dev->dev, "Soft command allocation failed\n");
+		goto lio_fetch_stats_exit;
+	}
+
+	resp = (struct oct_nic_stats_resp *)sc->virtrptr;
+	memset(resp, 0, sizeof(struct oct_nic_stats_resp));
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct_dev, sc, OPCODE_NIC,
+				    OPCODE_NIC_PORT_STATS, 0, 0, 0);
+
+	retval = octeon_send_soft_command(oct_dev, sc);
+	if (retval == IQ_SEND_FAILED) {
+		octeon_free_soft_command(oct_dev, sc);
+		goto lio_fetch_stats_exit;
+	}
+
+	retval = wait_for_sc_completion_timeout(oct_dev, sc,
+						(2 * LIO_SC_MAX_TMO_MS));
+	if (retval)  {
+		dev_err(&oct_dev->pci_dev->dev, "sc OPCODE_NIC_PORT_STATS command failed\n");
+		goto lio_fetch_stats_exit;
+	}
+
+	octnet_nic_stats_callback(oct_dev, sc->sc_status, sc);
+	WRITE_ONCE(sc->caller_is_done, true);
+
+lio_fetch_stats_exit:
+	time_in_jiffies = msecs_to_jiffies(LIQUIDIO_NDEV_STATS_POLL_TIME_MS);
+	if (ifstate_check(lio, LIO_IFSTATE_RUNNING))
+		schedule_delayed_work(&lio->stats_wk.work, time_in_jiffies);
+
+	return;
+}
+
+int liquidio_set_speed(struct lio *lio, int speed)
+{
+	struct octeon_device *oct = lio->oct_dev;
+	struct oct_nic_seapi_resp *resp;
+	struct octeon_soft_command *sc;
+	union octnet_cmd *ncmd;
+	int retval;
+	u32 var;
+
+	if (oct->speed_setting == speed)
+		return 0;
+
+	if (!OCTEON_CN23XX_PF(oct)) {
+		dev_err(&oct->pci_dev->dev, "%s: SET SPEED only for PF\n",
+			__func__);
+		return -EOPNOTSUPP;
+	}
+
+	sc = octeon_alloc_soft_command(oct, OCTNET_CMD_SIZE,
+				       sizeof(struct oct_nic_seapi_resp),
+				       0);
+	if (!sc)
+		return -ENOMEM;
+
+	ncmd = sc->virtdptr;
+	resp = sc->virtrptr;
+	memset(resp, 0, sizeof(struct oct_nic_seapi_resp));
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	ncmd->u64 = 0;
+	ncmd->s.cmd = SEAPI_CMD_SPEED_SET;
+	ncmd->s.param1 = speed;
+
+	octeon_swap_8B_data((u64 *)ncmd, (OCTNET_CMD_SIZE >> 3));
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC,
+				    OPCODE_NIC_UBOOT_CTL, 0, 0, 0);
+
+	retval = octeon_send_soft_command(oct, sc);
+	if (retval == IQ_SEND_FAILED) {
+		dev_info(&oct->pci_dev->dev, "Failed to send soft command\n");
+		octeon_free_soft_command(oct, sc);
+		retval = -EBUSY;
+	} else {
+		/* Wait for response or timeout */
+		retval = wait_for_sc_completion_timeout(oct, sc, 0);
+		if (retval)
+			return retval;
+
+		retval = resp->status;
+
+		if (retval) {
+			dev_err(&oct->pci_dev->dev, "%s failed, retval=%d\n",
+				__func__, retval);
+			WRITE_ONCE(sc->caller_is_done, true);
+
+			return -EIO;
+		}
+
+		var = be32_to_cpu((__force __be32)resp->speed);
+		if (var != speed) {
+			dev_err(&oct->pci_dev->dev,
+				"%s: setting failed speed= %x, expect %x\n",
+				__func__, var, speed);
+		}
+
+		oct->speed_setting = var;
+		WRITE_ONCE(sc->caller_is_done, true);
+	}
+
+	return retval;
+}
+
+int liquidio_get_speed(struct lio *lio)
+{
+	struct octeon_device *oct = lio->oct_dev;
+	struct oct_nic_seapi_resp *resp;
+	struct octeon_soft_command *sc;
+	union octnet_cmd *ncmd;
+	int retval;
+
+	sc = octeon_alloc_soft_command(oct, OCTNET_CMD_SIZE,
+				       sizeof(struct oct_nic_seapi_resp),
+				       0);
+	if (!sc)
+		return -ENOMEM;
+
+	ncmd = sc->virtdptr;
+	resp = sc->virtrptr;
+	memset(resp, 0, sizeof(struct oct_nic_seapi_resp));
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	ncmd->u64 = 0;
+	ncmd->s.cmd = SEAPI_CMD_SPEED_GET;
+
+	octeon_swap_8B_data((u64 *)ncmd, (OCTNET_CMD_SIZE >> 3));
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC,
+				    OPCODE_NIC_UBOOT_CTL, 0, 0, 0);
+
+	retval = octeon_send_soft_command(oct, sc);
+	if (retval == IQ_SEND_FAILED) {
+		dev_info(&oct->pci_dev->dev, "Failed to send soft command\n");
+		octeon_free_soft_command(oct, sc);
+		retval = -EIO;
+	} else {
+		retval = wait_for_sc_completion_timeout(oct, sc, 0);
+		if (retval)
+			return retval;
+
+		retval = resp->status;
+		if (retval) {
+			dev_err(&oct->pci_dev->dev,
+				"%s failed retval=%d\n", __func__, retval);
+			retval = -EIO;
+		} else {
+			u32 var;
+
+			var = be32_to_cpu((__force __be32)resp->speed);
+			oct->speed_setting = var;
+			if (var == 0xffff) {
+				/* unable to access boot variables
+				 * get the default value based on the NIC type
+				 */
+				if (oct->subsystem_id ==
+						OCTEON_CN2350_25GB_SUBSYS_ID ||
+				    oct->subsystem_id ==
+						OCTEON_CN2360_25GB_SUBSYS_ID) {
+					oct->no_speed_setting = 1;
+					oct->speed_setting = 25;
+				} else {
+					oct->speed_setting = 10;
+				}
+			}
+
+		}
+		WRITE_ONCE(sc->caller_is_done, true);
+	}
+
+	return retval;
+}
+
+int liquidio_set_fec(struct lio *lio, int on_off)
+{
+	struct oct_nic_seapi_resp *resp;
+	struct octeon_soft_command *sc;
+	struct octeon_device *oct;
+	union octnet_cmd *ncmd;
+	int retval;
+	u32 var;
+
+	oct = lio->oct_dev;
+
+	if (oct->props[lio->ifidx].fec == on_off)
+		return 0;
+
+	if (!OCTEON_CN23XX_PF(oct)) {
+		dev_err(&oct->pci_dev->dev, "%s: SET FEC only for PF\n",
+			__func__);
+		return -1;
+	}
+
+	if (oct->speed_boot != 25)  {
+		dev_err(&oct->pci_dev->dev,
+			"Set FEC only when link speed is 25G during insmod\n");
+		return -1;
+	}
+
+	sc = octeon_alloc_soft_command(oct, OCTNET_CMD_SIZE,
+				       sizeof(struct oct_nic_seapi_resp), 0);
+	if (!sc) {
+		dev_err(&oct->pci_dev->dev,
+			"Failed to allocate soft command\n");
+		return -ENOMEM;
+	}
+
+	ncmd = sc->virtdptr;
+	resp = sc->virtrptr;
+	memset(resp, 0, sizeof(struct oct_nic_seapi_resp));
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	ncmd->u64 = 0;
+	ncmd->s.cmd = SEAPI_CMD_FEC_SET;
+	ncmd->s.param1 = on_off;
+	/* SEAPI_CMD_FEC_DISABLE(0) or SEAPI_CMD_FEC_RS(1) */
+
+	octeon_swap_8B_data((u64 *)ncmd, (OCTNET_CMD_SIZE >> 3));
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC,
+				    OPCODE_NIC_UBOOT_CTL, 0, 0, 0);
+
+	retval = octeon_send_soft_command(oct, sc);
+	if (retval == IQ_SEND_FAILED) {
+		dev_info(&oct->pci_dev->dev, "Failed to send soft command\n");
+		octeon_free_soft_command(oct, sc);
+		return -EIO;
+	}
+
+	retval = wait_for_sc_completion_timeout(oct, sc, 0);
+	if (retval)
+		return (-EIO);
+
+	var = be32_to_cpu(resp->fec_setting);
+	resp->fec_setting = var;
+	if (var != on_off) {
+		dev_err(&oct->pci_dev->dev,
+			"Setting failed fec= %x, expect %x\n",
+			var, on_off);
+		oct->props[lio->ifidx].fec = var;
+		if (resp->fec_setting == SEAPI_CMD_FEC_SET_RS)
+			oct->props[lio->ifidx].fec = 1;
+		else
+			oct->props[lio->ifidx].fec = 0;
+	}
+
+	WRITE_ONCE(sc->caller_is_done, true);
+
+	if (oct->props[lio->ifidx].fec !=
+	    oct->props[lio->ifidx].fec_boot) {
+		dev_dbg(&oct->pci_dev->dev,
+			"Reload driver to change fec to %s\n",
+			oct->props[lio->ifidx].fec ? "on" : "off");
+	}
+
+	return retval;
+}
+
+int liquidio_get_fec(struct lio *lio)
+{
+	struct oct_nic_seapi_resp *resp;
+	struct octeon_soft_command *sc;
+	struct octeon_device *oct;
+	union octnet_cmd *ncmd;
+	int retval;
+	u32 var;
+
+	oct = lio->oct_dev;
+
+	sc = octeon_alloc_soft_command(oct, OCTNET_CMD_SIZE,
+				       sizeof(struct oct_nic_seapi_resp), 0);
+	if (!sc)
+		return -ENOMEM;
+
+	ncmd = sc->virtdptr;
+	resp = sc->virtrptr;
+	memset(resp, 0, sizeof(struct oct_nic_seapi_resp));
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	ncmd->u64 = 0;
+	ncmd->s.cmd = SEAPI_CMD_FEC_GET;
+
+	octeon_swap_8B_data((u64 *)ncmd, (OCTNET_CMD_SIZE >> 3));
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC,
+				    OPCODE_NIC_UBOOT_CTL, 0, 0, 0);
+
+	retval = octeon_send_soft_command(oct, sc);
+	if (retval == IQ_SEND_FAILED) {
+		dev_info(&oct->pci_dev->dev,
+			 "%s: Failed to send soft command\n", __func__);
+		octeon_free_soft_command(oct, sc);
+		return -EIO;
+	}
+
+	retval = wait_for_sc_completion_timeout(oct, sc, 0);
+	if (retval)
+		return retval;
+
+	var = be32_to_cpu(resp->fec_setting);
+	resp->fec_setting = var;
+	if (resp->fec_setting == SEAPI_CMD_FEC_SET_RS)
+		oct->props[lio->ifidx].fec = 1;
+	else
+		oct->props[lio->ifidx].fec = 0;
+
+	WRITE_ONCE(sc->caller_is_done, true);
+
+	if (oct->props[lio->ifidx].fec !=
+	    oct->props[lio->ifidx].fec_boot) {
+		dev_dbg(&oct->pci_dev->dev,
+			"Reload driver to change fec to %s\n",
+			oct->props[lio->ifidx].fec ? "on" : "off");
+	}
+
+	return retval;
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/lio_ethtool.c b/drivers/net/ethernet/cavium/liquidio/lio_ethtool.c
new file mode 100644
index 000000000..2c10ae3f7
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/lio_ethtool.c
@@ -0,0 +1,3182 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/ethtool.h>
+#include <linux/netdevice.h>
+#include <linux/net_tstamp.h>
+#include <linux/pci.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_nic.h"
+#include "octeon_main.h"
+#include "octeon_network.h"
+#include "cn66xx_regs.h"
+#include "cn66xx_device.h"
+#include "cn23xx_pf_device.h"
+#include "cn23xx_vf_device.h"
+
+static int lio_reset_queues(struct net_device *netdev, uint32_t num_qs);
+
+struct oct_intrmod_resp {
+	u64     rh;
+	struct oct_intrmod_cfg intrmod;
+	u64     status;
+};
+
+struct oct_mdio_cmd_resp {
+	u64 rh;
+	struct oct_mdio_cmd resp;
+	u64 status;
+};
+
+#define OCT_MDIO45_RESP_SIZE   (sizeof(struct oct_mdio_cmd_resp))
+
+/* Octeon's interface mode of operation */
+enum {
+	INTERFACE_MODE_DISABLED,
+	INTERFACE_MODE_RGMII,
+	INTERFACE_MODE_GMII,
+	INTERFACE_MODE_SPI,
+	INTERFACE_MODE_PCIE,
+	INTERFACE_MODE_XAUI,
+	INTERFACE_MODE_SGMII,
+	INTERFACE_MODE_PICMG,
+	INTERFACE_MODE_NPI,
+	INTERFACE_MODE_LOOP,
+	INTERFACE_MODE_SRIO,
+	INTERFACE_MODE_ILK,
+	INTERFACE_MODE_RXAUI,
+	INTERFACE_MODE_QSGMII,
+	INTERFACE_MODE_AGL,
+	INTERFACE_MODE_XLAUI,
+	INTERFACE_MODE_XFI,
+	INTERFACE_MODE_10G_KR,
+	INTERFACE_MODE_40G_KR4,
+	INTERFACE_MODE_MIXED,
+};
+
+#define OCT_ETHTOOL_REGDUMP_LEN  4096
+#define OCT_ETHTOOL_REGDUMP_LEN_23XX  (4096 * 11)
+#define OCT_ETHTOOL_REGDUMP_LEN_23XX_VF  (4096 * 2)
+#define OCT_ETHTOOL_REGSVER  1
+
+/* statistics of PF */
+static const char oct_stats_strings[][ETH_GSTRING_LEN] = {
+	"rx_packets",
+	"tx_packets",
+	"rx_bytes",
+	"tx_bytes",
+	"rx_errors",
+	"tx_errors",
+	"rx_dropped",
+	"tx_dropped",
+
+	"tx_total_sent",
+	"tx_total_fwd",
+	"tx_err_pko",
+	"tx_err_pki",
+	"tx_err_link",
+	"tx_err_drop",
+
+	"tx_tso",
+	"tx_tso_packets",
+	"tx_tso_err",
+	"tx_vxlan",
+
+	"tx_mcast",
+	"tx_bcast",
+
+	"mac_tx_total_pkts",
+	"mac_tx_total_bytes",
+	"mac_tx_mcast_pkts",
+	"mac_tx_bcast_pkts",
+	"mac_tx_ctl_packets",
+	"mac_tx_total_collisions",
+	"mac_tx_one_collision",
+	"mac_tx_multi_collision",
+	"mac_tx_max_collision_fail",
+	"mac_tx_max_deferral_fail",
+	"mac_tx_fifo_err",
+	"mac_tx_runts",
+
+	"rx_total_rcvd",
+	"rx_total_fwd",
+	"rx_mcast",
+	"rx_bcast",
+	"rx_jabber_err",
+	"rx_l2_err",
+	"rx_frame_err",
+	"rx_err_pko",
+	"rx_err_link",
+	"rx_err_drop",
+
+	"rx_vxlan",
+	"rx_vxlan_err",
+
+	"rx_lro_pkts",
+	"rx_lro_bytes",
+	"rx_total_lro",
+
+	"rx_lro_aborts",
+	"rx_lro_aborts_port",
+	"rx_lro_aborts_seq",
+	"rx_lro_aborts_tsval",
+	"rx_lro_aborts_timer",
+	"rx_fwd_rate",
+
+	"mac_rx_total_rcvd",
+	"mac_rx_bytes",
+	"mac_rx_total_bcst",
+	"mac_rx_total_mcst",
+	"mac_rx_runts",
+	"mac_rx_ctl_packets",
+	"mac_rx_fifo_err",
+	"mac_rx_dma_drop",
+	"mac_rx_fcs_err",
+
+	"link_state_changes",
+};
+
+/* statistics of VF */
+static const char oct_vf_stats_strings[][ETH_GSTRING_LEN] = {
+	"rx_packets",
+	"tx_packets",
+	"rx_bytes",
+	"tx_bytes",
+	"rx_errors",
+	"tx_errors",
+	"rx_dropped",
+	"tx_dropped",
+	"rx_mcast",
+	"tx_mcast",
+	"rx_bcast",
+	"tx_bcast",
+	"link_state_changes",
+};
+
+/* statistics of host tx queue */
+static const char oct_iq_stats_strings[][ETH_GSTRING_LEN] = {
+	"packets",
+	"bytes",
+	"dropped",
+	"iq_busy",
+	"sgentry_sent",
+
+	"fw_instr_posted",
+	"fw_instr_processed",
+	"fw_instr_dropped",
+	"fw_bytes_sent",
+
+	"tso",
+	"vxlan",
+	"txq_restart",
+};
+
+/* statistics of host rx queue */
+static const char oct_droq_stats_strings[][ETH_GSTRING_LEN] = {
+	"packets",
+	"bytes",
+	"dropped",
+	"dropped_nomem",
+	"dropped_toomany",
+	"fw_dropped",
+	"fw_pkts_received",
+	"fw_bytes_received",
+	"fw_dropped_nodispatch",
+
+	"vxlan",
+	"buffer_alloc_failure",
+};
+
+/* LiquidIO driver private flags */
+static const char oct_priv_flags_strings[][ETH_GSTRING_LEN] = {
+};
+
+#define OCTNIC_NCMD_AUTONEG_ON  0x1
+#define OCTNIC_NCMD_PHY_ON      0x2
+
+static int lio_get_link_ksettings(struct net_device *netdev,
+				  struct ethtool_link_ksettings *ecmd)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct oct_link_info *linfo;
+
+	linfo = &lio->linfo;
+
+	ethtool_link_ksettings_zero_link_mode(ecmd, supported);
+	ethtool_link_ksettings_zero_link_mode(ecmd, advertising);
+
+	switch (linfo->link.s.phy_type) {
+	case LIO_PHY_PORT_TP:
+		ecmd->base.port = PORT_TP;
+		ecmd->base.autoneg = AUTONEG_DISABLE;
+		ethtool_link_ksettings_add_link_mode(ecmd, supported, TP);
+		ethtool_link_ksettings_add_link_mode(ecmd, supported, Pause);
+		ethtool_link_ksettings_add_link_mode(ecmd, supported,
+						     10000baseT_Full);
+
+		ethtool_link_ksettings_add_link_mode(ecmd, advertising, Pause);
+		ethtool_link_ksettings_add_link_mode(ecmd, advertising,
+						     10000baseT_Full);
+
+		break;
+
+	case LIO_PHY_PORT_FIBRE:
+		if (linfo->link.s.if_mode == INTERFACE_MODE_XAUI ||
+		    linfo->link.s.if_mode == INTERFACE_MODE_RXAUI ||
+		    linfo->link.s.if_mode == INTERFACE_MODE_XLAUI ||
+		    linfo->link.s.if_mode == INTERFACE_MODE_XFI) {
+			dev_dbg(&oct->pci_dev->dev, "ecmd->base.transceiver is XCVR_EXTERNAL\n");
+			ecmd->base.transceiver = XCVR_EXTERNAL;
+		} else {
+			dev_err(&oct->pci_dev->dev, "Unknown link interface mode: %d\n",
+				linfo->link.s.if_mode);
+		}
+
+		ecmd->base.port = PORT_FIBRE;
+		ecmd->base.autoneg = AUTONEG_DISABLE;
+		ethtool_link_ksettings_add_link_mode(ecmd, supported, FIBRE);
+
+		ethtool_link_ksettings_add_link_mode(ecmd, supported, Pause);
+		ethtool_link_ksettings_add_link_mode(ecmd, advertising, Pause);
+		if (oct->subsystem_id == OCTEON_CN2350_25GB_SUBSYS_ID ||
+		    oct->subsystem_id == OCTEON_CN2360_25GB_SUBSYS_ID) {
+			if (OCTEON_CN23XX_PF(oct)) {
+				ethtool_link_ksettings_add_link_mode
+					(ecmd, supported, 25000baseSR_Full);
+				ethtool_link_ksettings_add_link_mode
+					(ecmd, supported, 25000baseKR_Full);
+				ethtool_link_ksettings_add_link_mode
+					(ecmd, supported, 25000baseCR_Full);
+
+				if (oct->no_speed_setting == 0)  {
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, supported,
+						 10000baseSR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, supported,
+						 10000baseKR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, supported,
+						 10000baseCR_Full);
+				}
+
+				if (oct->no_speed_setting == 0) {
+					liquidio_get_speed(lio);
+					liquidio_get_fec(lio);
+				} else {
+					oct->speed_setting = 25;
+				}
+
+				if (oct->speed_setting == 10) {
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 10000baseSR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 10000baseKR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 10000baseCR_Full);
+				}
+				if (oct->speed_setting == 25) {
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 25000baseSR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 25000baseKR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 25000baseCR_Full);
+				}
+
+				if (oct->no_speed_setting)
+					break;
+
+				ethtool_link_ksettings_add_link_mode
+					(ecmd, supported, FEC_RS);
+				ethtool_link_ksettings_add_link_mode
+					(ecmd, supported, FEC_NONE);
+					/*FEC_OFF*/
+				if (oct->props[lio->ifidx].fec == 1) {
+					/* ETHTOOL_FEC_RS */
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising, FEC_RS);
+				} else {
+					/* ETHTOOL_FEC_OFF */
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising, FEC_NONE);
+				}
+			} else { /* VF */
+				if (linfo->link.s.speed == 10000) {
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, supported,
+						 10000baseSR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, supported,
+						 10000baseKR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, supported,
+						 10000baseCR_Full);
+
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 10000baseSR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 10000baseKR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 10000baseCR_Full);
+				}
+
+				if (linfo->link.s.speed == 25000) {
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, supported,
+						 25000baseSR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, supported,
+						 25000baseKR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, supported,
+						 25000baseCR_Full);
+
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 25000baseSR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 25000baseKR_Full);
+					ethtool_link_ksettings_add_link_mode
+						(ecmd, advertising,
+						 25000baseCR_Full);
+				}
+			}
+		} else {
+			ethtool_link_ksettings_add_link_mode(ecmd, supported,
+							     10000baseT_Full);
+			ethtool_link_ksettings_add_link_mode(ecmd, advertising,
+							     10000baseT_Full);
+		}
+		break;
+	}
+
+	if (linfo->link.s.link_up) {
+		ecmd->base.speed = linfo->link.s.speed;
+		ecmd->base.duplex = linfo->link.s.duplex;
+	} else {
+		ecmd->base.speed = SPEED_UNKNOWN;
+		ecmd->base.duplex = DUPLEX_UNKNOWN;
+	}
+
+	return 0;
+}
+
+static int lio_set_link_ksettings(struct net_device *netdev,
+				  const struct ethtool_link_ksettings *ecmd)
+{
+	const int speed = ecmd->base.speed;
+	struct lio *lio = GET_LIO(netdev);
+	struct oct_link_info *linfo;
+	struct octeon_device *oct;
+
+	oct = lio->oct_dev;
+
+	linfo = &lio->linfo;
+
+	if (!(oct->subsystem_id == OCTEON_CN2350_25GB_SUBSYS_ID ||
+	      oct->subsystem_id == OCTEON_CN2360_25GB_SUBSYS_ID))
+		return -EOPNOTSUPP;
+
+	if (oct->no_speed_setting) {
+		dev_err(&oct->pci_dev->dev, "%s: Changing speed is not supported\n",
+			__func__);
+		return -EOPNOTSUPP;
+	}
+
+	if ((ecmd->base.duplex != DUPLEX_UNKNOWN &&
+	     ecmd->base.duplex != linfo->link.s.duplex) ||
+	     ecmd->base.autoneg != AUTONEG_DISABLE ||
+	    (ecmd->base.speed != 10000 && ecmd->base.speed != 25000 &&
+	     ecmd->base.speed != SPEED_UNKNOWN))
+		return -EOPNOTSUPP;
+
+	if ((oct->speed_boot == speed / 1000) &&
+	    oct->speed_boot == oct->speed_setting)
+		return 0;
+
+	liquidio_set_speed(lio, speed / 1000);
+
+	dev_dbg(&oct->pci_dev->dev, "Port speed is set to %dG\n",
+		oct->speed_setting);
+
+	return 0;
+}
+
+static void
+lio_get_drvinfo(struct net_device *netdev, struct ethtool_drvinfo *drvinfo)
+{
+	struct lio *lio;
+	struct octeon_device *oct;
+
+	lio = GET_LIO(netdev);
+	oct = lio->oct_dev;
+
+	memset(drvinfo, 0, sizeof(struct ethtool_drvinfo));
+	strcpy(drvinfo->driver, "liquidio");
+	strncpy(drvinfo->fw_version, oct->fw_info.liquidio_firmware_version,
+		ETHTOOL_FWVERS_LEN);
+	strncpy(drvinfo->bus_info, pci_name(oct->pci_dev), 32);
+}
+
+static void
+lio_get_vf_drvinfo(struct net_device *netdev, struct ethtool_drvinfo *drvinfo)
+{
+	struct octeon_device *oct;
+	struct lio *lio;
+
+	lio = GET_LIO(netdev);
+	oct = lio->oct_dev;
+
+	memset(drvinfo, 0, sizeof(struct ethtool_drvinfo));
+	strcpy(drvinfo->driver, "liquidio_vf");
+	strncpy(drvinfo->fw_version, oct->fw_info.liquidio_firmware_version,
+		ETHTOOL_FWVERS_LEN);
+	strncpy(drvinfo->bus_info, pci_name(oct->pci_dev), 32);
+}
+
+static int
+lio_send_queue_count_update(struct net_device *netdev, uint32_t num_queues)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_QUEUE_COUNT_CTL;
+	nctrl.ncmd.s.param1 = num_queues;
+	nctrl.ncmd.s.param2 = num_queues;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "Failed to send Queue reset command (ret: 0x%x)\n",
+			ret);
+		return -1;
+	}
+
+	return 0;
+}
+
+static void
+lio_ethtool_get_channels(struct net_device *dev,
+			 struct ethtool_channels *channel)
+{
+	struct lio *lio = GET_LIO(dev);
+	struct octeon_device *oct = lio->oct_dev;
+	u32 max_rx = 0, max_tx = 0, tx_count = 0, rx_count = 0;
+	u32 combined_count = 0, max_combined = 0;
+
+	if (OCTEON_CN6XXX(oct)) {
+		struct octeon_config *conf6x = CHIP_CONF(oct, cn6xxx);
+
+		max_rx = CFG_GET_OQ_MAX_Q(conf6x);
+		max_tx = CFG_GET_IQ_MAX_Q(conf6x);
+		rx_count = CFG_GET_NUM_RXQS_NIC_IF(conf6x, lio->ifidx);
+		tx_count = CFG_GET_NUM_TXQS_NIC_IF(conf6x, lio->ifidx);
+	} else if (OCTEON_CN23XX_PF(oct)) {
+		if (oct->sriov_info.sriov_enabled) {
+			max_combined = lio->linfo.num_txpciq;
+		} else {
+			struct octeon_config *conf23_pf =
+				CHIP_CONF(oct, cn23xx_pf);
+
+			max_combined = CFG_GET_IQ_MAX_Q(conf23_pf);
+		}
+		combined_count = oct->num_iqs;
+	} else if (OCTEON_CN23XX_VF(oct)) {
+		u64 reg_val = 0ULL;
+		u64 ctrl = CN23XX_VF_SLI_IQ_PKT_CONTROL64(0);
+
+		reg_val = octeon_read_csr64(oct, ctrl);
+		reg_val = reg_val >> CN23XX_PKT_INPUT_CTL_RPVF_POS;
+		max_combined = reg_val & CN23XX_PKT_INPUT_CTL_RPVF_MASK;
+		combined_count = oct->num_iqs;
+	}
+
+	channel->max_rx = max_rx;
+	channel->max_tx = max_tx;
+	channel->max_combined = max_combined;
+	channel->rx_count = rx_count;
+	channel->tx_count = tx_count;
+	channel->combined_count = combined_count;
+}
+
+static int
+lio_irq_reallocate_irqs(struct octeon_device *oct, uint32_t num_ioqs)
+{
+	struct msix_entry *msix_entries;
+	int num_msix_irqs = 0;
+	int i;
+
+	if (!oct->msix_on)
+		return 0;
+
+	/* Disable the input and output queues now. No more packets will
+	 * arrive from Octeon.
+	 */
+	oct->fn_list.disable_interrupt(oct, OCTEON_ALL_INTR);
+
+	if (oct->msix_on) {
+		if (OCTEON_CN23XX_PF(oct))
+			num_msix_irqs = oct->num_msix_irqs - 1;
+		else if (OCTEON_CN23XX_VF(oct))
+			num_msix_irqs = oct->num_msix_irqs;
+
+		msix_entries = (struct msix_entry *)oct->msix_entries;
+		for (i = 0; i < num_msix_irqs; i++) {
+			if (oct->ioq_vector[i].vector) {
+				/* clear the affinity_cpumask */
+				irq_set_affinity_hint(msix_entries[i].vector,
+						      NULL);
+				free_irq(msix_entries[i].vector,
+					 &oct->ioq_vector[i]);
+				oct->ioq_vector[i].vector = 0;
+			}
+		}
+
+		/* non-iov vector's argument is oct struct */
+		if (OCTEON_CN23XX_PF(oct))
+			free_irq(msix_entries[i].vector, oct);
+
+		pci_disable_msix(oct->pci_dev);
+		kfree(oct->msix_entries);
+		oct->msix_entries = NULL;
+	}
+
+	kfree(oct->irq_name_storage);
+	oct->irq_name_storage = NULL;
+
+	if (octeon_allocate_ioq_vector(oct, num_ioqs)) {
+		dev_err(&oct->pci_dev->dev, "OCTEON: ioq vector allocation failed\n");
+		return -1;
+	}
+
+	if (octeon_setup_interrupt(oct, num_ioqs)) {
+		dev_info(&oct->pci_dev->dev, "Setup interrupt failed\n");
+		return -1;
+	}
+
+	/* Enable Octeon device interrupts */
+	oct->fn_list.enable_interrupt(oct, OCTEON_ALL_INTR);
+
+	return 0;
+}
+
+static int
+lio_ethtool_set_channels(struct net_device *dev,
+			 struct ethtool_channels *channel)
+{
+	u32 combined_count, max_combined;
+	struct lio *lio = GET_LIO(dev);
+	struct octeon_device *oct = lio->oct_dev;
+	int stopped = 0;
+
+	if (strcmp(oct->fw_info.liquidio_firmware_version, "1.6.1") < 0) {
+		dev_err(&oct->pci_dev->dev, "Minimum firmware version required is 1.6.1\n");
+		return -EINVAL;
+	}
+
+	if (!channel->combined_count || channel->other_count ||
+	    channel->rx_count || channel->tx_count)
+		return -EINVAL;
+
+	combined_count = channel->combined_count;
+
+	if (OCTEON_CN23XX_PF(oct)) {
+		if (oct->sriov_info.sriov_enabled) {
+			max_combined = lio->linfo.num_txpciq;
+		} else {
+			struct octeon_config *conf23_pf =
+				CHIP_CONF(oct,
+					  cn23xx_pf);
+
+			max_combined =
+				CFG_GET_IQ_MAX_Q(conf23_pf);
+		}
+	} else if (OCTEON_CN23XX_VF(oct)) {
+		u64 reg_val = 0ULL;
+		u64 ctrl = CN23XX_VF_SLI_IQ_PKT_CONTROL64(0);
+
+		reg_val = octeon_read_csr64(oct, ctrl);
+		reg_val = reg_val >> CN23XX_PKT_INPUT_CTL_RPVF_POS;
+		max_combined = reg_val & CN23XX_PKT_INPUT_CTL_RPVF_MASK;
+	} else {
+		return -EINVAL;
+	}
+
+	if (combined_count > max_combined || combined_count < 1)
+		return -EINVAL;
+
+	if (combined_count == oct->num_iqs)
+		return 0;
+
+	ifstate_set(lio, LIO_IFSTATE_RESETTING);
+
+	if (netif_running(dev)) {
+		dev->netdev_ops->ndo_stop(dev);
+		stopped = 1;
+	}
+
+	if (lio_reset_queues(dev, combined_count))
+		return -EINVAL;
+
+	if (stopped)
+		dev->netdev_ops->ndo_open(dev);
+
+	ifstate_reset(lio, LIO_IFSTATE_RESETTING);
+
+	return 0;
+}
+
+static int lio_get_eeprom_len(struct net_device *netdev)
+{
+	u8 buf[192];
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct_dev = lio->oct_dev;
+	struct octeon_board_info *board_info;
+	int len;
+
+	board_info = (struct octeon_board_info *)(&oct_dev->boardinfo);
+	len = sprintf(buf, "boardname:%s serialnum:%s maj:%lld min:%lld\n",
+		      board_info->name, board_info->serial_number,
+		      board_info->major, board_info->minor);
+
+	return len;
+}
+
+static int
+lio_get_eeprom(struct net_device *netdev, struct ethtool_eeprom *eeprom,
+	       u8 *bytes)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct_dev = lio->oct_dev;
+	struct octeon_board_info *board_info;
+
+	if (eeprom->offset)
+		return -EINVAL;
+
+	eeprom->magic = oct_dev->pci_dev->vendor;
+	board_info = (struct octeon_board_info *)(&oct_dev->boardinfo);
+	sprintf((char *)bytes,
+		"boardname:%s serialnum:%s maj:%lld min:%lld\n",
+		board_info->name, board_info->serial_number,
+		board_info->major, board_info->minor);
+
+	return 0;
+}
+
+static int octnet_gpio_access(struct net_device *netdev, int addr, int val)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_GPIO_ACCESS;
+	nctrl.ncmd.s.param1 = addr;
+	nctrl.ncmd.s.param2 = val;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev,
+			"Failed to configure gpio value, ret=%d\n", ret);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int octnet_id_active(struct net_device *netdev, int val)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_ID_ACTIVE;
+	nctrl.ncmd.s.param1 = val;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev,
+			"Failed to configure gpio value, ret=%d\n", ret);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+/* This routine provides PHY access routines for
+ * mdio  clause45 .
+ */
+static int
+octnet_mdio45_access(struct lio *lio, int op, int loc, int *value)
+{
+	struct octeon_device *oct_dev = lio->oct_dev;
+	struct octeon_soft_command *sc;
+	struct oct_mdio_cmd_resp *mdio_cmd_rsp;
+	struct oct_mdio_cmd *mdio_cmd;
+	int retval = 0;
+
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct_dev,
+					  sizeof(struct oct_mdio_cmd),
+					  sizeof(struct oct_mdio_cmd_resp), 0);
+
+	if (!sc)
+		return -ENOMEM;
+
+	mdio_cmd_rsp = (struct oct_mdio_cmd_resp *)sc->virtrptr;
+	mdio_cmd = (struct oct_mdio_cmd *)sc->virtdptr;
+
+	mdio_cmd->op = op;
+	mdio_cmd->mdio_addr = loc;
+	if (op)
+		mdio_cmd->value1 = *value;
+	octeon_swap_8B_data((u64 *)mdio_cmd, sizeof(struct oct_mdio_cmd) / 8);
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct_dev, sc, OPCODE_NIC, OPCODE_NIC_MDIO45,
+				    0, 0, 0);
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	retval = octeon_send_soft_command(oct_dev, sc);
+	if (retval == IQ_SEND_FAILED) {
+		dev_err(&oct_dev->pci_dev->dev,
+			"octnet_mdio45_access instruction failed status: %x\n",
+			retval);
+		octeon_free_soft_command(oct_dev, sc);
+		return -EBUSY;
+	} else {
+		/* Sleep on a wait queue till the cond flag indicates that the
+		 * response arrived
+		 */
+		retval = wait_for_sc_completion_timeout(oct_dev, sc, 0);
+		if (retval)
+			return retval;
+
+		retval = mdio_cmd_rsp->status;
+		if (retval) {
+			dev_err(&oct_dev->pci_dev->dev,
+				"octnet mdio45 access failed: %x\n", retval);
+			WRITE_ONCE(sc->caller_is_done, true);
+			return -EBUSY;
+		}
+
+		octeon_swap_8B_data((u64 *)(&mdio_cmd_rsp->resp),
+				    sizeof(struct oct_mdio_cmd) / 8);
+
+		if (!op)
+			*value = mdio_cmd_rsp->resp.value1;
+
+		WRITE_ONCE(sc->caller_is_done, true);
+	}
+
+	return retval;
+}
+
+static int lio_set_phys_id(struct net_device *netdev,
+			   enum ethtool_phys_id_state state)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct oct_link_info *linfo;
+	int value, ret;
+	u32 cur_ver;
+
+	linfo = &lio->linfo;
+	cur_ver = OCT_FW_VER(oct->fw_info.ver.maj,
+			     oct->fw_info.ver.min,
+			     oct->fw_info.ver.rev);
+
+	switch (state) {
+	case ETHTOOL_ID_ACTIVE:
+		if (oct->chip_id == OCTEON_CN66XX) {
+			octnet_gpio_access(netdev, VITESSE_PHY_GPIO_CFG,
+					   VITESSE_PHY_GPIO_DRIVEON);
+			return 2;
+
+		} else if (oct->chip_id == OCTEON_CN68XX) {
+			/* Save the current LED settings */
+			ret = octnet_mdio45_access(lio, 0,
+						   LIO68XX_LED_BEACON_ADDR,
+						   &lio->phy_beacon_val);
+			if (ret)
+				return ret;
+
+			ret = octnet_mdio45_access(lio, 0,
+						   LIO68XX_LED_CTRL_ADDR,
+						   &lio->led_ctrl_val);
+			if (ret)
+				return ret;
+
+			/* Configure Beacon values */
+			value = LIO68XX_LED_BEACON_CFGON;
+			ret = octnet_mdio45_access(lio, 1,
+						   LIO68XX_LED_BEACON_ADDR,
+						   &value);
+			if (ret)
+				return ret;
+
+			value = LIO68XX_LED_CTRL_CFGON;
+			ret = octnet_mdio45_access(lio, 1,
+						   LIO68XX_LED_CTRL_ADDR,
+						   &value);
+			if (ret)
+				return ret;
+		} else if (oct->chip_id == OCTEON_CN23XX_PF_VID) {
+			octnet_id_active(netdev, LED_IDENTIFICATION_ON);
+			if (linfo->link.s.phy_type == LIO_PHY_PORT_TP &&
+			    cur_ver > OCT_FW_VER(1, 7, 2))
+				return 2;
+			else
+				return 0;
+		} else {
+			return -EINVAL;
+		}
+		break;
+
+	case ETHTOOL_ID_ON:
+		if (oct->chip_id == OCTEON_CN23XX_PF_VID &&
+		    linfo->link.s.phy_type == LIO_PHY_PORT_TP &&
+		    cur_ver > OCT_FW_VER(1, 7, 2))
+			octnet_id_active(netdev, LED_IDENTIFICATION_ON);
+		else if (oct->chip_id == OCTEON_CN66XX)
+			octnet_gpio_access(netdev, VITESSE_PHY_GPIO_CFG,
+					   VITESSE_PHY_GPIO_HIGH);
+		else
+			return -EINVAL;
+
+		break;
+
+	case ETHTOOL_ID_OFF:
+		if (oct->chip_id == OCTEON_CN23XX_PF_VID &&
+		    linfo->link.s.phy_type == LIO_PHY_PORT_TP &&
+		    cur_ver > OCT_FW_VER(1, 7, 2))
+			octnet_id_active(netdev, LED_IDENTIFICATION_OFF);
+		else if (oct->chip_id == OCTEON_CN66XX)
+			octnet_gpio_access(netdev, VITESSE_PHY_GPIO_CFG,
+					   VITESSE_PHY_GPIO_LOW);
+		else
+			return -EINVAL;
+
+		break;
+
+	case ETHTOOL_ID_INACTIVE:
+		if (oct->chip_id == OCTEON_CN66XX) {
+			octnet_gpio_access(netdev, VITESSE_PHY_GPIO_CFG,
+					   VITESSE_PHY_GPIO_DRIVEOFF);
+		} else if (oct->chip_id == OCTEON_CN68XX) {
+			/* Restore LED settings */
+			ret = octnet_mdio45_access(lio, 1,
+						   LIO68XX_LED_CTRL_ADDR,
+						   &lio->led_ctrl_val);
+			if (ret)
+				return ret;
+
+			ret = octnet_mdio45_access(lio, 1,
+						   LIO68XX_LED_BEACON_ADDR,
+						   &lio->phy_beacon_val);
+			if (ret)
+				return ret;
+		} else if (oct->chip_id == OCTEON_CN23XX_PF_VID) {
+			octnet_id_active(netdev, LED_IDENTIFICATION_OFF);
+
+			return 0;
+		} else {
+			return -EINVAL;
+		}
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static void
+lio_ethtool_get_ringparam(struct net_device *netdev,
+			  struct ethtool_ringparam *ering,
+			  struct kernel_ethtool_ringparam *kernel_ering,
+			  struct netlink_ext_ack *extack)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	u32 tx_max_pending = 0, rx_max_pending = 0, tx_pending = 0,
+	    rx_pending = 0;
+
+	if (ifstate_check(lio, LIO_IFSTATE_RESETTING))
+		return;
+
+	if (OCTEON_CN6XXX(oct)) {
+		struct octeon_config *conf6x = CHIP_CONF(oct, cn6xxx);
+
+		tx_max_pending = CN6XXX_MAX_IQ_DESCRIPTORS;
+		rx_max_pending = CN6XXX_MAX_OQ_DESCRIPTORS;
+		rx_pending = CFG_GET_NUM_RX_DESCS_NIC_IF(conf6x, lio->ifidx);
+		tx_pending = CFG_GET_NUM_TX_DESCS_NIC_IF(conf6x, lio->ifidx);
+	} else if (OCTEON_CN23XX_PF(oct) || OCTEON_CN23XX_VF(oct)) {
+		tx_max_pending = CN23XX_MAX_IQ_DESCRIPTORS;
+		rx_max_pending = CN23XX_MAX_OQ_DESCRIPTORS;
+		rx_pending = oct->droq[0]->max_count;
+		tx_pending = oct->instr_queue[0]->max_count;
+	}
+
+	ering->tx_pending = tx_pending;
+	ering->tx_max_pending = tx_max_pending;
+	ering->rx_pending = rx_pending;
+	ering->rx_max_pending = rx_max_pending;
+	ering->rx_mini_pending = 0;
+	ering->rx_jumbo_pending = 0;
+	ering->rx_mini_max_pending = 0;
+	ering->rx_jumbo_max_pending = 0;
+}
+
+static int lio_23xx_reconfigure_queue_count(struct lio *lio)
+{
+	struct octeon_device *oct = lio->oct_dev;
+	u32 resp_size, data_size;
+	struct liquidio_if_cfg_resp *resp;
+	struct octeon_soft_command *sc;
+	union oct_nic_if_cfg if_cfg;
+	struct lio_version *vdata;
+	u32 ifidx_or_pfnum;
+	int retval;
+	int j;
+
+	resp_size = sizeof(struct liquidio_if_cfg_resp);
+	data_size = sizeof(struct lio_version);
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct, data_size,
+					  resp_size, 0);
+	if (!sc) {
+		dev_err(&oct->pci_dev->dev, "%s: Failed to allocate soft command\n",
+			__func__);
+		return -1;
+	}
+
+	resp = (struct liquidio_if_cfg_resp *)sc->virtrptr;
+	vdata = (struct lio_version *)sc->virtdptr;
+
+	vdata->major = (__force u16)cpu_to_be16(LIQUIDIO_BASE_MAJOR_VERSION);
+	vdata->minor = (__force u16)cpu_to_be16(LIQUIDIO_BASE_MINOR_VERSION);
+	vdata->micro = (__force u16)cpu_to_be16(LIQUIDIO_BASE_MICRO_VERSION);
+
+	ifidx_or_pfnum = oct->pf_num;
+
+	if_cfg.u64 = 0;
+	if_cfg.s.num_iqueues = oct->sriov_info.num_pf_rings;
+	if_cfg.s.num_oqueues = oct->sriov_info.num_pf_rings;
+	if_cfg.s.base_queue = oct->sriov_info.pf_srn;
+	if_cfg.s.gmx_port_id = oct->pf_num;
+
+	sc->iq_no = 0;
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC,
+				    OPCODE_NIC_QCOUNT_UPDATE, 0,
+				    if_cfg.u64, 0);
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	retval = octeon_send_soft_command(oct, sc);
+	if (retval == IQ_SEND_FAILED) {
+		dev_err(&oct->pci_dev->dev,
+			"Sending iq/oq config failed status: %x\n",
+			retval);
+		octeon_free_soft_command(oct, sc);
+		return -EIO;
+	}
+
+	retval = wait_for_sc_completion_timeout(oct, sc, 0);
+	if (retval)
+		return retval;
+
+	retval = resp->status;
+	if (retval) {
+		dev_err(&oct->pci_dev->dev,
+			"iq/oq config failed: %x\n", retval);
+		WRITE_ONCE(sc->caller_is_done, true);
+		return -1;
+	}
+
+	octeon_swap_8B_data((u64 *)(&resp->cfg_info),
+			    (sizeof(struct liquidio_if_cfg_info)) >> 3);
+
+	lio->ifidx = ifidx_or_pfnum;
+	lio->linfo.num_rxpciq = hweight64(resp->cfg_info.iqmask);
+	lio->linfo.num_txpciq = hweight64(resp->cfg_info.iqmask);
+	for (j = 0; j < lio->linfo.num_rxpciq; j++) {
+		lio->linfo.rxpciq[j].u64 =
+			resp->cfg_info.linfo.rxpciq[j].u64;
+	}
+
+	for (j = 0; j < lio->linfo.num_txpciq; j++) {
+		lio->linfo.txpciq[j].u64 =
+			resp->cfg_info.linfo.txpciq[j].u64;
+	}
+
+	lio->linfo.hw_addr = resp->cfg_info.linfo.hw_addr;
+	lio->linfo.gmxport = resp->cfg_info.linfo.gmxport;
+	lio->linfo.link.u64 = resp->cfg_info.linfo.link.u64;
+	lio->txq = lio->linfo.txpciq[0].s.q_no;
+	lio->rxq = lio->linfo.rxpciq[0].s.q_no;
+
+	dev_info(&oct->pci_dev->dev, "Queue count updated to %d\n",
+		 lio->linfo.num_rxpciq);
+
+	WRITE_ONCE(sc->caller_is_done, true);
+
+	return 0;
+}
+
+static int lio_reset_queues(struct net_device *netdev, uint32_t num_qs)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	int i, queue_count_update = 0;
+	struct napi_struct *napi, *n;
+	int ret;
+
+	schedule_timeout_uninterruptible(msecs_to_jiffies(100));
+
+	if (wait_for_pending_requests(oct))
+		dev_err(&oct->pci_dev->dev, "There were pending requests\n");
+
+	if (lio_wait_for_instr_fetch(oct))
+		dev_err(&oct->pci_dev->dev, "IQ had pending instructions\n");
+
+	if (octeon_set_io_queues_off(oct)) {
+		dev_err(&oct->pci_dev->dev, "Setting io queues off failed\n");
+		return -1;
+	}
+
+	/* Disable the input and output queues now. No more packets will
+	 * arrive from Octeon.
+	 */
+	oct->fn_list.disable_io_queues(oct);
+	/* Delete NAPI */
+	list_for_each_entry_safe(napi, n, &netdev->napi_list, dev_list)
+		netif_napi_del(napi);
+
+	if (num_qs != oct->num_iqs) {
+		ret = netif_set_real_num_rx_queues(netdev, num_qs);
+		if (ret) {
+			dev_err(&oct->pci_dev->dev,
+				"Setting real number rx failed\n");
+			return ret;
+		}
+
+		ret = netif_set_real_num_tx_queues(netdev, num_qs);
+		if (ret) {
+			dev_err(&oct->pci_dev->dev,
+				"Setting real number tx failed\n");
+			return ret;
+		}
+
+		/* The value of queue_count_update decides whether it is the
+		 * queue count or the descriptor count that is being
+		 * re-configured.
+		 */
+		queue_count_update = 1;
+	}
+
+	/* Re-configuration of queues can happen in two scenarios, SRIOV enabled
+	 * and SRIOV disabled. Few things like recreating queue zero, resetting
+	 * glists and IRQs are required for both. For the latter, some more
+	 * steps like updating sriov_info for the octeon device need to be done.
+	 */
+	if (queue_count_update) {
+		cleanup_rx_oom_poll_fn(netdev);
+
+		lio_delete_glists(lio);
+
+		/* Delete mbox for PF which is SRIOV disabled because sriov_info
+		 * will be now changed.
+		 */
+		if ((OCTEON_CN23XX_PF(oct)) && !oct->sriov_info.sriov_enabled)
+			oct->fn_list.free_mbox(oct);
+	}
+
+	for (i = 0; i < MAX_OCTEON_OUTPUT_QUEUES(oct); i++) {
+		if (!(oct->io_qmask.oq & BIT_ULL(i)))
+			continue;
+		octeon_delete_droq(oct, i);
+	}
+
+	for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct); i++) {
+		if (!(oct->io_qmask.iq & BIT_ULL(i)))
+			continue;
+		octeon_delete_instr_queue(oct, i);
+	}
+
+	if (queue_count_update) {
+		/* For PF re-configure sriov related information */
+		if ((OCTEON_CN23XX_PF(oct)) &&
+		    !oct->sriov_info.sriov_enabled) {
+			oct->sriov_info.num_pf_rings = num_qs;
+			if (cn23xx_sriov_config(oct)) {
+				dev_err(&oct->pci_dev->dev,
+					"Queue reset aborted: SRIOV config failed\n");
+				return -1;
+			}
+
+			num_qs = oct->sriov_info.num_pf_rings;
+		}
+	}
+
+	if (oct->fn_list.setup_device_regs(oct)) {
+		dev_err(&oct->pci_dev->dev, "Failed to configure device registers\n");
+		return -1;
+	}
+
+	/* The following are needed in case of queue count re-configuration and
+	 * not for descriptor count re-configuration.
+	 */
+	if (queue_count_update) {
+		if (octeon_setup_instr_queues(oct))
+			return -1;
+
+		if (octeon_setup_output_queues(oct))
+			return -1;
+
+		/* Recreating mbox for PF that is SRIOV disabled */
+		if (OCTEON_CN23XX_PF(oct) && !oct->sriov_info.sriov_enabled) {
+			if (oct->fn_list.setup_mbox(oct)) {
+				dev_err(&oct->pci_dev->dev, "Mailbox setup failed\n");
+				return -1;
+			}
+		}
+
+		/* Deleting and recreating IRQs whether the interface is SRIOV
+		 * enabled or disabled.
+		 */
+		if (lio_irq_reallocate_irqs(oct, num_qs)) {
+			dev_err(&oct->pci_dev->dev, "IRQs could not be allocated\n");
+			return -1;
+		}
+
+		/* Enable the input and output queues for this Octeon device */
+		if (oct->fn_list.enable_io_queues(oct)) {
+			dev_err(&oct->pci_dev->dev, "Failed to enable input/output queues\n");
+			return -1;
+		}
+
+		for (i = 0; i < oct->num_oqs; i++)
+			writel(oct->droq[i]->max_count,
+			       oct->droq[i]->pkts_credit_reg);
+
+		/* Informing firmware about the new queue count. It is required
+		 * for firmware to allocate more number of queues than those at
+		 * load time.
+		 */
+		if (OCTEON_CN23XX_PF(oct) && !oct->sriov_info.sriov_enabled) {
+			if (lio_23xx_reconfigure_queue_count(lio))
+				return -1;
+		}
+	}
+
+	/* Once firmware is aware of the new value, queues can be recreated */
+	if (liquidio_setup_io_queues(oct, 0, num_qs, num_qs)) {
+		dev_err(&oct->pci_dev->dev, "I/O queues creation failed\n");
+		return -1;
+	}
+
+	if (queue_count_update) {
+		if (lio_setup_glists(oct, lio, num_qs)) {
+			dev_err(&oct->pci_dev->dev, "Gather list allocation failed\n");
+			return -1;
+		}
+
+		if (setup_rx_oom_poll_fn(netdev)) {
+			dev_err(&oct->pci_dev->dev, "lio_setup_rx_oom_poll_fn failed\n");
+			return 1;
+		}
+
+		/* Send firmware the information about new number of queues
+		 * if the interface is a VF or a PF that is SRIOV enabled.
+		 */
+		if (oct->sriov_info.sriov_enabled || OCTEON_CN23XX_VF(oct))
+			if (lio_send_queue_count_update(netdev, num_qs))
+				return -1;
+	}
+
+	return 0;
+}
+
+static int
+lio_ethtool_set_ringparam(struct net_device *netdev,
+			  struct ethtool_ringparam *ering,
+			  struct kernel_ethtool_ringparam *kernel_ering,
+			  struct netlink_ext_ack *extack)
+{
+	u32 rx_count, tx_count, rx_count_old, tx_count_old;
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	int stopped = 0;
+
+	if (!OCTEON_CN23XX_PF(oct) && !OCTEON_CN23XX_VF(oct))
+		return -EINVAL;
+
+	if (ering->rx_mini_pending || ering->rx_jumbo_pending)
+		return -EINVAL;
+
+	rx_count = clamp_t(u32, ering->rx_pending, CN23XX_MIN_OQ_DESCRIPTORS,
+			   CN23XX_MAX_OQ_DESCRIPTORS);
+	tx_count = clamp_t(u32, ering->tx_pending, CN23XX_MIN_IQ_DESCRIPTORS,
+			   CN23XX_MAX_IQ_DESCRIPTORS);
+
+	rx_count_old = oct->droq[0]->max_count;
+	tx_count_old = oct->instr_queue[0]->max_count;
+
+	if (rx_count == rx_count_old && tx_count == tx_count_old)
+		return 0;
+
+	ifstate_set(lio, LIO_IFSTATE_RESETTING);
+
+	if (netif_running(netdev)) {
+		netdev->netdev_ops->ndo_stop(netdev);
+		stopped = 1;
+	}
+
+	/* Change RX/TX DESCS  count */
+	if (tx_count != tx_count_old)
+		CFG_SET_NUM_TX_DESCS_NIC_IF(octeon_get_conf(oct), lio->ifidx,
+					    tx_count);
+	if (rx_count != rx_count_old)
+		CFG_SET_NUM_RX_DESCS_NIC_IF(octeon_get_conf(oct), lio->ifidx,
+					    rx_count);
+
+	if (lio_reset_queues(netdev, oct->num_iqs))
+		goto err_lio_reset_queues;
+
+	if (stopped)
+		netdev->netdev_ops->ndo_open(netdev);
+
+	ifstate_reset(lio, LIO_IFSTATE_RESETTING);
+
+	return 0;
+
+err_lio_reset_queues:
+	if (tx_count != tx_count_old)
+		CFG_SET_NUM_TX_DESCS_NIC_IF(octeon_get_conf(oct), lio->ifidx,
+					    tx_count_old);
+	if (rx_count != rx_count_old)
+		CFG_SET_NUM_RX_DESCS_NIC_IF(octeon_get_conf(oct), lio->ifidx,
+					    rx_count_old);
+	return -EINVAL;
+}
+
+static u32 lio_get_msglevel(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+
+	return lio->msg_enable;
+}
+
+static void lio_set_msglevel(struct net_device *netdev, u32 msglvl)
+{
+	struct lio *lio = GET_LIO(netdev);
+
+	if ((msglvl ^ lio->msg_enable) & NETIF_MSG_HW) {
+		if (msglvl & NETIF_MSG_HW)
+			liquidio_set_feature(netdev,
+					     OCTNET_CMD_VERBOSE_ENABLE, 0);
+		else
+			liquidio_set_feature(netdev,
+					     OCTNET_CMD_VERBOSE_DISABLE, 0);
+	}
+
+	lio->msg_enable = msglvl;
+}
+
+static void lio_vf_set_msglevel(struct net_device *netdev, u32 msglvl)
+{
+	struct lio *lio = GET_LIO(netdev);
+
+	lio->msg_enable = msglvl;
+}
+
+static void
+lio_get_pauseparam(struct net_device *netdev, struct ethtool_pauseparam *pause)
+{
+	/* Notes: Not supporting any auto negotiation in these
+	 * drivers. Just report pause frame support.
+	 */
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+
+	pause->autoneg = 0;
+
+	pause->tx_pause = oct->tx_pause;
+	pause->rx_pause = oct->rx_pause;
+}
+
+static int
+lio_set_pauseparam(struct net_device *netdev, struct ethtool_pauseparam *pause)
+{
+	/* Notes: Not supporting any auto negotiation in these
+	 * drivers.
+	 */
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	struct oct_link_info *linfo = &lio->linfo;
+
+	int ret = 0;
+
+	if (oct->chip_id != OCTEON_CN23XX_PF_VID)
+		return -EINVAL;
+
+	if (linfo->link.s.duplex == 0) {
+		/*no flow control for half duplex*/
+		if (pause->rx_pause || pause->tx_pause)
+			return -EINVAL;
+	}
+
+	/*do not support autoneg of link flow control*/
+	if (pause->autoneg == AUTONEG_ENABLE)
+		return -EINVAL;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_SET_FLOW_CTL;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	if (pause->rx_pause) {
+		/*enable rx pause*/
+		nctrl.ncmd.s.param1 = 1;
+	} else {
+		/*disable rx pause*/
+		nctrl.ncmd.s.param1 = 0;
+	}
+
+	if (pause->tx_pause) {
+		/*enable tx pause*/
+		nctrl.ncmd.s.param2 = 1;
+	} else {
+		/*disable tx pause*/
+		nctrl.ncmd.s.param2 = 0;
+	}
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev,
+			"Failed to set pause parameter, ret=%d\n", ret);
+		return -EINVAL;
+	}
+
+	oct->rx_pause = pause->rx_pause;
+	oct->tx_pause = pause->tx_pause;
+
+	return 0;
+}
+
+static void
+lio_get_ethtool_stats(struct net_device *netdev,
+		      struct ethtool_stats *stats  __attribute__((unused)),
+		      u64 *data)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct_dev = lio->oct_dev;
+	struct rtnl_link_stats64 lstats;
+	int i = 0, j;
+
+	if (ifstate_check(lio, LIO_IFSTATE_RESETTING))
+		return;
+
+	netdev->netdev_ops->ndo_get_stats64(netdev, &lstats);
+	/*sum of oct->droq[oq_no]->stats->rx_pkts_received */
+	data[i++] = lstats.rx_packets;
+	/*sum of oct->instr_queue[iq_no]->stats.tx_done */
+	data[i++] = lstats.tx_packets;
+	/*sum of oct->droq[oq_no]->stats->rx_bytes_received */
+	data[i++] = lstats.rx_bytes;
+	/*sum of oct->instr_queue[iq_no]->stats.tx_tot_bytes */
+	data[i++] = lstats.tx_bytes;
+	data[i++] = lstats.rx_errors +
+			oct_dev->link_stats.fromwire.fcs_err +
+			oct_dev->link_stats.fromwire.jabber_err +
+			oct_dev->link_stats.fromwire.l2_err +
+			oct_dev->link_stats.fromwire.frame_err;
+	data[i++] = lstats.tx_errors;
+	/*sum of oct->droq[oq_no]->stats->rx_dropped +
+	 *oct->droq[oq_no]->stats->dropped_nodispatch +
+	 *oct->droq[oq_no]->stats->dropped_toomany +
+	 *oct->droq[oq_no]->stats->dropped_nomem
+	 */
+	data[i++] = lstats.rx_dropped +
+			oct_dev->link_stats.fromwire.fifo_err +
+			oct_dev->link_stats.fromwire.dmac_drop +
+			oct_dev->link_stats.fromwire.red_drops +
+			oct_dev->link_stats.fromwire.fw_err_pko +
+			oct_dev->link_stats.fromwire.fw_err_link +
+			oct_dev->link_stats.fromwire.fw_err_drop;
+	/*sum of oct->instr_queue[iq_no]->stats.tx_dropped */
+	data[i++] = lstats.tx_dropped +
+			oct_dev->link_stats.fromhost.max_collision_fail +
+			oct_dev->link_stats.fromhost.max_deferral_fail +
+			oct_dev->link_stats.fromhost.total_collisions +
+			oct_dev->link_stats.fromhost.fw_err_pko +
+			oct_dev->link_stats.fromhost.fw_err_link +
+			oct_dev->link_stats.fromhost.fw_err_drop +
+			oct_dev->link_stats.fromhost.fw_err_pki;
+
+	/* firmware tx stats */
+	/*per_core_stats[cvmx_get_core_num()].link_stats[mdata->from_ifidx].
+	 *fromhost.fw_total_sent
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.fw_total_sent);
+	/*per_core_stats[i].link_stats[port].fromwire.fw_total_fwd */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.fw_total_fwd);
+	/*per_core_stats[j].link_stats[i].fromhost.fw_err_pko */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.fw_err_pko);
+	/*per_core_stats[j].link_stats[i].fromhost.fw_err_pki */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.fw_err_pki);
+	/*per_core_stats[j].link_stats[i].fromhost.fw_err_link */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.fw_err_link);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[idx].fromhost.
+	 *fw_err_drop
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.fw_err_drop);
+
+	/*per_core_stats[cvmx_get_core_num()].link_stats[idx].fromhost.fw_tso */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.fw_tso);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[idx].fromhost.
+	 *fw_tso_fwd
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.fw_tso_fwd);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[idx].fromhost.
+	 *fw_err_tso
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.fw_err_tso);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[idx].fromhost.
+	 *fw_tx_vxlan
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.fw_tx_vxlan);
+
+	/* Multicast packets sent by this port */
+	data[i++] = oct_dev->link_stats.fromhost.fw_total_mcast_sent;
+	data[i++] = oct_dev->link_stats.fromhost.fw_total_bcast_sent;
+
+	/* mac tx statistics */
+	/*CVMX_BGXX_CMRX_TX_STAT5 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.total_pkts_sent);
+	/*CVMX_BGXX_CMRX_TX_STAT4 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.total_bytes_sent);
+	/*CVMX_BGXX_CMRX_TX_STAT15 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.mcast_pkts_sent);
+	/*CVMX_BGXX_CMRX_TX_STAT14 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.bcast_pkts_sent);
+	/*CVMX_BGXX_CMRX_TX_STAT17 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.ctl_sent);
+	/*CVMX_BGXX_CMRX_TX_STAT0 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.total_collisions);
+	/*CVMX_BGXX_CMRX_TX_STAT3 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.one_collision_sent);
+	/*CVMX_BGXX_CMRX_TX_STAT2 */
+	data[i++] =
+		CVM_CAST64(oct_dev->link_stats.fromhost.multi_collision_sent);
+	/*CVMX_BGXX_CMRX_TX_STAT0 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.max_collision_fail);
+	/*CVMX_BGXX_CMRX_TX_STAT1 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.max_deferral_fail);
+	/*CVMX_BGXX_CMRX_TX_STAT16 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.fifo_err);
+	/*CVMX_BGXX_CMRX_TX_STAT6 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromhost.runts);
+
+	/* RX firmware stats */
+	/*per_core_stats[cvmx_get_core_num()].link_stats[ifidx].fromwire.
+	 *fw_total_rcvd
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_total_rcvd);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[ifidx].fromwire.
+	 *fw_total_fwd
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_total_fwd);
+	/* Multicast packets received on this port */
+	data[i++] = oct_dev->link_stats.fromwire.fw_total_mcast;
+	data[i++] = oct_dev->link_stats.fromwire.fw_total_bcast;
+	/*per_core_stats[core_id].link_stats[ifidx].fromwire.jabber_err */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.jabber_err);
+	/*per_core_stats[core_id].link_stats[ifidx].fromwire.l2_err */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.l2_err);
+	/*per_core_stats[core_id].link_stats[ifidx].fromwire.frame_err */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.frame_err);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[ifidx].fromwire.
+	 *fw_err_pko
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_err_pko);
+	/*per_core_stats[j].link_stats[i].fromwire.fw_err_link */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_err_link);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[lro_ctx->ifidx].
+	 *fromwire.fw_err_drop
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_err_drop);
+
+	/*per_core_stats[cvmx_get_core_num()].link_stats[lro_ctx->ifidx].
+	 *fromwire.fw_rx_vxlan
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_rx_vxlan);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[lro_ctx->ifidx].
+	 *fromwire.fw_rx_vxlan_err
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_rx_vxlan_err);
+
+	/* LRO */
+	/*per_core_stats[cvmx_get_core_num()].link_stats[ifidx].fromwire.
+	 *fw_lro_pkts
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_lro_pkts);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[ifidx].fromwire.
+	 *fw_lro_octs
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_lro_octs);
+	/*per_core_stats[j].link_stats[i].fromwire.fw_total_lro */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_total_lro);
+	/*per_core_stats[j].link_stats[i].fromwire.fw_lro_aborts */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_lro_aborts);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[ifidx].fromwire.
+	 *fw_lro_aborts_port
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_lro_aborts_port);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[ifidx].fromwire.
+	 *fw_lro_aborts_seq
+	 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fw_lro_aborts_seq);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[ifidx].fromwire.
+	 *fw_lro_aborts_tsval
+	 */
+	data[i++] =
+		CVM_CAST64(oct_dev->link_stats.fromwire.fw_lro_aborts_tsval);
+	/*per_core_stats[cvmx_get_core_num()].link_stats[ifidx].fromwire.
+	 *fw_lro_aborts_timer
+	 */
+	/* intrmod: packet forward rate */
+	data[i++] =
+		CVM_CAST64(oct_dev->link_stats.fromwire.fw_lro_aborts_timer);
+	/*per_core_stats[j].link_stats[i].fromwire.fw_lro_aborts */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fwd_rate);
+
+	/* mac: link-level stats */
+	/*CVMX_BGXX_CMRX_RX_STAT0 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.total_rcvd);
+	/*CVMX_BGXX_CMRX_RX_STAT1 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.bytes_rcvd);
+	/*CVMX_PKI_STATX_STAT5 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.total_bcst);
+	/*CVMX_PKI_STATX_STAT5 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.total_mcst);
+	/*wqe->word2.err_code or wqe->word2.err_level */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.runts);
+	/*CVMX_BGXX_CMRX_RX_STAT2 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.ctl_rcvd);
+	/*CVMX_BGXX_CMRX_RX_STAT6 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fifo_err);
+	/*CVMX_BGXX_CMRX_RX_STAT4 */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.dmac_drop);
+	/*wqe->word2.err_code or wqe->word2.err_level */
+	data[i++] = CVM_CAST64(oct_dev->link_stats.fromwire.fcs_err);
+	/*lio->link_changes*/
+	data[i++] = CVM_CAST64(lio->link_changes);
+
+	for (j = 0; j < MAX_OCTEON_INSTR_QUEUES(oct_dev); j++) {
+		if (!(oct_dev->io_qmask.iq & BIT_ULL(j)))
+			continue;
+		/*packets to network port*/
+		/*# of packets tx to network */
+		data[i++] = CVM_CAST64(oct_dev->instr_queue[j]->stats.tx_done);
+		/*# of bytes tx to network */
+		data[i++] =
+			CVM_CAST64(oct_dev->instr_queue[j]->stats.tx_tot_bytes);
+		/*# of packets dropped */
+		data[i++] =
+			CVM_CAST64(oct_dev->instr_queue[j]->stats.tx_dropped);
+		/*# of tx fails due to queue full */
+		data[i++] =
+			CVM_CAST64(oct_dev->instr_queue[j]->stats.tx_iq_busy);
+		/*XXX gather entries sent */
+		data[i++] =
+			CVM_CAST64(oct_dev->instr_queue[j]->stats.sgentry_sent);
+
+		/*instruction to firmware: data and control */
+		/*# of instructions to the queue */
+		data[i++] =
+			CVM_CAST64(oct_dev->instr_queue[j]->stats.instr_posted);
+		/*# of instructions processed */
+		data[i++] = CVM_CAST64(
+				oct_dev->instr_queue[j]->stats.instr_processed);
+		/*# of instructions could not be processed */
+		data[i++] = CVM_CAST64(
+				oct_dev->instr_queue[j]->stats.instr_dropped);
+		/*bytes sent through the queue */
+		data[i++] =
+			CVM_CAST64(oct_dev->instr_queue[j]->stats.bytes_sent);
+
+		/*tso request*/
+		data[i++] = CVM_CAST64(oct_dev->instr_queue[j]->stats.tx_gso);
+		/*vxlan request*/
+		data[i++] = CVM_CAST64(oct_dev->instr_queue[j]->stats.tx_vxlan);
+		/*txq restart*/
+		data[i++] =
+			CVM_CAST64(oct_dev->instr_queue[j]->stats.tx_restart);
+	}
+
+	/* RX */
+	for (j = 0; j < MAX_OCTEON_OUTPUT_QUEUES(oct_dev); j++) {
+		if (!(oct_dev->io_qmask.oq & BIT_ULL(j)))
+			continue;
+
+		/*packets send to TCP/IP network stack */
+		/*# of packets to network stack */
+		data[i++] =
+			CVM_CAST64(oct_dev->droq[j]->stats.rx_pkts_received);
+		/*# of bytes to network stack */
+		data[i++] =
+			CVM_CAST64(oct_dev->droq[j]->stats.rx_bytes_received);
+		/*# of packets dropped */
+		data[i++] = CVM_CAST64(oct_dev->droq[j]->stats.dropped_nomem +
+				       oct_dev->droq[j]->stats.dropped_toomany +
+				       oct_dev->droq[j]->stats.rx_dropped);
+		data[i++] =
+			CVM_CAST64(oct_dev->droq[j]->stats.dropped_nomem);
+		data[i++] =
+			CVM_CAST64(oct_dev->droq[j]->stats.dropped_toomany);
+		data[i++] =
+			CVM_CAST64(oct_dev->droq[j]->stats.rx_dropped);
+
+		/*control and data path*/
+		data[i++] =
+			CVM_CAST64(oct_dev->droq[j]->stats.pkts_received);
+		data[i++] =
+			CVM_CAST64(oct_dev->droq[j]->stats.bytes_received);
+		data[i++] =
+			CVM_CAST64(oct_dev->droq[j]->stats.dropped_nodispatch);
+
+		data[i++] =
+			CVM_CAST64(oct_dev->droq[j]->stats.rx_vxlan);
+		data[i++] =
+			CVM_CAST64(oct_dev->droq[j]->stats.rx_alloc_failure);
+	}
+}
+
+static void lio_vf_get_ethtool_stats(struct net_device *netdev,
+				     struct ethtool_stats *stats
+				     __attribute__((unused)),
+				     u64 *data)
+{
+	struct rtnl_link_stats64 lstats;
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct_dev = lio->oct_dev;
+	int i = 0, j, vj;
+
+	if (ifstate_check(lio, LIO_IFSTATE_RESETTING))
+		return;
+
+	netdev->netdev_ops->ndo_get_stats64(netdev, &lstats);
+	/* sum of oct->droq[oq_no]->stats->rx_pkts_received */
+	data[i++] = lstats.rx_packets;
+	/* sum of oct->instr_queue[iq_no]->stats.tx_done */
+	data[i++] = lstats.tx_packets;
+	/* sum of oct->droq[oq_no]->stats->rx_bytes_received */
+	data[i++] = lstats.rx_bytes;
+	/* sum of oct->instr_queue[iq_no]->stats.tx_tot_bytes */
+	data[i++] = lstats.tx_bytes;
+	data[i++] = lstats.rx_errors;
+	data[i++] = lstats.tx_errors;
+	 /* sum of oct->droq[oq_no]->stats->rx_dropped +
+	  * oct->droq[oq_no]->stats->dropped_nodispatch +
+	  * oct->droq[oq_no]->stats->dropped_toomany +
+	  * oct->droq[oq_no]->stats->dropped_nomem
+	  */
+	data[i++] = lstats.rx_dropped;
+	/* sum of oct->instr_queue[iq_no]->stats.tx_dropped */
+	data[i++] = lstats.tx_dropped +
+		oct_dev->link_stats.fromhost.fw_err_drop;
+
+	data[i++] = oct_dev->link_stats.fromwire.fw_total_mcast;
+	data[i++] = oct_dev->link_stats.fromhost.fw_total_mcast_sent;
+	data[i++] = oct_dev->link_stats.fromwire.fw_total_bcast;
+	data[i++] = oct_dev->link_stats.fromhost.fw_total_bcast_sent;
+
+	/* lio->link_changes */
+	data[i++] = CVM_CAST64(lio->link_changes);
+
+	for (vj = 0; vj < oct_dev->num_iqs; vj++) {
+		j = lio->linfo.txpciq[vj].s.q_no;
+
+		/* packets to network port */
+		/* # of packets tx to network */
+		data[i++] = CVM_CAST64(oct_dev->instr_queue[j]->stats.tx_done);
+		 /* # of bytes tx to network */
+		data[i++] = CVM_CAST64(
+				oct_dev->instr_queue[j]->stats.tx_tot_bytes);
+		/* # of packets dropped */
+		data[i++] = CVM_CAST64(
+				oct_dev->instr_queue[j]->stats.tx_dropped);
+		/* # of tx fails due to queue full */
+		data[i++] = CVM_CAST64(
+				oct_dev->instr_queue[j]->stats.tx_iq_busy);
+		/* XXX gather entries sent */
+		data[i++] = CVM_CAST64(
+				oct_dev->instr_queue[j]->stats.sgentry_sent);
+
+		/* instruction to firmware: data and control */
+		/* # of instructions to the queue */
+		data[i++] = CVM_CAST64(
+				oct_dev->instr_queue[j]->stats.instr_posted);
+		/* # of instructions processed */
+		data[i++] =
+		    CVM_CAST64(oct_dev->instr_queue[j]->stats.instr_processed);
+		/* # of instructions could not be processed */
+		data[i++] =
+		    CVM_CAST64(oct_dev->instr_queue[j]->stats.instr_dropped);
+		/* bytes sent through the queue */
+		data[i++] = CVM_CAST64(
+				oct_dev->instr_queue[j]->stats.bytes_sent);
+		/* tso request */
+		data[i++] = CVM_CAST64(oct_dev->instr_queue[j]->stats.tx_gso);
+		/* vxlan request */
+		data[i++] = CVM_CAST64(oct_dev->instr_queue[j]->stats.tx_vxlan);
+		/* txq restart */
+		data[i++] = CVM_CAST64(
+				oct_dev->instr_queue[j]->stats.tx_restart);
+	}
+
+	/* RX */
+	for (vj = 0; vj < oct_dev->num_oqs; vj++) {
+		j = lio->linfo.rxpciq[vj].s.q_no;
+
+		/* packets send to TCP/IP network stack */
+		/* # of packets to network stack */
+		data[i++] = CVM_CAST64(
+				oct_dev->droq[j]->stats.rx_pkts_received);
+		/* # of bytes to network stack */
+		data[i++] = CVM_CAST64(
+				oct_dev->droq[j]->stats.rx_bytes_received);
+		data[i++] = CVM_CAST64(oct_dev->droq[j]->stats.dropped_nomem +
+				       oct_dev->droq[j]->stats.dropped_toomany +
+				       oct_dev->droq[j]->stats.rx_dropped);
+		data[i++] = CVM_CAST64(oct_dev->droq[j]->stats.dropped_nomem);
+		data[i++] = CVM_CAST64(oct_dev->droq[j]->stats.dropped_toomany);
+		data[i++] = CVM_CAST64(oct_dev->droq[j]->stats.rx_dropped);
+
+		/* control and data path */
+		data[i++] = CVM_CAST64(oct_dev->droq[j]->stats.pkts_received);
+		data[i++] = CVM_CAST64(oct_dev->droq[j]->stats.bytes_received);
+		data[i++] =
+			CVM_CAST64(oct_dev->droq[j]->stats.dropped_nodispatch);
+
+		data[i++] = CVM_CAST64(oct_dev->droq[j]->stats.rx_vxlan);
+		data[i++] =
+		    CVM_CAST64(oct_dev->droq[j]->stats.rx_alloc_failure);
+	}
+}
+
+static void lio_get_priv_flags_strings(struct lio *lio, u8 *data)
+{
+	struct octeon_device *oct_dev = lio->oct_dev;
+	int i;
+
+	switch (oct_dev->chip_id) {
+	case OCTEON_CN23XX_PF_VID:
+	case OCTEON_CN23XX_VF_VID:
+		for (i = 0; i < ARRAY_SIZE(oct_priv_flags_strings); i++) {
+			sprintf(data, "%s", oct_priv_flags_strings[i]);
+			data += ETH_GSTRING_LEN;
+		}
+		break;
+	case OCTEON_CN68XX:
+	case OCTEON_CN66XX:
+		break;
+	default:
+		netif_info(lio, drv, lio->netdev, "Unknown Chip !!\n");
+		break;
+	}
+}
+
+static void lio_get_strings(struct net_device *netdev, u32 stringset, u8 *data)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct_dev = lio->oct_dev;
+	int num_iq_stats, num_oq_stats, i, j;
+	int num_stats;
+
+	switch (stringset) {
+	case ETH_SS_STATS:
+		num_stats = ARRAY_SIZE(oct_stats_strings);
+		for (j = 0; j < num_stats; j++) {
+			sprintf(data, "%s", oct_stats_strings[j]);
+			data += ETH_GSTRING_LEN;
+		}
+
+		num_iq_stats = ARRAY_SIZE(oct_iq_stats_strings);
+		for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct_dev); i++) {
+			if (!(oct_dev->io_qmask.iq & BIT_ULL(i)))
+				continue;
+			for (j = 0; j < num_iq_stats; j++) {
+				sprintf(data, "tx-%d-%s", i,
+					oct_iq_stats_strings[j]);
+				data += ETH_GSTRING_LEN;
+			}
+		}
+
+		num_oq_stats = ARRAY_SIZE(oct_droq_stats_strings);
+		for (i = 0; i < MAX_OCTEON_OUTPUT_QUEUES(oct_dev); i++) {
+			if (!(oct_dev->io_qmask.oq & BIT_ULL(i)))
+				continue;
+			for (j = 0; j < num_oq_stats; j++) {
+				sprintf(data, "rx-%d-%s", i,
+					oct_droq_stats_strings[j]);
+				data += ETH_GSTRING_LEN;
+			}
+		}
+		break;
+
+	case ETH_SS_PRIV_FLAGS:
+		lio_get_priv_flags_strings(lio, data);
+		break;
+	default:
+		netif_info(lio, drv, lio->netdev, "Unknown Stringset !!\n");
+		break;
+	}
+}
+
+static void lio_vf_get_strings(struct net_device *netdev, u32 stringset,
+			       u8 *data)
+{
+	int num_iq_stats, num_oq_stats, i, j;
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct_dev = lio->oct_dev;
+	int num_stats;
+
+	switch (stringset) {
+	case ETH_SS_STATS:
+		num_stats = ARRAY_SIZE(oct_vf_stats_strings);
+		for (j = 0; j < num_stats; j++) {
+			sprintf(data, "%s", oct_vf_stats_strings[j]);
+			data += ETH_GSTRING_LEN;
+		}
+
+		num_iq_stats = ARRAY_SIZE(oct_iq_stats_strings);
+		for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct_dev); i++) {
+			if (!(oct_dev->io_qmask.iq & BIT_ULL(i)))
+				continue;
+			for (j = 0; j < num_iq_stats; j++) {
+				sprintf(data, "tx-%d-%s", i,
+					oct_iq_stats_strings[j]);
+				data += ETH_GSTRING_LEN;
+			}
+		}
+
+		num_oq_stats = ARRAY_SIZE(oct_droq_stats_strings);
+		for (i = 0; i < MAX_OCTEON_OUTPUT_QUEUES(oct_dev); i++) {
+			if (!(oct_dev->io_qmask.oq & BIT_ULL(i)))
+				continue;
+			for (j = 0; j < num_oq_stats; j++) {
+				sprintf(data, "rx-%d-%s", i,
+					oct_droq_stats_strings[j]);
+				data += ETH_GSTRING_LEN;
+			}
+		}
+		break;
+
+	case ETH_SS_PRIV_FLAGS:
+		lio_get_priv_flags_strings(lio, data);
+		break;
+	default:
+		netif_info(lio, drv, lio->netdev, "Unknown Stringset !!\n");
+		break;
+	}
+}
+
+static int lio_get_priv_flags_ss_count(struct lio *lio)
+{
+	struct octeon_device *oct_dev = lio->oct_dev;
+
+	switch (oct_dev->chip_id) {
+	case OCTEON_CN23XX_PF_VID:
+	case OCTEON_CN23XX_VF_VID:
+		return ARRAY_SIZE(oct_priv_flags_strings);
+	case OCTEON_CN68XX:
+	case OCTEON_CN66XX:
+		return -EOPNOTSUPP;
+	default:
+		netif_info(lio, drv, lio->netdev, "Unknown Chip !!\n");
+		return -EOPNOTSUPP;
+	}
+}
+
+static int lio_get_sset_count(struct net_device *netdev, int sset)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct_dev = lio->oct_dev;
+
+	switch (sset) {
+	case ETH_SS_STATS:
+		return (ARRAY_SIZE(oct_stats_strings) +
+			ARRAY_SIZE(oct_iq_stats_strings) * oct_dev->num_iqs +
+			ARRAY_SIZE(oct_droq_stats_strings) * oct_dev->num_oqs);
+	case ETH_SS_PRIV_FLAGS:
+		return lio_get_priv_flags_ss_count(lio);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static int lio_vf_get_sset_count(struct net_device *netdev, int sset)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct_dev = lio->oct_dev;
+
+	switch (sset) {
+	case ETH_SS_STATS:
+		return (ARRAY_SIZE(oct_vf_stats_strings) +
+			ARRAY_SIZE(oct_iq_stats_strings) * oct_dev->num_iqs +
+			ARRAY_SIZE(oct_droq_stats_strings) * oct_dev->num_oqs);
+	case ETH_SS_PRIV_FLAGS:
+		return lio_get_priv_flags_ss_count(lio);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+/*  get interrupt moderation parameters */
+static int octnet_get_intrmod_cfg(struct lio *lio,
+				  struct oct_intrmod_cfg *intr_cfg)
+{
+	struct octeon_soft_command *sc;
+	struct oct_intrmod_resp *resp;
+	int retval;
+	struct octeon_device *oct_dev = lio->oct_dev;
+
+	/* Alloc soft command */
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct_dev,
+					  0,
+					  sizeof(struct oct_intrmod_resp), 0);
+
+	if (!sc)
+		return -ENOMEM;
+
+	resp = (struct oct_intrmod_resp *)sc->virtrptr;
+	memset(resp, 0, sizeof(struct oct_intrmod_resp));
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct_dev, sc, OPCODE_NIC,
+				    OPCODE_NIC_INTRMOD_PARAMS, 0, 0, 0);
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	retval = octeon_send_soft_command(oct_dev, sc);
+	if (retval == IQ_SEND_FAILED) {
+		octeon_free_soft_command(oct_dev, sc);
+		return -EINVAL;
+	}
+
+	/* Sleep on a wait queue till the cond flag indicates that the
+	 * response arrived or timed-out.
+	 */
+	retval = wait_for_sc_completion_timeout(oct_dev, sc, 0);
+	if (retval)
+		return -ENODEV;
+
+	if (resp->status) {
+		dev_err(&oct_dev->pci_dev->dev,
+			"Get interrupt moderation parameters failed\n");
+		WRITE_ONCE(sc->caller_is_done, true);
+		return -ENODEV;
+	}
+
+	octeon_swap_8B_data((u64 *)&resp->intrmod,
+			    (sizeof(struct oct_intrmod_cfg)) / 8);
+	memcpy(intr_cfg, &resp->intrmod, sizeof(struct oct_intrmod_cfg));
+	WRITE_ONCE(sc->caller_is_done, true);
+
+	return 0;
+}
+
+/*  Configure interrupt moderation parameters */
+static int octnet_set_intrmod_cfg(struct lio *lio,
+				  struct oct_intrmod_cfg *intr_cfg)
+{
+	struct octeon_soft_command *sc;
+	struct oct_intrmod_cfg *cfg;
+	int retval;
+	struct octeon_device *oct_dev = lio->oct_dev;
+
+	/* Alloc soft command */
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct_dev,
+					  sizeof(struct oct_intrmod_cfg),
+					  16, 0);
+
+	if (!sc)
+		return -ENOMEM;
+
+	cfg = (struct oct_intrmod_cfg *)sc->virtdptr;
+
+	memcpy(cfg, intr_cfg, sizeof(struct oct_intrmod_cfg));
+	octeon_swap_8B_data((u64 *)cfg, (sizeof(struct oct_intrmod_cfg)) / 8);
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct_dev, sc, OPCODE_NIC,
+				    OPCODE_NIC_INTRMOD_CFG, 0, 0, 0);
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	retval = octeon_send_soft_command(oct_dev, sc);
+	if (retval == IQ_SEND_FAILED) {
+		octeon_free_soft_command(oct_dev, sc);
+		return -EINVAL;
+	}
+
+	/* Sleep on a wait queue till the cond flag indicates that the
+	 * response arrived or timed-out.
+	 */
+	retval = wait_for_sc_completion_timeout(oct_dev, sc, 0);
+	if (retval)
+		return retval;
+
+	retval = sc->sc_status;
+	if (retval == 0) {
+		dev_info(&oct_dev->pci_dev->dev,
+			 "Rx-Adaptive Interrupt moderation %s\n",
+			 (intr_cfg->rx_enable) ?
+			 "enabled" : "disabled");
+		WRITE_ONCE(sc->caller_is_done, true);
+		return 0;
+	}
+
+	dev_err(&oct_dev->pci_dev->dev,
+		"intrmod config failed. Status: %x\n", retval);
+	WRITE_ONCE(sc->caller_is_done, true);
+	return -ENODEV;
+}
+
+static int lio_get_intr_coalesce(struct net_device *netdev,
+				 struct ethtool_coalesce *intr_coal,
+				 struct kernel_ethtool_coalesce *kernel_coal,
+				 struct netlink_ext_ack *extack)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octeon_instr_queue *iq;
+	struct oct_intrmod_cfg intrmod_cfg;
+
+	if (octnet_get_intrmod_cfg(lio, &intrmod_cfg))
+		return -ENODEV;
+
+	switch (oct->chip_id) {
+	case OCTEON_CN23XX_PF_VID:
+	case OCTEON_CN23XX_VF_VID: {
+		if (!intrmod_cfg.rx_enable) {
+			intr_coal->rx_coalesce_usecs = oct->rx_coalesce_usecs;
+			intr_coal->rx_max_coalesced_frames =
+				oct->rx_max_coalesced_frames;
+		}
+		if (!intrmod_cfg.tx_enable)
+			intr_coal->tx_max_coalesced_frames =
+				oct->tx_max_coalesced_frames;
+		break;
+	}
+	case OCTEON_CN68XX:
+	case OCTEON_CN66XX: {
+		struct octeon_cn6xxx *cn6xxx =
+			(struct octeon_cn6xxx *)oct->chip;
+
+		if (!intrmod_cfg.rx_enable) {
+			intr_coal->rx_coalesce_usecs =
+				CFG_GET_OQ_INTR_TIME(cn6xxx->conf);
+			intr_coal->rx_max_coalesced_frames =
+				CFG_GET_OQ_INTR_PKT(cn6xxx->conf);
+		}
+		iq = oct->instr_queue[lio->linfo.txpciq[0].s.q_no];
+		intr_coal->tx_max_coalesced_frames = iq->fill_threshold;
+		break;
+	}
+	default:
+		netif_info(lio, drv, lio->netdev, "Unknown Chip !!\n");
+		return -EINVAL;
+	}
+	if (intrmod_cfg.rx_enable) {
+		intr_coal->use_adaptive_rx_coalesce =
+			intrmod_cfg.rx_enable;
+		intr_coal->rate_sample_interval =
+			intrmod_cfg.check_intrvl;
+		intr_coal->pkt_rate_high =
+			intrmod_cfg.maxpkt_ratethr;
+		intr_coal->pkt_rate_low =
+			intrmod_cfg.minpkt_ratethr;
+		intr_coal->rx_max_coalesced_frames_high =
+			intrmod_cfg.rx_maxcnt_trigger;
+		intr_coal->rx_coalesce_usecs_high =
+			intrmod_cfg.rx_maxtmr_trigger;
+		intr_coal->rx_coalesce_usecs_low =
+			intrmod_cfg.rx_mintmr_trigger;
+		intr_coal->rx_max_coalesced_frames_low =
+			intrmod_cfg.rx_mincnt_trigger;
+	}
+	if ((OCTEON_CN23XX_PF(oct) || OCTEON_CN23XX_VF(oct)) &&
+	    (intrmod_cfg.tx_enable)) {
+		intr_coal->use_adaptive_tx_coalesce =
+			intrmod_cfg.tx_enable;
+		intr_coal->tx_max_coalesced_frames_high =
+			intrmod_cfg.tx_maxcnt_trigger;
+		intr_coal->tx_max_coalesced_frames_low =
+			intrmod_cfg.tx_mincnt_trigger;
+	}
+	return 0;
+}
+
+/* Enable/Disable auto interrupt Moderation */
+static int oct_cfg_adaptive_intr(struct lio *lio,
+				 struct oct_intrmod_cfg *intrmod_cfg,
+				 struct ethtool_coalesce *intr_coal)
+{
+	int ret = 0;
+
+	if (intrmod_cfg->rx_enable || intrmod_cfg->tx_enable) {
+		intrmod_cfg->check_intrvl = intr_coal->rate_sample_interval;
+		intrmod_cfg->maxpkt_ratethr = intr_coal->pkt_rate_high;
+		intrmod_cfg->minpkt_ratethr = intr_coal->pkt_rate_low;
+	}
+	if (intrmod_cfg->rx_enable) {
+		intrmod_cfg->rx_maxcnt_trigger =
+			intr_coal->rx_max_coalesced_frames_high;
+		intrmod_cfg->rx_maxtmr_trigger =
+			intr_coal->rx_coalesce_usecs_high;
+		intrmod_cfg->rx_mintmr_trigger =
+			intr_coal->rx_coalesce_usecs_low;
+		intrmod_cfg->rx_mincnt_trigger =
+			intr_coal->rx_max_coalesced_frames_low;
+	}
+	if (intrmod_cfg->tx_enable) {
+		intrmod_cfg->tx_maxcnt_trigger =
+			intr_coal->tx_max_coalesced_frames_high;
+		intrmod_cfg->tx_mincnt_trigger =
+			intr_coal->tx_max_coalesced_frames_low;
+	}
+
+	ret = octnet_set_intrmod_cfg(lio, intrmod_cfg);
+
+	return ret;
+}
+
+static int
+oct_cfg_rx_intrcnt(struct lio *lio,
+		   struct oct_intrmod_cfg *intrmod,
+		   struct ethtool_coalesce *intr_coal)
+{
+	struct octeon_device *oct = lio->oct_dev;
+	u32 rx_max_coalesced_frames;
+
+	/* Config Cnt based interrupt values */
+	switch (oct->chip_id) {
+	case OCTEON_CN68XX:
+	case OCTEON_CN66XX: {
+		struct octeon_cn6xxx *cn6xxx =
+			(struct octeon_cn6xxx *)oct->chip;
+
+		if (!intr_coal->rx_max_coalesced_frames)
+			rx_max_coalesced_frames = CN6XXX_OQ_INTR_PKT;
+		else
+			rx_max_coalesced_frames =
+				intr_coal->rx_max_coalesced_frames;
+		octeon_write_csr(oct, CN6XXX_SLI_OQ_INT_LEVEL_PKTS,
+				 rx_max_coalesced_frames);
+		CFG_SET_OQ_INTR_PKT(cn6xxx->conf, rx_max_coalesced_frames);
+		break;
+	}
+	case OCTEON_CN23XX_PF_VID: {
+		int q_no;
+
+		if (!intr_coal->rx_max_coalesced_frames)
+			rx_max_coalesced_frames = intrmod->rx_frames;
+		else
+			rx_max_coalesced_frames =
+			    intr_coal->rx_max_coalesced_frames;
+		for (q_no = 0; q_no < oct->num_oqs; q_no++) {
+			q_no += oct->sriov_info.pf_srn;
+			octeon_write_csr64(
+			    oct, CN23XX_SLI_OQ_PKT_INT_LEVELS(q_no),
+			    (octeon_read_csr64(
+				 oct, CN23XX_SLI_OQ_PKT_INT_LEVELS(q_no)) &
+			     (0x3fffff00000000UL)) |
+				(rx_max_coalesced_frames - 1));
+			/*consider setting resend bit*/
+		}
+		intrmod->rx_frames = rx_max_coalesced_frames;
+		oct->rx_max_coalesced_frames = rx_max_coalesced_frames;
+		break;
+	}
+	case OCTEON_CN23XX_VF_VID: {
+		int q_no;
+
+		if (!intr_coal->rx_max_coalesced_frames)
+			rx_max_coalesced_frames = intrmod->rx_frames;
+		else
+			rx_max_coalesced_frames =
+			    intr_coal->rx_max_coalesced_frames;
+		for (q_no = 0; q_no < oct->num_oqs; q_no++) {
+			octeon_write_csr64(
+			    oct, CN23XX_VF_SLI_OQ_PKT_INT_LEVELS(q_no),
+			    (octeon_read_csr64(
+				 oct, CN23XX_VF_SLI_OQ_PKT_INT_LEVELS(q_no)) &
+			     (0x3fffff00000000UL)) |
+				(rx_max_coalesced_frames - 1));
+			/*consider writing to resend bit here*/
+		}
+		intrmod->rx_frames = rx_max_coalesced_frames;
+		oct->rx_max_coalesced_frames = rx_max_coalesced_frames;
+		break;
+	}
+	default:
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static int oct_cfg_rx_intrtime(struct lio *lio,
+			       struct oct_intrmod_cfg *intrmod,
+			       struct ethtool_coalesce *intr_coal)
+{
+	struct octeon_device *oct = lio->oct_dev;
+	u32 time_threshold, rx_coalesce_usecs;
+
+	/* Config Time based interrupt values */
+	switch (oct->chip_id) {
+	case OCTEON_CN68XX:
+	case OCTEON_CN66XX: {
+		struct octeon_cn6xxx *cn6xxx =
+			(struct octeon_cn6xxx *)oct->chip;
+		if (!intr_coal->rx_coalesce_usecs)
+			rx_coalesce_usecs = CN6XXX_OQ_INTR_TIME;
+		else
+			rx_coalesce_usecs = intr_coal->rx_coalesce_usecs;
+
+		time_threshold = lio_cn6xxx_get_oq_ticks(oct,
+							 rx_coalesce_usecs);
+		octeon_write_csr(oct,
+				 CN6XXX_SLI_OQ_INT_LEVEL_TIME,
+				 time_threshold);
+
+		CFG_SET_OQ_INTR_TIME(cn6xxx->conf, rx_coalesce_usecs);
+		break;
+	}
+	case OCTEON_CN23XX_PF_VID: {
+		u64 time_threshold;
+		int q_no;
+
+		if (!intr_coal->rx_coalesce_usecs)
+			rx_coalesce_usecs = intrmod->rx_usecs;
+		else
+			rx_coalesce_usecs = intr_coal->rx_coalesce_usecs;
+		time_threshold =
+		    cn23xx_pf_get_oq_ticks(oct, (u32)rx_coalesce_usecs);
+		for (q_no = 0; q_no < oct->num_oqs; q_no++) {
+			q_no += oct->sriov_info.pf_srn;
+			octeon_write_csr64(oct,
+					   CN23XX_SLI_OQ_PKT_INT_LEVELS(q_no),
+					   (intrmod->rx_frames |
+					    ((u64)time_threshold << 32)));
+			/*consider writing to resend bit here*/
+		}
+		intrmod->rx_usecs = rx_coalesce_usecs;
+		oct->rx_coalesce_usecs = rx_coalesce_usecs;
+		break;
+	}
+	case OCTEON_CN23XX_VF_VID: {
+		u64 time_threshold;
+		int q_no;
+
+		if (!intr_coal->rx_coalesce_usecs)
+			rx_coalesce_usecs = intrmod->rx_usecs;
+		else
+			rx_coalesce_usecs = intr_coal->rx_coalesce_usecs;
+
+		time_threshold =
+		    cn23xx_vf_get_oq_ticks(oct, (u32)rx_coalesce_usecs);
+		for (q_no = 0; q_no < oct->num_oqs; q_no++) {
+			octeon_write_csr64(
+				oct, CN23XX_VF_SLI_OQ_PKT_INT_LEVELS(q_no),
+				(intrmod->rx_frames |
+				 ((u64)time_threshold << 32)));
+			/*consider setting resend bit*/
+		}
+		intrmod->rx_usecs = rx_coalesce_usecs;
+		oct->rx_coalesce_usecs = rx_coalesce_usecs;
+		break;
+	}
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int
+oct_cfg_tx_intrcnt(struct lio *lio,
+		   struct oct_intrmod_cfg *intrmod,
+		   struct ethtool_coalesce *intr_coal)
+{
+	struct octeon_device *oct = lio->oct_dev;
+	u32 iq_intr_pkt;
+	void __iomem *inst_cnt_reg;
+	u64 val;
+
+	/* Config Cnt based interrupt values */
+	switch (oct->chip_id) {
+	case OCTEON_CN68XX:
+	case OCTEON_CN66XX:
+		break;
+	case OCTEON_CN23XX_VF_VID:
+	case OCTEON_CN23XX_PF_VID: {
+		int q_no;
+
+		if (!intr_coal->tx_max_coalesced_frames)
+			iq_intr_pkt = CN23XX_DEF_IQ_INTR_THRESHOLD &
+				      CN23XX_PKT_IN_DONE_WMARK_MASK;
+		else
+			iq_intr_pkt = intr_coal->tx_max_coalesced_frames &
+				      CN23XX_PKT_IN_DONE_WMARK_MASK;
+		for (q_no = 0; q_no < oct->num_iqs; q_no++) {
+			inst_cnt_reg = (oct->instr_queue[q_no])->inst_cnt_reg;
+			val = readq(inst_cnt_reg);
+			/*clear wmark and count.dont want to write count back*/
+			val = (val & 0xFFFF000000000000ULL) |
+			      ((u64)(iq_intr_pkt - 1)
+			       << CN23XX_PKT_IN_DONE_WMARK_BIT_POS);
+			writeq(val, inst_cnt_reg);
+			/*consider setting resend bit*/
+		}
+		intrmod->tx_frames = iq_intr_pkt;
+		oct->tx_max_coalesced_frames = iq_intr_pkt;
+		break;
+	}
+	default:
+		return -EINVAL;
+	}
+	return 0;
+}
+
+static int lio_set_intr_coalesce(struct net_device *netdev,
+				 struct ethtool_coalesce *intr_coal,
+				 struct kernel_ethtool_coalesce *kernel_coal,
+				 struct netlink_ext_ack *extack)
+{
+	struct lio *lio = GET_LIO(netdev);
+	int ret;
+	struct octeon_device *oct = lio->oct_dev;
+	struct oct_intrmod_cfg intrmod = {0};
+	u32 j, q_no;
+	int db_max, db_min;
+
+	switch (oct->chip_id) {
+	case OCTEON_CN68XX:
+	case OCTEON_CN66XX:
+		db_min = CN6XXX_DB_MIN;
+		db_max = CN6XXX_DB_MAX;
+		if ((intr_coal->tx_max_coalesced_frames >= db_min) &&
+		    (intr_coal->tx_max_coalesced_frames <= db_max)) {
+			for (j = 0; j < lio->linfo.num_txpciq; j++) {
+				q_no = lio->linfo.txpciq[j].s.q_no;
+				oct->instr_queue[q_no]->fill_threshold =
+					intr_coal->tx_max_coalesced_frames;
+			}
+		} else {
+			dev_err(&oct->pci_dev->dev,
+				"LIQUIDIO: Invalid tx-frames:%d. Range is min:%d max:%d\n",
+				intr_coal->tx_max_coalesced_frames,
+				db_min, db_max);
+			return -EINVAL;
+		}
+		break;
+	case OCTEON_CN23XX_PF_VID:
+	case OCTEON_CN23XX_VF_VID:
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	intrmod.rx_enable = intr_coal->use_adaptive_rx_coalesce ? 1 : 0;
+	intrmod.tx_enable = intr_coal->use_adaptive_tx_coalesce ? 1 : 0;
+	intrmod.rx_frames = CFG_GET_OQ_INTR_PKT(octeon_get_conf(oct));
+	intrmod.rx_usecs = CFG_GET_OQ_INTR_TIME(octeon_get_conf(oct));
+	intrmod.tx_frames = CFG_GET_IQ_INTR_PKT(octeon_get_conf(oct));
+
+	ret = oct_cfg_adaptive_intr(lio, &intrmod, intr_coal);
+
+	if (!intr_coal->use_adaptive_rx_coalesce) {
+		ret = oct_cfg_rx_intrtime(lio, &intrmod, intr_coal);
+		if (ret)
+			goto ret_intrmod;
+
+		ret = oct_cfg_rx_intrcnt(lio, &intrmod, intr_coal);
+		if (ret)
+			goto ret_intrmod;
+	} else {
+		oct->rx_coalesce_usecs =
+			CFG_GET_OQ_INTR_TIME(octeon_get_conf(oct));
+		oct->rx_max_coalesced_frames =
+			CFG_GET_OQ_INTR_PKT(octeon_get_conf(oct));
+	}
+
+	if (!intr_coal->use_adaptive_tx_coalesce) {
+		ret = oct_cfg_tx_intrcnt(lio, &intrmod, intr_coal);
+		if (ret)
+			goto ret_intrmod;
+	} else {
+		oct->tx_max_coalesced_frames =
+			CFG_GET_IQ_INTR_PKT(octeon_get_conf(oct));
+	}
+
+	return 0;
+ret_intrmod:
+	return ret;
+}
+
+static int lio_get_ts_info(struct net_device *netdev,
+			   struct ethtool_ts_info *info)
+{
+	struct lio *lio = GET_LIO(netdev);
+
+	info->so_timestamping =
+#ifdef PTP_HARDWARE_TIMESTAMPING
+		SOF_TIMESTAMPING_TX_HARDWARE |
+		SOF_TIMESTAMPING_RX_HARDWARE |
+		SOF_TIMESTAMPING_RAW_HARDWARE |
+		SOF_TIMESTAMPING_TX_SOFTWARE |
+#endif
+		SOF_TIMESTAMPING_RX_SOFTWARE |
+		SOF_TIMESTAMPING_SOFTWARE;
+
+	if (lio->ptp_clock)
+		info->phc_index = ptp_clock_index(lio->ptp_clock);
+	else
+		info->phc_index = -1;
+
+#ifdef PTP_HARDWARE_TIMESTAMPING
+	info->tx_types = (1 << HWTSTAMP_TX_OFF) | (1 << HWTSTAMP_TX_ON);
+
+	info->rx_filters = (1 << HWTSTAMP_FILTER_NONE) |
+			   (1 << HWTSTAMP_FILTER_PTP_V1_L4_EVENT) |
+			   (1 << HWTSTAMP_FILTER_PTP_V2_L2_EVENT) |
+			   (1 << HWTSTAMP_FILTER_PTP_V2_L4_EVENT);
+#endif
+
+	return 0;
+}
+
+/* Return register dump len. */
+static int lio_get_regs_len(struct net_device *dev)
+{
+	struct lio *lio = GET_LIO(dev);
+	struct octeon_device *oct = lio->oct_dev;
+
+	switch (oct->chip_id) {
+	case OCTEON_CN23XX_PF_VID:
+		return OCT_ETHTOOL_REGDUMP_LEN_23XX;
+	case OCTEON_CN23XX_VF_VID:
+		return OCT_ETHTOOL_REGDUMP_LEN_23XX_VF;
+	default:
+		return OCT_ETHTOOL_REGDUMP_LEN;
+	}
+}
+
+static int cn23xx_read_csr_reg(char *s, struct octeon_device *oct)
+{
+	u32 reg;
+	u8 pf_num = oct->pf_num;
+	int len = 0;
+	int i;
+
+	/* PCI  Window Registers */
+
+	len += sprintf(s + len, "\n\t Octeon CSR Registers\n\n");
+
+	/*0x29030 or 0x29040*/
+	reg = CN23XX_SLI_PKT_MAC_RINFO64(oct->pcie_port, oct->pf_num);
+	len += sprintf(s + len,
+		       "\n[%08x] (SLI_PKT_MAC%d_PF%d_RINFO): %016llx\n",
+		       reg, oct->pcie_port, oct->pf_num,
+		       (u64)octeon_read_csr64(oct, reg));
+
+	/*0x27080 or 0x27090*/
+	reg = CN23XX_SLI_MAC_PF_INT_ENB64(oct->pcie_port, oct->pf_num);
+	len +=
+	    sprintf(s + len, "\n[%08x] (SLI_MAC%d_PF%d_INT_ENB): %016llx\n",
+		    reg, oct->pcie_port, oct->pf_num,
+		    (u64)octeon_read_csr64(oct, reg));
+
+	/*0x27000 or 0x27010*/
+	reg = CN23XX_SLI_MAC_PF_INT_SUM64(oct->pcie_port, oct->pf_num);
+	len +=
+	    sprintf(s + len, "\n[%08x] (SLI_MAC%d_PF%d_INT_SUM): %016llx\n",
+		    reg, oct->pcie_port, oct->pf_num,
+		    (u64)octeon_read_csr64(oct, reg));
+
+	/*0x29120*/
+	reg = 0x29120;
+	len += sprintf(s + len, "\n[%08x] (SLI_PKT_MEM_CTL): %016llx\n", reg,
+		       (u64)octeon_read_csr64(oct, reg));
+
+	/*0x27300*/
+	reg = 0x27300 + oct->pcie_port * CN23XX_MAC_INT_OFFSET +
+	      (oct->pf_num) * CN23XX_PF_INT_OFFSET;
+	len += sprintf(
+	    s + len, "\n[%08x] (SLI_MAC%d_PF%d_PKT_VF_INT): %016llx\n", reg,
+	    oct->pcie_port, oct->pf_num, (u64)octeon_read_csr64(oct, reg));
+
+	/*0x27200*/
+	reg = 0x27200 + oct->pcie_port * CN23XX_MAC_INT_OFFSET +
+	      (oct->pf_num) * CN23XX_PF_INT_OFFSET;
+	len += sprintf(s + len,
+		       "\n[%08x] (SLI_MAC%d_PF%d_PP_VF_INT): %016llx\n",
+		       reg, oct->pcie_port, oct->pf_num,
+		       (u64)octeon_read_csr64(oct, reg));
+
+	/*29130*/
+	reg = CN23XX_SLI_PKT_CNT_INT;
+	len += sprintf(s + len, "\n[%08x] (SLI_PKT_CNT_INT): %016llx\n", reg,
+		       (u64)octeon_read_csr64(oct, reg));
+
+	/*0x29140*/
+	reg = CN23XX_SLI_PKT_TIME_INT;
+	len += sprintf(s + len, "\n[%08x] (SLI_PKT_TIME_INT): %016llx\n", reg,
+		       (u64)octeon_read_csr64(oct, reg));
+
+	/*0x29160*/
+	reg = 0x29160;
+	len += sprintf(s + len, "\n[%08x] (SLI_PKT_INT): %016llx\n", reg,
+		       (u64)octeon_read_csr64(oct, reg));
+
+	/*0x29180*/
+	reg = CN23XX_SLI_OQ_WMARK;
+	len += sprintf(s + len, "\n[%08x] (SLI_PKT_OUTPUT_WMARK): %016llx\n",
+		       reg, (u64)octeon_read_csr64(oct, reg));
+
+	/*0x291E0*/
+	reg = CN23XX_SLI_PKT_IOQ_RING_RST;
+	len += sprintf(s + len, "\n[%08x] (SLI_PKT_RING_RST): %016llx\n", reg,
+		       (u64)octeon_read_csr64(oct, reg));
+
+	/*0x29210*/
+	reg = CN23XX_SLI_GBL_CONTROL;
+	len += sprintf(s + len,
+		       "\n[%08x] (SLI_PKT_GBL_CONTROL): %016llx\n", reg,
+		       (u64)octeon_read_csr64(oct, reg));
+
+	/*0x29220*/
+	reg = 0x29220;
+	len += sprintf(s + len, "\n[%08x] (SLI_PKT_BIST_STATUS): %016llx\n",
+		       reg, (u64)octeon_read_csr64(oct, reg));
+
+	/*PF only*/
+	if (pf_num == 0) {
+		/*0x29260*/
+		reg = CN23XX_SLI_OUT_BP_EN_W1S;
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT_OUT_BP_EN_W1S):  %016llx\n",
+			       reg, (u64)octeon_read_csr64(oct, reg));
+	} else if (pf_num == 1) {
+		/*0x29270*/
+		reg = CN23XX_SLI_OUT_BP_EN2_W1S;
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT_OUT_BP_EN2_W1S): %016llx\n",
+			       reg, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < CN23XX_MAX_OUTPUT_QUEUES; i++) {
+		reg = CN23XX_SLI_OQ_BUFF_INFO_SIZE(i);
+		len +=
+		    sprintf(s + len, "\n[%08x] (SLI_PKT%d_OUT_SIZE): %016llx\n",
+			    reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	/*0x10040*/
+	for (i = 0; i < CN23XX_MAX_INPUT_QUEUES; i++) {
+		reg = CN23XX_SLI_IQ_INSTR_COUNT64(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT_IN_DONE%d_CNTS): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	/*0x10080*/
+	for (i = 0; i < CN23XX_MAX_OUTPUT_QUEUES; i++) {
+		reg = CN23XX_SLI_OQ_PKTS_CREDIT(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_SLIST_BAOFF_DBELL): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	/*0x10090*/
+	for (i = 0; i < CN23XX_MAX_OUTPUT_QUEUES; i++) {
+		reg = CN23XX_SLI_OQ_SIZE(i);
+		len += sprintf(
+		    s + len, "\n[%08x] (SLI_PKT%d_SLIST_FIFO_RSIZE): %016llx\n",
+		    reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	/*0x10050*/
+	for (i = 0; i < CN23XX_MAX_OUTPUT_QUEUES; i++) {
+		reg = CN23XX_SLI_OQ_PKT_CONTROL(i);
+		len += sprintf(
+			s + len,
+			"\n[%08x] (SLI_PKT%d__OUTPUT_CONTROL): %016llx\n",
+			reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	/*0x10070*/
+	for (i = 0; i < CN23XX_MAX_OUTPUT_QUEUES; i++) {
+		reg = CN23XX_SLI_OQ_BASE_ADDR64(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_SLIST_BADDR): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	/*0x100a0*/
+	for (i = 0; i < CN23XX_MAX_OUTPUT_QUEUES; i++) {
+		reg = CN23XX_SLI_OQ_PKT_INT_LEVELS(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_INT_LEVELS): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	/*0x100b0*/
+	for (i = 0; i < CN23XX_MAX_OUTPUT_QUEUES; i++) {
+		reg = CN23XX_SLI_OQ_PKTS_SENT(i);
+		len += sprintf(s + len, "\n[%08x] (SLI_PKT%d_CNTS): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	/*0x100c0*/
+	for (i = 0; i < CN23XX_MAX_OUTPUT_QUEUES; i++) {
+		reg = 0x100c0 + i * CN23XX_OQ_OFFSET;
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_ERROR_INFO): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+
+		/*0x10000*/
+		for (i = 0; i < CN23XX_MAX_INPUT_QUEUES; i++) {
+			reg = CN23XX_SLI_IQ_PKT_CONTROL64(i);
+			len += sprintf(
+				s + len,
+				"\n[%08x] (SLI_PKT%d_INPUT_CONTROL): %016llx\n",
+				reg, i, (u64)octeon_read_csr64(oct, reg));
+		}
+
+		/*0x10010*/
+		for (i = 0; i < CN23XX_MAX_INPUT_QUEUES; i++) {
+			reg = CN23XX_SLI_IQ_BASE_ADDR64(i);
+			len += sprintf(
+			    s + len,
+			    "\n[%08x] (SLI_PKT%d_INSTR_BADDR): %016llx\n", reg,
+			    i, (u64)octeon_read_csr64(oct, reg));
+		}
+
+		/*0x10020*/
+		for (i = 0; i < CN23XX_MAX_INPUT_QUEUES; i++) {
+			reg = CN23XX_SLI_IQ_DOORBELL(i);
+			len += sprintf(
+			    s + len,
+			    "\n[%08x] (SLI_PKT%d_INSTR_BAOFF_DBELL): %016llx\n",
+			    reg, i, (u64)octeon_read_csr64(oct, reg));
+		}
+
+		/*0x10030*/
+		for (i = 0; i < CN23XX_MAX_INPUT_QUEUES; i++) {
+			reg = CN23XX_SLI_IQ_SIZE(i);
+			len += sprintf(
+			    s + len,
+			    "\n[%08x] (SLI_PKT%d_INSTR_FIFO_RSIZE): %016llx\n",
+			    reg, i, (u64)octeon_read_csr64(oct, reg));
+		}
+
+		/*0x10040*/
+		for (i = 0; i < CN23XX_MAX_INPUT_QUEUES; i++)
+			reg = CN23XX_SLI_IQ_INSTR_COUNT64(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT_IN_DONE%d_CNTS): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	return len;
+}
+
+static int cn23xx_vf_read_csr_reg(char *s, struct octeon_device *oct)
+{
+	int len = 0;
+	u32 reg;
+	int i;
+
+	/* PCI  Window Registers */
+
+	len += sprintf(s + len, "\n\t Octeon CSR Registers\n\n");
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_OQ_BUFF_INFO_SIZE(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_OUT_SIZE): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_IQ_INSTR_COUNT64(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT_IN_DONE%d_CNTS): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_OQ_PKTS_CREDIT(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_SLIST_BAOFF_DBELL): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_OQ_SIZE(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_SLIST_FIFO_RSIZE): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_OQ_PKT_CONTROL(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d__OUTPUT_CONTROL): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_OQ_BASE_ADDR64(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_SLIST_BADDR): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_OQ_PKT_INT_LEVELS(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_INT_LEVELS): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_OQ_PKTS_SENT(i);
+		len += sprintf(s + len, "\n[%08x] (SLI_PKT%d_CNTS): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = 0x100c0 + i * CN23XX_VF_OQ_OFFSET;
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_ERROR_INFO): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = 0x100d0 + i * CN23XX_VF_IQ_OFFSET;
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_VF_INT_SUM): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_IQ_PKT_CONTROL64(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_INPUT_CONTROL): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_IQ_BASE_ADDR64(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_INSTR_BADDR): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_IQ_DOORBELL(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_INSTR_BAOFF_DBELL): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_IQ_SIZE(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT%d_INSTR_FIFO_RSIZE): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	for (i = 0; i < (oct->sriov_info.rings_per_vf); i++) {
+		reg = CN23XX_VF_SLI_IQ_INSTR_COUNT64(i);
+		len += sprintf(s + len,
+			       "\n[%08x] (SLI_PKT_IN_DONE%d_CNTS): %016llx\n",
+			       reg, i, (u64)octeon_read_csr64(oct, reg));
+	}
+
+	return len;
+}
+
+static int cn6xxx_read_csr_reg(char *s, struct octeon_device *oct)
+{
+	u32 reg;
+	int i, len = 0;
+
+	/* PCI  Window Registers */
+
+	len += sprintf(s + len, "\n\t Octeon CSR Registers\n\n");
+	reg = CN6XXX_WIN_WR_ADDR_LO;
+	len += sprintf(s + len, "\n[%02x] (WIN_WR_ADDR_LO): %08x\n",
+		       CN6XXX_WIN_WR_ADDR_LO, octeon_read_csr(oct, reg));
+	reg = CN6XXX_WIN_WR_ADDR_HI;
+	len += sprintf(s + len, "[%02x] (WIN_WR_ADDR_HI): %08x\n",
+		       CN6XXX_WIN_WR_ADDR_HI, octeon_read_csr(oct, reg));
+	reg = CN6XXX_WIN_RD_ADDR_LO;
+	len += sprintf(s + len, "[%02x] (WIN_RD_ADDR_LO): %08x\n",
+		       CN6XXX_WIN_RD_ADDR_LO, octeon_read_csr(oct, reg));
+	reg = CN6XXX_WIN_RD_ADDR_HI;
+	len += sprintf(s + len, "[%02x] (WIN_RD_ADDR_HI): %08x\n",
+		       CN6XXX_WIN_RD_ADDR_HI, octeon_read_csr(oct, reg));
+	reg = CN6XXX_WIN_WR_DATA_LO;
+	len += sprintf(s + len, "[%02x] (WIN_WR_DATA_LO): %08x\n",
+		       CN6XXX_WIN_WR_DATA_LO, octeon_read_csr(oct, reg));
+	reg = CN6XXX_WIN_WR_DATA_HI;
+	len += sprintf(s + len, "[%02x] (WIN_WR_DATA_HI): %08x\n",
+		       CN6XXX_WIN_WR_DATA_HI, octeon_read_csr(oct, reg));
+	len += sprintf(s + len, "[%02x] (WIN_WR_MASK_REG): %08x\n",
+		       CN6XXX_WIN_WR_MASK_REG,
+		       octeon_read_csr(oct, CN6XXX_WIN_WR_MASK_REG));
+
+	/* PCI  Interrupt Register */
+	len += sprintf(s + len, "\n[%x] (INT_ENABLE PORT 0): %08x\n",
+		       CN6XXX_SLI_INT_ENB64_PORT0, octeon_read_csr(oct,
+						CN6XXX_SLI_INT_ENB64_PORT0));
+	len += sprintf(s + len, "\n[%x] (INT_ENABLE PORT 1): %08x\n",
+		       CN6XXX_SLI_INT_ENB64_PORT1,
+		       octeon_read_csr(oct, CN6XXX_SLI_INT_ENB64_PORT1));
+	len += sprintf(s + len, "[%x] (INT_SUM): %08x\n", CN6XXX_SLI_INT_SUM64,
+		       octeon_read_csr(oct, CN6XXX_SLI_INT_SUM64));
+
+	/* PCI  Output queue registers */
+	for (i = 0; i < oct->num_oqs; i++) {
+		reg = CN6XXX_SLI_OQ_PKTS_SENT(i);
+		len += sprintf(s + len, "\n[%x] (PKTS_SENT_%d): %08x\n",
+			       reg, i, octeon_read_csr(oct, reg));
+		reg = CN6XXX_SLI_OQ_PKTS_CREDIT(i);
+		len += sprintf(s + len, "[%x] (PKT_CREDITS_%d): %08x\n",
+			       reg, i, octeon_read_csr(oct, reg));
+	}
+	reg = CN6XXX_SLI_OQ_INT_LEVEL_PKTS;
+	len += sprintf(s + len, "\n[%x] (PKTS_SENT_INT_LEVEL): %08x\n",
+		       reg, octeon_read_csr(oct, reg));
+	reg = CN6XXX_SLI_OQ_INT_LEVEL_TIME;
+	len += sprintf(s + len, "[%x] (PKTS_SENT_TIME): %08x\n",
+		       reg, octeon_read_csr(oct, reg));
+
+	/* PCI  Input queue registers */
+	for (i = 0; i <= 3; i++) {
+		u32 reg;
+
+		reg = CN6XXX_SLI_IQ_DOORBELL(i);
+		len += sprintf(s + len, "\n[%x] (INSTR_DOORBELL_%d): %08x\n",
+			       reg, i, octeon_read_csr(oct, reg));
+		reg = CN6XXX_SLI_IQ_INSTR_COUNT(i);
+		len += sprintf(s + len, "[%x] (INSTR_COUNT_%d): %08x\n",
+			       reg, i, octeon_read_csr(oct, reg));
+	}
+
+	/* PCI  DMA registers */
+
+	len += sprintf(s + len, "\n[%x] (DMA_CNT_0): %08x\n",
+		       CN6XXX_DMA_CNT(0),
+		       octeon_read_csr(oct, CN6XXX_DMA_CNT(0)));
+	reg = CN6XXX_DMA_PKT_INT_LEVEL(0);
+	len += sprintf(s + len, "[%x] (DMA_INT_LEV_0): %08x\n",
+		       CN6XXX_DMA_PKT_INT_LEVEL(0), octeon_read_csr(oct, reg));
+	reg = CN6XXX_DMA_TIME_INT_LEVEL(0);
+	len += sprintf(s + len, "[%x] (DMA_TIME_0): %08x\n",
+		       CN6XXX_DMA_TIME_INT_LEVEL(0),
+		       octeon_read_csr(oct, reg));
+
+	len += sprintf(s + len, "\n[%x] (DMA_CNT_1): %08x\n",
+		       CN6XXX_DMA_CNT(1),
+		       octeon_read_csr(oct, CN6XXX_DMA_CNT(1)));
+	reg = CN6XXX_DMA_PKT_INT_LEVEL(1);
+	len += sprintf(s + len, "[%x] (DMA_INT_LEV_1): %08x\n",
+		       CN6XXX_DMA_PKT_INT_LEVEL(1),
+		       octeon_read_csr(oct, reg));
+	reg = CN6XXX_DMA_PKT_INT_LEVEL(1);
+	len += sprintf(s + len, "[%x] (DMA_TIME_1): %08x\n",
+		       CN6XXX_DMA_TIME_INT_LEVEL(1),
+		       octeon_read_csr(oct, reg));
+
+	/* PCI  Index registers */
+
+	len += sprintf(s + len, "\n");
+
+	for (i = 0; i < 16; i++) {
+		reg = lio_pci_readq(oct, CN6XXX_BAR1_REG(i, oct->pcie_port));
+		len += sprintf(s + len, "[%llx] (BAR1_INDEX_%02d): %08x\n",
+			       CN6XXX_BAR1_REG(i, oct->pcie_port), i, reg);
+	}
+
+	return len;
+}
+
+static int cn6xxx_read_config_reg(char *s, struct octeon_device *oct)
+{
+	u32 val;
+	int i, len = 0;
+
+	/* PCI CONFIG Registers */
+
+	len += sprintf(s + len,
+		       "\n\t Octeon Config space Registers\n\n");
+
+	for (i = 0; i <= 13; i++) {
+		pci_read_config_dword(oct->pci_dev, (i * 4), &val);
+		len += sprintf(s + len, "[0x%x] (Config[%d]): 0x%08x\n",
+			       (i * 4), i, val);
+	}
+
+	for (i = 30; i <= 34; i++) {
+		pci_read_config_dword(oct->pci_dev, (i * 4), &val);
+		len += sprintf(s + len, "[0x%x] (Config[%d]): 0x%08x\n",
+			       (i * 4), i, val);
+	}
+
+	return len;
+}
+
+/*  Return register dump user app.  */
+static void lio_get_regs(struct net_device *dev,
+			 struct ethtool_regs *regs, void *regbuf)
+{
+	struct lio *lio = GET_LIO(dev);
+	int len = 0;
+	struct octeon_device *oct = lio->oct_dev;
+
+	regs->version = OCT_ETHTOOL_REGSVER;
+
+	switch (oct->chip_id) {
+	case OCTEON_CN23XX_PF_VID:
+		memset(regbuf, 0, OCT_ETHTOOL_REGDUMP_LEN_23XX);
+		len += cn23xx_read_csr_reg(regbuf + len, oct);
+		break;
+	case OCTEON_CN23XX_VF_VID:
+		memset(regbuf, 0, OCT_ETHTOOL_REGDUMP_LEN_23XX_VF);
+		len += cn23xx_vf_read_csr_reg(regbuf + len, oct);
+		break;
+	case OCTEON_CN68XX:
+	case OCTEON_CN66XX:
+		memset(regbuf, 0, OCT_ETHTOOL_REGDUMP_LEN);
+		len += cn6xxx_read_csr_reg(regbuf + len, oct);
+		len += cn6xxx_read_config_reg(regbuf + len, oct);
+		break;
+	default:
+		dev_err(&oct->pci_dev->dev, "%s Unknown chipid: %d\n",
+			__func__, oct->chip_id);
+	}
+}
+
+static u32 lio_get_priv_flags(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+
+	return lio->oct_dev->priv_flags;
+}
+
+static int lio_set_priv_flags(struct net_device *netdev, u32 flags)
+{
+	struct lio *lio = GET_LIO(netdev);
+	bool intr_by_tx_bytes = !!(flags & (0x1 << OCT_PRIV_FLAG_TX_BYTES));
+
+	lio_set_priv_flag(lio->oct_dev, OCT_PRIV_FLAG_TX_BYTES,
+			  intr_by_tx_bytes);
+	return 0;
+}
+
+static int lio_get_fecparam(struct net_device *netdev,
+			    struct ethtool_fecparam *fec)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+
+	fec->active_fec = ETHTOOL_FEC_NONE;
+	fec->fec = ETHTOOL_FEC_NONE;
+
+	if (oct->subsystem_id == OCTEON_CN2350_25GB_SUBSYS_ID ||
+	    oct->subsystem_id == OCTEON_CN2360_25GB_SUBSYS_ID) {
+		if (oct->no_speed_setting == 1)
+			return 0;
+
+		liquidio_get_fec(lio);
+		fec->fec = (ETHTOOL_FEC_RS | ETHTOOL_FEC_OFF);
+		if (oct->props[lio->ifidx].fec == 1)
+			fec->active_fec = ETHTOOL_FEC_RS;
+		else
+			fec->active_fec = ETHTOOL_FEC_OFF;
+	}
+
+	return 0;
+}
+
+static int lio_set_fecparam(struct net_device *netdev,
+			    struct ethtool_fecparam *fec)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+
+	if (oct->subsystem_id == OCTEON_CN2350_25GB_SUBSYS_ID ||
+	    oct->subsystem_id == OCTEON_CN2360_25GB_SUBSYS_ID) {
+		if (oct->no_speed_setting == 1)
+			return -EOPNOTSUPP;
+
+		if (fec->fec & ETHTOOL_FEC_OFF)
+			liquidio_set_fec(lio, 0);
+		else if (fec->fec & ETHTOOL_FEC_RS)
+			liquidio_set_fec(lio, 1);
+		else
+			return -EOPNOTSUPP;
+	} else {
+		return -EOPNOTSUPP;
+	}
+
+	return 0;
+}
+
+#define LIO_ETHTOOL_COALESCE	(ETHTOOL_COALESCE_RX_USECS |		\
+				 ETHTOOL_COALESCE_MAX_FRAMES |		\
+				 ETHTOOL_COALESCE_USE_ADAPTIVE |	\
+				 ETHTOOL_COALESCE_RX_MAX_FRAMES_LOW |	\
+				 ETHTOOL_COALESCE_TX_MAX_FRAMES_LOW |	\
+				 ETHTOOL_COALESCE_RX_MAX_FRAMES_HIGH |	\
+				 ETHTOOL_COALESCE_TX_MAX_FRAMES_HIGH |	\
+				 ETHTOOL_COALESCE_PKT_RATE_RX_USECS)
+
+static const struct ethtool_ops lio_ethtool_ops = {
+	.supported_coalesce_params = LIO_ETHTOOL_COALESCE,
+	.get_link_ksettings	= lio_get_link_ksettings,
+	.set_link_ksettings	= lio_set_link_ksettings,
+	.get_fecparam		= lio_get_fecparam,
+	.set_fecparam		= lio_set_fecparam,
+	.get_link		= ethtool_op_get_link,
+	.get_drvinfo		= lio_get_drvinfo,
+	.get_ringparam		= lio_ethtool_get_ringparam,
+	.set_ringparam		= lio_ethtool_set_ringparam,
+	.get_channels		= lio_ethtool_get_channels,
+	.set_channels		= lio_ethtool_set_channels,
+	.set_phys_id		= lio_set_phys_id,
+	.get_eeprom_len		= lio_get_eeprom_len,
+	.get_eeprom		= lio_get_eeprom,
+	.get_strings		= lio_get_strings,
+	.get_ethtool_stats	= lio_get_ethtool_stats,
+	.get_pauseparam		= lio_get_pauseparam,
+	.set_pauseparam		= lio_set_pauseparam,
+	.get_regs_len		= lio_get_regs_len,
+	.get_regs		= lio_get_regs,
+	.get_msglevel		= lio_get_msglevel,
+	.set_msglevel		= lio_set_msglevel,
+	.get_sset_count		= lio_get_sset_count,
+	.get_coalesce		= lio_get_intr_coalesce,
+	.set_coalesce		= lio_set_intr_coalesce,
+	.get_priv_flags		= lio_get_priv_flags,
+	.set_priv_flags		= lio_set_priv_flags,
+	.get_ts_info		= lio_get_ts_info,
+};
+
+static const struct ethtool_ops lio_vf_ethtool_ops = {
+	.supported_coalesce_params = LIO_ETHTOOL_COALESCE,
+	.get_link_ksettings	= lio_get_link_ksettings,
+	.get_link		= ethtool_op_get_link,
+	.get_drvinfo		= lio_get_vf_drvinfo,
+	.get_ringparam		= lio_ethtool_get_ringparam,
+	.set_ringparam          = lio_ethtool_set_ringparam,
+	.get_channels		= lio_ethtool_get_channels,
+	.set_channels		= lio_ethtool_set_channels,
+	.get_strings		= lio_vf_get_strings,
+	.get_ethtool_stats	= lio_vf_get_ethtool_stats,
+	.get_regs_len		= lio_get_regs_len,
+	.get_regs		= lio_get_regs,
+	.get_msglevel		= lio_get_msglevel,
+	.set_msglevel		= lio_vf_set_msglevel,
+	.get_sset_count		= lio_vf_get_sset_count,
+	.get_coalesce		= lio_get_intr_coalesce,
+	.set_coalesce		= lio_set_intr_coalesce,
+	.get_priv_flags		= lio_get_priv_flags,
+	.set_priv_flags		= lio_set_priv_flags,
+	.get_ts_info		= lio_get_ts_info,
+};
+
+void liquidio_set_ethtool_ops(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+
+	if (OCTEON_CN23XX_VF(oct))
+		netdev->ethtool_ops = &lio_vf_ethtool_ops;
+	else
+		netdev->ethtool_ops = &lio_ethtool_ops;
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/lio_main.c b/drivers/net/ethernet/cavium/liquidio/lio_main.c
new file mode 100644
index 000000000..98793b2ac
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/lio_main.c
@@ -0,0 +1,4366 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include <linux/firmware.h>
+#include <net/vxlan.h>
+#include <linux/kthread.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_nic.h"
+#include "octeon_main.h"
+#include "octeon_network.h"
+#include "cn66xx_regs.h"
+#include "cn66xx_device.h"
+#include "cn68xx_device.h"
+#include "cn23xx_pf_device.h"
+#include "liquidio_image.h"
+#include "lio_vf_rep.h"
+
+MODULE_AUTHOR("Cavium Networks, <support@cavium.com>");
+MODULE_DESCRIPTION("Cavium LiquidIO Intelligent Server Adapter Driver");
+MODULE_LICENSE("GPL");
+MODULE_FIRMWARE(LIO_FW_DIR LIO_FW_BASE_NAME LIO_210SV_NAME
+		"_" LIO_FW_NAME_TYPE_NIC LIO_FW_NAME_SUFFIX);
+MODULE_FIRMWARE(LIO_FW_DIR LIO_FW_BASE_NAME LIO_210NV_NAME
+		"_" LIO_FW_NAME_TYPE_NIC LIO_FW_NAME_SUFFIX);
+MODULE_FIRMWARE(LIO_FW_DIR LIO_FW_BASE_NAME LIO_410NV_NAME
+		"_" LIO_FW_NAME_TYPE_NIC LIO_FW_NAME_SUFFIX);
+MODULE_FIRMWARE(LIO_FW_DIR LIO_FW_BASE_NAME LIO_23XX_NAME
+		"_" LIO_FW_NAME_TYPE_NIC LIO_FW_NAME_SUFFIX);
+
+static int ddr_timeout = 10000;
+module_param(ddr_timeout, int, 0644);
+MODULE_PARM_DESC(ddr_timeout,
+		 "Number of milliseconds to wait for DDR initialization. 0 waits for ddr_timeout to be set to non-zero value before starting to check");
+
+#define DEFAULT_MSG_ENABLE (NETIF_MSG_DRV | NETIF_MSG_PROBE | NETIF_MSG_LINK)
+
+static int debug = -1;
+module_param(debug, int, 0644);
+MODULE_PARM_DESC(debug, "NETIF_MSG debug bits");
+
+static char fw_type[LIO_MAX_FW_TYPE_LEN] = LIO_FW_NAME_TYPE_AUTO;
+module_param_string(fw_type, fw_type, sizeof(fw_type), 0444);
+MODULE_PARM_DESC(fw_type, "Type of firmware to be loaded (default is \"auto\"), which uses firmware in flash, if present, else loads \"nic\".");
+
+static u32 console_bitmask;
+module_param(console_bitmask, int, 0644);
+MODULE_PARM_DESC(console_bitmask,
+		 "Bitmask indicating which consoles have debug output redirected to syslog.");
+
+/**
+ * octeon_console_debug_enabled - determines if a given console has debug enabled.
+ * @console: console to check
+ * Return:  1 = enabled. 0 otherwise
+ */
+static int octeon_console_debug_enabled(u32 console)
+{
+	return (console_bitmask >> (console)) & 0x1;
+}
+
+/* Polling interval for determining when NIC application is alive */
+#define LIQUIDIO_STARTER_POLL_INTERVAL_MS 100
+
+/* runtime link query interval */
+#define LIQUIDIO_LINK_QUERY_INTERVAL_MS         1000
+/* update localtime to octeon firmware every 60 seconds.
+ * make firmware to use same time reference, so that it will be easy to
+ * correlate firmware logged events/errors with host events, for debugging.
+ */
+#define LIO_SYNC_OCTEON_TIME_INTERVAL_MS 60000
+
+/* time to wait for possible in-flight requests in milliseconds */
+#define WAIT_INFLIGHT_REQUEST	msecs_to_jiffies(1000)
+
+struct oct_link_status_resp {
+	u64 rh;
+	struct oct_link_info link_info;
+	u64 status;
+};
+
+struct oct_timestamp_resp {
+	u64 rh;
+	u64 timestamp;
+	u64 status;
+};
+
+#define OCT_TIMESTAMP_RESP_SIZE (sizeof(struct oct_timestamp_resp))
+
+union tx_info {
+	u64 u64;
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		u16 gso_size;
+		u16 gso_segs;
+		u32 reserved;
+#else
+		u32 reserved;
+		u16 gso_segs;
+		u16 gso_size;
+#endif
+	} s;
+};
+
+/* Octeon device properties to be used by the NIC module.
+ * Each octeon device in the system will be represented
+ * by this structure in the NIC module.
+ */
+
+#define OCTNIC_GSO_MAX_HEADER_SIZE 128
+#define OCTNIC_GSO_MAX_SIZE                                                    \
+	(CN23XX_DEFAULT_INPUT_JABBER - OCTNIC_GSO_MAX_HEADER_SIZE)
+
+struct handshake {
+	struct completion init;
+	struct completion started;
+	struct pci_dev *pci_dev;
+	int init_ok;
+	int started_ok;
+};
+
+#ifdef CONFIG_PCI_IOV
+static int liquidio_enable_sriov(struct pci_dev *dev, int num_vfs);
+#endif
+
+static int octeon_dbg_console_print(struct octeon_device *oct, u32 console_num,
+				    char *prefix, char *suffix);
+
+static int octeon_device_init(struct octeon_device *);
+static int liquidio_stop(struct net_device *netdev);
+static void liquidio_remove(struct pci_dev *pdev);
+static int liquidio_probe(struct pci_dev *pdev,
+			  const struct pci_device_id *ent);
+static int liquidio_set_vf_link_state(struct net_device *netdev, int vfidx,
+				      int linkstate);
+
+static struct handshake handshake[MAX_OCTEON_DEVICES];
+static struct completion first_stage;
+
+static void octeon_droq_bh(struct tasklet_struct *t)
+{
+	int q_no;
+	int reschedule = 0;
+	struct octeon_device_priv *oct_priv = from_tasklet(oct_priv, t,
+							  droq_tasklet);
+	struct octeon_device *oct = oct_priv->dev;
+
+	for (q_no = 0; q_no < MAX_OCTEON_OUTPUT_QUEUES(oct); q_no++) {
+		if (!(oct->io_qmask.oq & BIT_ULL(q_no)))
+			continue;
+		reschedule |= octeon_droq_process_packets(oct, oct->droq[q_no],
+							  MAX_PACKET_BUDGET);
+		lio_enable_irq(oct->droq[q_no], NULL);
+
+		if (OCTEON_CN23XX_PF(oct) && oct->msix_on) {
+			/* set time and cnt interrupt thresholds for this DROQ
+			 * for NAPI
+			 */
+			int adjusted_q_no = q_no + oct->sriov_info.pf_srn;
+
+			octeon_write_csr64(
+			    oct, CN23XX_SLI_OQ_PKT_INT_LEVELS(adjusted_q_no),
+			    0x5700000040ULL);
+			octeon_write_csr64(
+			    oct, CN23XX_SLI_OQ_PKTS_SENT(adjusted_q_no), 0);
+		}
+	}
+
+	if (reschedule)
+		tasklet_schedule(&oct_priv->droq_tasklet);
+}
+
+static int lio_wait_for_oq_pkts(struct octeon_device *oct)
+{
+	struct octeon_device_priv *oct_priv =
+		(struct octeon_device_priv *)oct->priv;
+	int retry = 100, pkt_cnt = 0, pending_pkts = 0;
+	int i;
+
+	do {
+		pending_pkts = 0;
+
+		for (i = 0; i < MAX_OCTEON_OUTPUT_QUEUES(oct); i++) {
+			if (!(oct->io_qmask.oq & BIT_ULL(i)))
+				continue;
+			pkt_cnt += octeon_droq_check_hw_for_pkts(oct->droq[i]);
+		}
+		if (pkt_cnt > 0) {
+			pending_pkts += pkt_cnt;
+			tasklet_schedule(&oct_priv->droq_tasklet);
+		}
+		pkt_cnt = 0;
+		schedule_timeout_uninterruptible(1);
+
+	} while (retry-- && pending_pkts);
+
+	return pkt_cnt;
+}
+
+/**
+ * force_io_queues_off - Forces all IO queues off on a given device
+ * @oct: Pointer to Octeon device
+ */
+static void force_io_queues_off(struct octeon_device *oct)
+{
+	if ((oct->chip_id == OCTEON_CN66XX) ||
+	    (oct->chip_id == OCTEON_CN68XX)) {
+		/* Reset the Enable bits for Input Queues. */
+		octeon_write_csr(oct, CN6XXX_SLI_PKT_INSTR_ENB, 0);
+
+		/* Reset the Enable bits for Output Queues. */
+		octeon_write_csr(oct, CN6XXX_SLI_PKT_OUT_ENB, 0);
+	}
+}
+
+/**
+ * pcierror_quiesce_device - Cause device to go quiet so it can be safely removed/reset/etc
+ * @oct: Pointer to Octeon device
+ */
+static inline void pcierror_quiesce_device(struct octeon_device *oct)
+{
+	int i;
+
+	/* Disable the input and output queues now. No more packets will
+	 * arrive from Octeon, but we should wait for all packet processing
+	 * to finish.
+	 */
+	force_io_queues_off(oct);
+
+	/* To allow for in-flight requests */
+	schedule_timeout_uninterruptible(WAIT_INFLIGHT_REQUEST);
+
+	if (wait_for_pending_requests(oct))
+		dev_err(&oct->pci_dev->dev, "There were pending requests\n");
+
+	/* Force all requests waiting to be fetched by OCTEON to complete. */
+	for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct); i++) {
+		struct octeon_instr_queue *iq;
+
+		if (!(oct->io_qmask.iq & BIT_ULL(i)))
+			continue;
+		iq = oct->instr_queue[i];
+
+		if (atomic_read(&iq->instr_pending)) {
+			spin_lock_bh(&iq->lock);
+			iq->fill_cnt = 0;
+			iq->octeon_read_index = iq->host_write_index;
+			iq->stats.instr_processed +=
+				atomic_read(&iq->instr_pending);
+			lio_process_iq_request_list(oct, iq, 0);
+			spin_unlock_bh(&iq->lock);
+		}
+	}
+
+	/* Force all pending ordered list requests to time out. */
+	lio_process_ordered_list(oct, 1);
+
+	/* We do not need to wait for output queue packets to be processed. */
+}
+
+/**
+ * cleanup_aer_uncorrect_error_status - Cleanup PCI AER uncorrectable error status
+ * @dev: Pointer to PCI device
+ */
+static void cleanup_aer_uncorrect_error_status(struct pci_dev *dev)
+{
+	int pos = 0x100;
+	u32 status, mask;
+
+	pr_info("%s :\n", __func__);
+
+	pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_STATUS, &status);
+	pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_SEVER, &mask);
+	if (dev->error_state == pci_channel_io_normal)
+		status &= ~mask;        /* Clear corresponding nonfatal bits */
+	else
+		status &= mask;         /* Clear corresponding fatal bits */
+	pci_write_config_dword(dev, pos + PCI_ERR_UNCOR_STATUS, status);
+}
+
+/**
+ * stop_pci_io - Stop all PCI IO to a given device
+ * @oct: Pointer to Octeon device
+ */
+static void stop_pci_io(struct octeon_device *oct)
+{
+	/* No more instructions will be forwarded. */
+	atomic_set(&oct->status, OCT_DEV_IN_RESET);
+
+	pci_disable_device(oct->pci_dev);
+
+	/* Disable interrupts  */
+	oct->fn_list.disable_interrupt(oct, OCTEON_ALL_INTR);
+
+	pcierror_quiesce_device(oct);
+
+	/* Release the interrupt line */
+	free_irq(oct->pci_dev->irq, oct);
+
+	if (oct->flags & LIO_FLAG_MSI_ENABLED)
+		pci_disable_msi(oct->pci_dev);
+
+	dev_dbg(&oct->pci_dev->dev, "Device state is now %s\n",
+		lio_get_state_string(&oct->status));
+
+	/* making it a common function for all OCTEON models */
+	cleanup_aer_uncorrect_error_status(oct->pci_dev);
+}
+
+/**
+ * liquidio_pcie_error_detected - called when PCI error is detected
+ * @pdev: Pointer to PCI device
+ * @state: The current pci connection state
+ *
+ * This function is called after a PCI bus error affecting
+ * this device has been detected.
+ */
+static pci_ers_result_t liquidio_pcie_error_detected(struct pci_dev *pdev,
+						     pci_channel_state_t state)
+{
+	struct octeon_device *oct = pci_get_drvdata(pdev);
+
+	/* Non-correctable Non-fatal errors */
+	if (state == pci_channel_io_normal) {
+		dev_err(&oct->pci_dev->dev, "Non-correctable non-fatal error reported:\n");
+		cleanup_aer_uncorrect_error_status(oct->pci_dev);
+		return PCI_ERS_RESULT_CAN_RECOVER;
+	}
+
+	/* Non-correctable Fatal errors */
+	dev_err(&oct->pci_dev->dev, "Non-correctable FATAL reported by PCI AER driver\n");
+	stop_pci_io(oct);
+
+	/* Always return a DISCONNECT. There is no support for recovery but only
+	 * for a clean shutdown.
+	 */
+	return PCI_ERS_RESULT_DISCONNECT;
+}
+
+/**
+ * liquidio_pcie_mmio_enabled - mmio handler
+ * @pdev: Pointer to PCI device
+ */
+static pci_ers_result_t liquidio_pcie_mmio_enabled(struct pci_dev __maybe_unused *pdev)
+{
+	/* We should never hit this since we never ask for a reset for a Fatal
+	 * Error. We always return DISCONNECT in io_error above.
+	 * But play safe and return RECOVERED for now.
+	 */
+	return PCI_ERS_RESULT_RECOVERED;
+}
+
+/**
+ * liquidio_pcie_slot_reset - called after the pci bus has been reset.
+ * @pdev: Pointer to PCI device
+ *
+ * Restart the card from scratch, as if from a cold-boot. Implementation
+ * resembles the first-half of the octeon_resume routine.
+ */
+static pci_ers_result_t liquidio_pcie_slot_reset(struct pci_dev __maybe_unused *pdev)
+{
+	/* We should never hit this since we never ask for a reset for a Fatal
+	 * Error. We always return DISCONNECT in io_error above.
+	 * But play safe and return RECOVERED for now.
+	 */
+	return PCI_ERS_RESULT_RECOVERED;
+}
+
+/**
+ * liquidio_pcie_resume - called when traffic can start flowing again.
+ * @pdev: Pointer to PCI device
+ *
+ * This callback is called when the error recovery driver tells us that
+ * its OK to resume normal operation. Implementation resembles the
+ * second-half of the octeon_resume routine.
+ */
+static void liquidio_pcie_resume(struct pci_dev __maybe_unused *pdev)
+{
+	/* Nothing to be done here. */
+}
+
+#define liquidio_suspend NULL
+#define liquidio_resume NULL
+
+/* For PCI-E Advanced Error Recovery (AER) Interface */
+static const struct pci_error_handlers liquidio_err_handler = {
+	.error_detected = liquidio_pcie_error_detected,
+	.mmio_enabled	= liquidio_pcie_mmio_enabled,
+	.slot_reset	= liquidio_pcie_slot_reset,
+	.resume		= liquidio_pcie_resume,
+};
+
+static const struct pci_device_id liquidio_pci_tbl[] = {
+	{       /* 68xx */
+		PCI_VENDOR_ID_CAVIUM, 0x91, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0
+	},
+	{       /* 66xx */
+		PCI_VENDOR_ID_CAVIUM, 0x92, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0
+	},
+	{       /* 23xx pf */
+		PCI_VENDOR_ID_CAVIUM, 0x9702, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0
+	},
+	{
+		0, 0, 0, 0, 0, 0, 0
+	}
+};
+MODULE_DEVICE_TABLE(pci, liquidio_pci_tbl);
+
+static SIMPLE_DEV_PM_OPS(liquidio_pm_ops, liquidio_suspend, liquidio_resume);
+
+static struct pci_driver liquidio_pci_driver = {
+	.name		= "LiquidIO",
+	.id_table	= liquidio_pci_tbl,
+	.probe		= liquidio_probe,
+	.remove		= liquidio_remove,
+	.err_handler	= &liquidio_err_handler,    /* For AER */
+	.driver.pm	= &liquidio_pm_ops,
+#ifdef CONFIG_PCI_IOV
+	.sriov_configure = liquidio_enable_sriov,
+#endif
+};
+
+/**
+ * liquidio_init_pci - register PCI driver
+ */
+static int liquidio_init_pci(void)
+{
+	return pci_register_driver(&liquidio_pci_driver);
+}
+
+/**
+ * liquidio_deinit_pci - unregister PCI driver
+ */
+static void liquidio_deinit_pci(void)
+{
+	pci_unregister_driver(&liquidio_pci_driver);
+}
+
+/**
+ * check_txq_status - Check Tx queue status, and take appropriate action
+ * @lio: per-network private data
+ * Return: 0 if full, number of queues woken up otherwise
+ */
+static inline int check_txq_status(struct lio *lio)
+{
+	int numqs = lio->netdev->real_num_tx_queues;
+	int ret_val = 0;
+	int q, iq;
+
+	/* check each sub-queue state */
+	for (q = 0; q < numqs; q++) {
+		iq = lio->linfo.txpciq[q %
+			lio->oct_dev->num_iqs].s.q_no;
+		if (octnet_iq_is_full(lio->oct_dev, iq))
+			continue;
+		if (__netif_subqueue_stopped(lio->netdev, q)) {
+			netif_wake_subqueue(lio->netdev, q);
+			INCR_INSTRQUEUE_PKT_COUNT(lio->oct_dev, iq,
+						  tx_restart, 1);
+			ret_val++;
+		}
+	}
+
+	return ret_val;
+}
+
+/**
+ * print_link_info -  Print link information
+ * @netdev: network device
+ */
+static void print_link_info(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+
+	if (!ifstate_check(lio, LIO_IFSTATE_RESETTING) &&
+	    ifstate_check(lio, LIO_IFSTATE_REGISTERED)) {
+		struct oct_link_info *linfo = &lio->linfo;
+
+		if (linfo->link.s.link_up) {
+			netif_info(lio, link, lio->netdev, "%d Mbps %s Duplex UP\n",
+				   linfo->link.s.speed,
+				   (linfo->link.s.duplex) ? "Full" : "Half");
+		} else {
+			netif_info(lio, link, lio->netdev, "Link Down\n");
+		}
+	}
+}
+
+/**
+ * octnet_link_status_change - Routine to notify MTU change
+ * @work: work_struct data structure
+ */
+static void octnet_link_status_change(struct work_struct *work)
+{
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct lio *lio = (struct lio *)wk->ctxptr;
+
+	/* lio->linfo.link.s.mtu always contains max MTU of the lio interface.
+	 * this API is invoked only when new max-MTU of the interface is
+	 * less than current MTU.
+	 */
+	rtnl_lock();
+	dev_set_mtu(lio->netdev, lio->linfo.link.s.mtu);
+	rtnl_unlock();
+}
+
+/**
+ * setup_link_status_change_wq - Sets up the mtu status change work
+ * @netdev: network device
+ */
+static inline int setup_link_status_change_wq(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+
+	lio->link_status_wq.wq = alloc_workqueue("link-status",
+						 WQ_MEM_RECLAIM, 0);
+	if (!lio->link_status_wq.wq) {
+		dev_err(&oct->pci_dev->dev, "unable to create cavium link status wq\n");
+		return -1;
+	}
+	INIT_DELAYED_WORK(&lio->link_status_wq.wk.work,
+			  octnet_link_status_change);
+	lio->link_status_wq.wk.ctxptr = lio;
+
+	return 0;
+}
+
+static inline void cleanup_link_status_change_wq(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+
+	if (lio->link_status_wq.wq) {
+		cancel_delayed_work_sync(&lio->link_status_wq.wk.work);
+		destroy_workqueue(lio->link_status_wq.wq);
+	}
+}
+
+/**
+ * update_link_status - Update link status
+ * @netdev: network device
+ * @ls: link status structure
+ *
+ * Called on receipt of a link status response from the core application to
+ * update each interface's link status.
+ */
+static inline void update_link_status(struct net_device *netdev,
+				      union oct_link_status *ls)
+{
+	struct lio *lio = GET_LIO(netdev);
+	int changed = (lio->linfo.link.u64 != ls->u64);
+	int current_max_mtu = lio->linfo.link.s.mtu;
+	struct octeon_device *oct = lio->oct_dev;
+
+	dev_dbg(&oct->pci_dev->dev, "%s: lio->linfo.link.u64=%llx, ls->u64=%llx\n",
+		__func__, lio->linfo.link.u64, ls->u64);
+	lio->linfo.link.u64 = ls->u64;
+
+	if ((lio->intf_open) && (changed)) {
+		print_link_info(netdev);
+		lio->link_changes++;
+
+		if (lio->linfo.link.s.link_up) {
+			dev_dbg(&oct->pci_dev->dev, "%s: link_up", __func__);
+			netif_carrier_on(netdev);
+			wake_txqs(netdev);
+		} else {
+			dev_dbg(&oct->pci_dev->dev, "%s: link_off", __func__);
+			netif_carrier_off(netdev);
+			stop_txqs(netdev);
+		}
+		if (lio->linfo.link.s.mtu != current_max_mtu) {
+			netif_info(lio, probe, lio->netdev, "Max MTU changed from %d to %d\n",
+				   current_max_mtu, lio->linfo.link.s.mtu);
+			netdev->max_mtu = lio->linfo.link.s.mtu;
+		}
+		if (lio->linfo.link.s.mtu < netdev->mtu) {
+			dev_warn(&oct->pci_dev->dev,
+				 "Current MTU is higher than new max MTU; Reducing the current mtu from %d to %d\n",
+				     netdev->mtu, lio->linfo.link.s.mtu);
+			queue_delayed_work(lio->link_status_wq.wq,
+					   &lio->link_status_wq.wk.work, 0);
+		}
+	}
+}
+
+/**
+ * lio_sync_octeon_time - send latest localtime to octeon firmware so that
+ * firmware will correct it's time, in case there is a time skew
+ *
+ * @work: work scheduled to send time update to octeon firmware
+ **/
+static void lio_sync_octeon_time(struct work_struct *work)
+{
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct lio *lio = (struct lio *)wk->ctxptr;
+	struct octeon_device *oct = lio->oct_dev;
+	struct octeon_soft_command *sc;
+	struct timespec64 ts;
+	struct lio_time *lt;
+	int ret;
+
+	sc = octeon_alloc_soft_command(oct, sizeof(struct lio_time), 16, 0);
+	if (!sc) {
+		dev_err(&oct->pci_dev->dev,
+			"Failed to sync time to octeon: soft command allocation failed\n");
+		return;
+	}
+
+	lt = (struct lio_time *)sc->virtdptr;
+
+	/* Get time of the day */
+	ktime_get_real_ts64(&ts);
+	lt->sec = ts.tv_sec;
+	lt->nsec = ts.tv_nsec;
+	octeon_swap_8B_data((u64 *)lt, (sizeof(struct lio_time)) / 8);
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC,
+				    OPCODE_NIC_SYNC_OCTEON_TIME, 0, 0, 0);
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	ret = octeon_send_soft_command(oct, sc);
+	if (ret == IQ_SEND_FAILED) {
+		dev_err(&oct->pci_dev->dev,
+			"Failed to sync time to octeon: failed to send soft command\n");
+		octeon_free_soft_command(oct, sc);
+	} else {
+		WRITE_ONCE(sc->caller_is_done, true);
+	}
+
+	queue_delayed_work(lio->sync_octeon_time_wq.wq,
+			   &lio->sync_octeon_time_wq.wk.work,
+			   msecs_to_jiffies(LIO_SYNC_OCTEON_TIME_INTERVAL_MS));
+}
+
+/**
+ * setup_sync_octeon_time_wq - prepare work to periodically update local time to octeon firmware
+ *
+ * @netdev: network device which should send time update to firmware
+ **/
+static inline int setup_sync_octeon_time_wq(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+
+	lio->sync_octeon_time_wq.wq =
+		alloc_workqueue("update-octeon-time", WQ_MEM_RECLAIM, 0);
+	if (!lio->sync_octeon_time_wq.wq) {
+		dev_err(&oct->pci_dev->dev, "Unable to create wq to update octeon time\n");
+		return -1;
+	}
+	INIT_DELAYED_WORK(&lio->sync_octeon_time_wq.wk.work,
+			  lio_sync_octeon_time);
+	lio->sync_octeon_time_wq.wk.ctxptr = lio;
+	queue_delayed_work(lio->sync_octeon_time_wq.wq,
+			   &lio->sync_octeon_time_wq.wk.work,
+			   msecs_to_jiffies(LIO_SYNC_OCTEON_TIME_INTERVAL_MS));
+
+	return 0;
+}
+
+/**
+ * cleanup_sync_octeon_time_wq - destroy wq
+ *
+ * @netdev: network device which should send time update to firmware
+ *
+ * Stop scheduling and destroy the work created to periodically update local
+ * time to octeon firmware.
+ **/
+static inline void cleanup_sync_octeon_time_wq(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct cavium_wq *time_wq = &lio->sync_octeon_time_wq;
+
+	if (time_wq->wq) {
+		cancel_delayed_work_sync(&time_wq->wk.work);
+		destroy_workqueue(time_wq->wq);
+	}
+}
+
+static struct octeon_device *get_other_octeon_device(struct octeon_device *oct)
+{
+	struct octeon_device *other_oct;
+
+	other_oct = lio_get_device(oct->octeon_id + 1);
+
+	if (other_oct && other_oct->pci_dev) {
+		int oct_busnum, other_oct_busnum;
+
+		oct_busnum = oct->pci_dev->bus->number;
+		other_oct_busnum = other_oct->pci_dev->bus->number;
+
+		if (oct_busnum == other_oct_busnum) {
+			int oct_slot, other_oct_slot;
+
+			oct_slot = PCI_SLOT(oct->pci_dev->devfn);
+			other_oct_slot = PCI_SLOT(other_oct->pci_dev->devfn);
+
+			if (oct_slot == other_oct_slot)
+				return other_oct;
+		}
+	}
+
+	return NULL;
+}
+
+static void disable_all_vf_links(struct octeon_device *oct)
+{
+	struct net_device *netdev;
+	int max_vfs, vf, i;
+
+	if (!oct)
+		return;
+
+	max_vfs = oct->sriov_info.max_vfs;
+
+	for (i = 0; i < oct->ifcount; i++) {
+		netdev = oct->props[i].netdev;
+		if (!netdev)
+			continue;
+
+		for (vf = 0; vf < max_vfs; vf++)
+			liquidio_set_vf_link_state(netdev, vf,
+						   IFLA_VF_LINK_STATE_DISABLE);
+	}
+}
+
+static int liquidio_watchdog(void *param)
+{
+	bool err_msg_was_printed[LIO_MAX_CORES];
+	u16 mask_of_crashed_or_stuck_cores = 0;
+	bool all_vf_links_are_disabled = false;
+	struct octeon_device *oct = param;
+	struct octeon_device *other_oct;
+#ifdef CONFIG_MODULE_UNLOAD
+	long refcount, vfs_referencing_pf;
+	u64 vfs_mask1, vfs_mask2;
+#endif
+	int core;
+
+	memset(err_msg_was_printed, 0, sizeof(err_msg_was_printed));
+
+	while (!kthread_should_stop()) {
+		/* sleep for a couple of seconds so that we don't hog the CPU */
+		set_current_state(TASK_INTERRUPTIBLE);
+		schedule_timeout(msecs_to_jiffies(2000));
+
+		mask_of_crashed_or_stuck_cores =
+		    (u16)octeon_read_csr64(oct, CN23XX_SLI_SCRATCH2);
+
+		if (!mask_of_crashed_or_stuck_cores)
+			continue;
+
+		WRITE_ONCE(oct->cores_crashed, true);
+		other_oct = get_other_octeon_device(oct);
+		if (other_oct)
+			WRITE_ONCE(other_oct->cores_crashed, true);
+
+		for (core = 0; core < LIO_MAX_CORES; core++) {
+			bool core_crashed_or_got_stuck;
+
+			core_crashed_or_got_stuck =
+						(mask_of_crashed_or_stuck_cores
+						 >> core) & 1;
+
+			if (core_crashed_or_got_stuck &&
+			    !err_msg_was_printed[core]) {
+				dev_err(&oct->pci_dev->dev,
+					"ERROR: Octeon core %d crashed or got stuck!  See oct-fwdump for details.\n",
+					core);
+				err_msg_was_printed[core] = true;
+			}
+		}
+
+		if (all_vf_links_are_disabled)
+			continue;
+
+		disable_all_vf_links(oct);
+		disable_all_vf_links(other_oct);
+		all_vf_links_are_disabled = true;
+
+#ifdef CONFIG_MODULE_UNLOAD
+		vfs_mask1 = READ_ONCE(oct->sriov_info.vf_drv_loaded_mask);
+		vfs_mask2 = READ_ONCE(other_oct->sriov_info.vf_drv_loaded_mask);
+
+		vfs_referencing_pf  = hweight64(vfs_mask1);
+		vfs_referencing_pf += hweight64(vfs_mask2);
+
+		refcount = module_refcount(THIS_MODULE);
+		if (refcount >= vfs_referencing_pf) {
+			while (vfs_referencing_pf) {
+				module_put(THIS_MODULE);
+				vfs_referencing_pf--;
+			}
+		}
+#endif
+	}
+
+	return 0;
+}
+
+/**
+ * liquidio_probe - PCI probe handler
+ * @pdev: PCI device structure
+ * @ent: unused
+ */
+static int
+liquidio_probe(struct pci_dev *pdev, const struct pci_device_id __maybe_unused *ent)
+{
+	struct octeon_device *oct_dev = NULL;
+	struct handshake *hs;
+
+	oct_dev = octeon_allocate_device(pdev->device,
+					 sizeof(struct octeon_device_priv));
+	if (!oct_dev) {
+		dev_err(&pdev->dev, "Unable to allocate device\n");
+		return -ENOMEM;
+	}
+
+	if (pdev->device == OCTEON_CN23XX_PF_VID)
+		oct_dev->msix_on = LIO_FLAG_MSIX_ENABLED;
+
+	/* Enable PTP for 6XXX Device */
+	if (((pdev->device == OCTEON_CN66XX) ||
+	     (pdev->device == OCTEON_CN68XX)))
+		oct_dev->ptp_enable = true;
+	else
+		oct_dev->ptp_enable = false;
+
+	dev_info(&pdev->dev, "Initializing device %x:%x.\n",
+		 (u32)pdev->vendor, (u32)pdev->device);
+
+	/* Assign octeon_device for this device to the private data area. */
+	pci_set_drvdata(pdev, oct_dev);
+
+	/* set linux specific device pointer */
+	oct_dev->pci_dev = (void *)pdev;
+
+	oct_dev->subsystem_id = pdev->subsystem_vendor |
+		(pdev->subsystem_device << 16);
+
+	hs = &handshake[oct_dev->octeon_id];
+	init_completion(&hs->init);
+	init_completion(&hs->started);
+	hs->pci_dev = pdev;
+
+	if (oct_dev->octeon_id == 0)
+		/* first LiquidIO NIC is detected */
+		complete(&first_stage);
+
+	if (octeon_device_init(oct_dev)) {
+		complete(&hs->init);
+		liquidio_remove(pdev);
+		return -ENOMEM;
+	}
+
+	if (OCTEON_CN23XX_PF(oct_dev)) {
+		u8 bus, device, function;
+
+		if (atomic_read(oct_dev->adapter_refcount) == 1) {
+			/* Each NIC gets one watchdog kernel thread.  The first
+			 * PF (of each NIC) that gets pci_driver->probe()'d
+			 * creates that thread.
+			 */
+			bus = pdev->bus->number;
+			device = PCI_SLOT(pdev->devfn);
+			function = PCI_FUNC(pdev->devfn);
+			oct_dev->watchdog_task = kthread_run(liquidio_watchdog,
+							     oct_dev,
+							     "liowd/%02hhx:%02hhx.%hhx",
+							     bus, device, function);
+			if (IS_ERR(oct_dev->watchdog_task)) {
+				oct_dev->watchdog_task = NULL;
+				dev_err(&oct_dev->pci_dev->dev,
+					"failed to create kernel_thread\n");
+				liquidio_remove(pdev);
+				return -1;
+			}
+		}
+	}
+
+	oct_dev->rx_pause = 1;
+	oct_dev->tx_pause = 1;
+
+	dev_dbg(&oct_dev->pci_dev->dev, "Device is ready\n");
+
+	return 0;
+}
+
+static bool fw_type_is_auto(void)
+{
+	return strncmp(fw_type, LIO_FW_NAME_TYPE_AUTO,
+		       sizeof(LIO_FW_NAME_TYPE_AUTO)) == 0;
+}
+
+/**
+ * octeon_pci_flr - PCI FLR for each Octeon device.
+ * @oct: octeon device
+ */
+static void octeon_pci_flr(struct octeon_device *oct)
+{
+	int rc;
+
+	pci_save_state(oct->pci_dev);
+
+	pci_cfg_access_lock(oct->pci_dev);
+
+	/* Quiesce the device completely */
+	pci_write_config_word(oct->pci_dev, PCI_COMMAND,
+			      PCI_COMMAND_INTX_DISABLE);
+
+	rc = __pci_reset_function_locked(oct->pci_dev);
+
+	if (rc != 0)
+		dev_err(&oct->pci_dev->dev, "Error %d resetting PCI function %d\n",
+			rc, oct->pf_num);
+
+	pci_cfg_access_unlock(oct->pci_dev);
+
+	pci_restore_state(oct->pci_dev);
+}
+
+/**
+ * octeon_destroy_resources - Destroy resources associated with octeon device
+ * @oct: octeon device
+ */
+static void octeon_destroy_resources(struct octeon_device *oct)
+{
+	int i, refcount;
+	struct msix_entry *msix_entries;
+	struct octeon_device_priv *oct_priv =
+		(struct octeon_device_priv *)oct->priv;
+
+	struct handshake *hs;
+
+	switch (atomic_read(&oct->status)) {
+	case OCT_DEV_RUNNING:
+	case OCT_DEV_CORE_OK:
+
+		/* No more instructions will be forwarded. */
+		atomic_set(&oct->status, OCT_DEV_IN_RESET);
+
+		oct->app_mode = CVM_DRV_INVALID_APP;
+		dev_dbg(&oct->pci_dev->dev, "Device state is now %s\n",
+			lio_get_state_string(&oct->status));
+
+		schedule_timeout_uninterruptible(HZ / 10);
+
+		fallthrough;
+	case OCT_DEV_HOST_OK:
+
+	case OCT_DEV_CONSOLE_INIT_DONE:
+		/* Remove any consoles */
+		octeon_remove_consoles(oct);
+
+		fallthrough;
+	case OCT_DEV_IO_QUEUES_DONE:
+		if (lio_wait_for_instr_fetch(oct))
+			dev_err(&oct->pci_dev->dev, "IQ had pending instructions\n");
+
+		if (wait_for_pending_requests(oct))
+			dev_err(&oct->pci_dev->dev, "There were pending requests\n");
+
+		/* Disable the input and output queues now. No more packets will
+		 * arrive from Octeon, but we should wait for all packet
+		 * processing to finish.
+		 */
+		oct->fn_list.disable_io_queues(oct);
+
+		if (lio_wait_for_oq_pkts(oct))
+			dev_err(&oct->pci_dev->dev, "OQ had pending packets\n");
+
+		/* Force all requests waiting to be fetched by OCTEON to
+		 * complete.
+		 */
+		for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct); i++) {
+			struct octeon_instr_queue *iq;
+
+			if (!(oct->io_qmask.iq & BIT_ULL(i)))
+				continue;
+			iq = oct->instr_queue[i];
+
+			if (atomic_read(&iq->instr_pending)) {
+				spin_lock_bh(&iq->lock);
+				iq->fill_cnt = 0;
+				iq->octeon_read_index = iq->host_write_index;
+				iq->stats.instr_processed +=
+					atomic_read(&iq->instr_pending);
+				lio_process_iq_request_list(oct, iq, 0);
+				spin_unlock_bh(&iq->lock);
+			}
+		}
+
+		lio_process_ordered_list(oct, 1);
+		octeon_free_sc_done_list(oct);
+		octeon_free_sc_zombie_list(oct);
+
+		fallthrough;
+	case OCT_DEV_INTR_SET_DONE:
+		/* Disable interrupts  */
+		oct->fn_list.disable_interrupt(oct, OCTEON_ALL_INTR);
+
+		if (oct->msix_on) {
+			msix_entries = (struct msix_entry *)oct->msix_entries;
+			for (i = 0; i < oct->num_msix_irqs - 1; i++) {
+				if (oct->ioq_vector[i].vector) {
+					/* clear the affinity_cpumask */
+					irq_set_affinity_hint(
+							msix_entries[i].vector,
+							NULL);
+					free_irq(msix_entries[i].vector,
+						 &oct->ioq_vector[i]);
+					oct->ioq_vector[i].vector = 0;
+				}
+			}
+			/* non-iov vector's argument is oct struct */
+			free_irq(msix_entries[i].vector, oct);
+
+			pci_disable_msix(oct->pci_dev);
+			kfree(oct->msix_entries);
+			oct->msix_entries = NULL;
+		} else {
+			/* Release the interrupt line */
+			free_irq(oct->pci_dev->irq, oct);
+
+			if (oct->flags & LIO_FLAG_MSI_ENABLED)
+				pci_disable_msi(oct->pci_dev);
+		}
+
+		kfree(oct->irq_name_storage);
+		oct->irq_name_storage = NULL;
+
+		fallthrough;
+	case OCT_DEV_MSIX_ALLOC_VECTOR_DONE:
+		if (OCTEON_CN23XX_PF(oct))
+			octeon_free_ioq_vector(oct);
+
+		fallthrough;
+	case OCT_DEV_MBOX_SETUP_DONE:
+		if (OCTEON_CN23XX_PF(oct))
+			oct->fn_list.free_mbox(oct);
+
+		fallthrough;
+	case OCT_DEV_IN_RESET:
+	case OCT_DEV_DROQ_INIT_DONE:
+		/* Wait for any pending operations */
+		mdelay(100);
+		for (i = 0; i < MAX_OCTEON_OUTPUT_QUEUES(oct); i++) {
+			if (!(oct->io_qmask.oq & BIT_ULL(i)))
+				continue;
+			octeon_delete_droq(oct, i);
+		}
+
+		/* Force any pending handshakes to complete */
+		for (i = 0; i < MAX_OCTEON_DEVICES; i++) {
+			hs = &handshake[i];
+
+			if (hs->pci_dev) {
+				handshake[oct->octeon_id].init_ok = 0;
+				complete(&handshake[oct->octeon_id].init);
+				handshake[oct->octeon_id].started_ok = 0;
+				complete(&handshake[oct->octeon_id].started);
+			}
+		}
+
+		fallthrough;
+	case OCT_DEV_RESP_LIST_INIT_DONE:
+		octeon_delete_response_list(oct);
+
+		fallthrough;
+	case OCT_DEV_INSTR_QUEUE_INIT_DONE:
+		for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct); i++) {
+			if (!(oct->io_qmask.iq & BIT_ULL(i)))
+				continue;
+			octeon_delete_instr_queue(oct, i);
+		}
+#ifdef CONFIG_PCI_IOV
+		if (oct->sriov_info.sriov_enabled)
+			pci_disable_sriov(oct->pci_dev);
+#endif
+		fallthrough;
+	case OCT_DEV_SC_BUFF_POOL_INIT_DONE:
+		octeon_free_sc_buffer_pool(oct);
+
+		fallthrough;
+	case OCT_DEV_DISPATCH_INIT_DONE:
+		octeon_delete_dispatch_list(oct);
+		cancel_delayed_work_sync(&oct->nic_poll_work.work);
+
+		fallthrough;
+	case OCT_DEV_PCI_MAP_DONE:
+		refcount = octeon_deregister_device(oct);
+
+		/* Soft reset the octeon device before exiting.
+		 * However, if fw was loaded from card (i.e. autoboot),
+		 * perform an FLR instead.
+		 * Implementation note: only soft-reset the device
+		 * if it is a CN6XXX OR the LAST CN23XX device.
+		 */
+		if (atomic_read(oct->adapter_fw_state) == FW_IS_PRELOADED)
+			octeon_pci_flr(oct);
+		else if (OCTEON_CN6XXX(oct) || !refcount)
+			oct->fn_list.soft_reset(oct);
+
+		octeon_unmap_pci_barx(oct, 0);
+		octeon_unmap_pci_barx(oct, 1);
+
+		fallthrough;
+	case OCT_DEV_PCI_ENABLE_DONE:
+		pci_clear_master(oct->pci_dev);
+		/* Disable the device, releasing the PCI INT */
+		pci_disable_device(oct->pci_dev);
+
+		fallthrough;
+	case OCT_DEV_BEGIN_STATE:
+		/* Nothing to be done here either */
+		break;
+	}                       /* end switch (oct->status) */
+
+	tasklet_kill(&oct_priv->droq_tasklet);
+}
+
+/**
+ * send_rx_ctrl_cmd - Send Rx control command
+ * @lio: per-network private data
+ * @start_stop: whether to start or stop
+ */
+static int send_rx_ctrl_cmd(struct lio *lio, int start_stop)
+{
+	struct octeon_soft_command *sc;
+	union octnet_cmd *ncmd;
+	struct octeon_device *oct = (struct octeon_device *)lio->oct_dev;
+	int retval;
+
+	if (oct->props[lio->ifidx].rx_on == start_stop)
+		return 0;
+
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct, OCTNET_CMD_SIZE,
+					  16, 0);
+	if (!sc) {
+		netif_info(lio, rx_err, lio->netdev,
+			   "Failed to allocate octeon_soft_command struct\n");
+		return -ENOMEM;
+	}
+
+	ncmd = (union octnet_cmd *)sc->virtdptr;
+
+	ncmd->u64 = 0;
+	ncmd->s.cmd = OCTNET_CMD_RX_CTL;
+	ncmd->s.param1 = start_stop;
+
+	octeon_swap_8B_data((u64 *)ncmd, (OCTNET_CMD_SIZE >> 3));
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC,
+				    OPCODE_NIC_CMD, 0, 0, 0);
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	retval = octeon_send_soft_command(oct, sc);
+	if (retval == IQ_SEND_FAILED) {
+		netif_info(lio, rx_err, lio->netdev, "Failed to send RX Control message\n");
+		octeon_free_soft_command(oct, sc);
+	} else {
+		/* Sleep on a wait queue till the cond flag indicates that the
+		 * response arrived or timed-out.
+		 */
+		retval = wait_for_sc_completion_timeout(oct, sc, 0);
+		if (retval)
+			return retval;
+
+		oct->props[lio->ifidx].rx_on = start_stop;
+		WRITE_ONCE(sc->caller_is_done, true);
+	}
+
+	return retval;
+}
+
+/**
+ * liquidio_destroy_nic_device - Destroy NIC device interface
+ * @oct: octeon device
+ * @ifidx: which interface to destroy
+ *
+ * Cleanup associated with each interface for an Octeon device  when NIC
+ * module is being unloaded or if initialization fails during load.
+ */
+static void liquidio_destroy_nic_device(struct octeon_device *oct, int ifidx)
+{
+	struct net_device *netdev = oct->props[ifidx].netdev;
+	struct octeon_device_priv *oct_priv =
+		(struct octeon_device_priv *)oct->priv;
+	struct napi_struct *napi, *n;
+	struct lio *lio;
+
+	if (!netdev) {
+		dev_err(&oct->pci_dev->dev, "%s No netdevice ptr for index %d\n",
+			__func__, ifidx);
+		return;
+	}
+
+	lio = GET_LIO(netdev);
+
+	dev_dbg(&oct->pci_dev->dev, "NIC device cleanup\n");
+
+	if (atomic_read(&lio->ifstate) & LIO_IFSTATE_RUNNING)
+		liquidio_stop(netdev);
+
+	if (oct->props[lio->ifidx].napi_enabled == 1) {
+		list_for_each_entry_safe(napi, n, &netdev->napi_list, dev_list)
+			napi_disable(napi);
+
+		oct->props[lio->ifidx].napi_enabled = 0;
+
+		if (OCTEON_CN23XX_PF(oct))
+			oct->droq[0]->ops.poll_mode = 0;
+	}
+
+	/* Delete NAPI */
+	list_for_each_entry_safe(napi, n, &netdev->napi_list, dev_list)
+		netif_napi_del(napi);
+
+	tasklet_enable(&oct_priv->droq_tasklet);
+
+	if (atomic_read(&lio->ifstate) & LIO_IFSTATE_REGISTERED)
+		unregister_netdev(netdev);
+
+	cleanup_sync_octeon_time_wq(netdev);
+	cleanup_link_status_change_wq(netdev);
+
+	cleanup_rx_oom_poll_fn(netdev);
+
+	lio_delete_glists(lio);
+
+	free_netdev(netdev);
+
+	oct->props[ifidx].gmxport = -1;
+
+	oct->props[ifidx].netdev = NULL;
+}
+
+/**
+ * liquidio_stop_nic_module - Stop complete NIC functionality
+ * @oct: octeon device
+ */
+static int liquidio_stop_nic_module(struct octeon_device *oct)
+{
+	int i, j;
+	struct lio *lio;
+
+	dev_dbg(&oct->pci_dev->dev, "Stopping network interfaces\n");
+	device_lock(&oct->pci_dev->dev);
+	if (oct->devlink) {
+		devlink_unregister(oct->devlink);
+		devlink_free(oct->devlink);
+		oct->devlink = NULL;
+	}
+	device_unlock(&oct->pci_dev->dev);
+
+	if (!oct->ifcount) {
+		dev_err(&oct->pci_dev->dev, "Init for Octeon was not completed\n");
+		return 1;
+	}
+
+	spin_lock_bh(&oct->cmd_resp_wqlock);
+	oct->cmd_resp_state = OCT_DRV_OFFLINE;
+	spin_unlock_bh(&oct->cmd_resp_wqlock);
+
+	lio_vf_rep_destroy(oct);
+
+	for (i = 0; i < oct->ifcount; i++) {
+		lio = GET_LIO(oct->props[i].netdev);
+		for (j = 0; j < oct->num_oqs; j++)
+			octeon_unregister_droq_ops(oct,
+						   lio->linfo.rxpciq[j].s.q_no);
+	}
+
+	for (i = 0; i < oct->ifcount; i++)
+		liquidio_destroy_nic_device(oct, i);
+
+	dev_dbg(&oct->pci_dev->dev, "Network interfaces stopped\n");
+	return 0;
+}
+
+/**
+ * liquidio_remove - Cleans up resources at unload time
+ * @pdev: PCI device structure
+ */
+static void liquidio_remove(struct pci_dev *pdev)
+{
+	struct octeon_device *oct_dev = pci_get_drvdata(pdev);
+
+	dev_dbg(&oct_dev->pci_dev->dev, "Stopping device\n");
+
+	if (oct_dev->watchdog_task)
+		kthread_stop(oct_dev->watchdog_task);
+
+	if (!oct_dev->octeon_id &&
+	    oct_dev->fw_info.app_cap_flags & LIQUIDIO_SWITCHDEV_CAP)
+		lio_vf_rep_modexit();
+
+	if (oct_dev->app_mode && (oct_dev->app_mode == CVM_DRV_NIC_APP))
+		liquidio_stop_nic_module(oct_dev);
+
+	/* Reset the octeon device and cleanup all memory allocated for
+	 * the octeon device by driver.
+	 */
+	octeon_destroy_resources(oct_dev);
+
+	dev_info(&oct_dev->pci_dev->dev, "Device removed\n");
+
+	/* This octeon device has been removed. Update the global
+	 * data structure to reflect this. Free the device structure.
+	 */
+	octeon_free_device_mem(oct_dev);
+}
+
+/**
+ * octeon_chip_specific_setup - Identify the Octeon device and to map the BAR address space
+ * @oct: octeon device
+ */
+static int octeon_chip_specific_setup(struct octeon_device *oct)
+{
+	u32 dev_id, rev_id;
+	int ret = 1;
+
+	pci_read_config_dword(oct->pci_dev, 0, &dev_id);
+	pci_read_config_dword(oct->pci_dev, 8, &rev_id);
+	oct->rev_id = rev_id & 0xff;
+
+	switch (dev_id) {
+	case OCTEON_CN68XX_PCIID:
+		oct->chip_id = OCTEON_CN68XX;
+		ret = lio_setup_cn68xx_octeon_device(oct);
+		break;
+
+	case OCTEON_CN66XX_PCIID:
+		oct->chip_id = OCTEON_CN66XX;
+		ret = lio_setup_cn66xx_octeon_device(oct);
+		break;
+
+	case OCTEON_CN23XX_PCIID_PF:
+		oct->chip_id = OCTEON_CN23XX_PF_VID;
+		ret = setup_cn23xx_octeon_pf_device(oct);
+		if (ret)
+			break;
+#ifdef CONFIG_PCI_IOV
+		if (!ret)
+			pci_sriov_set_totalvfs(oct->pci_dev,
+					       oct->sriov_info.max_vfs);
+#endif
+		break;
+
+	default:
+		dev_err(&oct->pci_dev->dev, "Unknown device found (dev_id: %x)\n",
+			dev_id);
+	}
+
+	return ret;
+}
+
+/**
+ * octeon_pci_os_setup - PCI initialization for each Octeon device.
+ * @oct: octeon device
+ */
+static int octeon_pci_os_setup(struct octeon_device *oct)
+{
+	/* setup PCI stuff first */
+	if (pci_enable_device(oct->pci_dev)) {
+		dev_err(&oct->pci_dev->dev, "pci_enable_device failed\n");
+		return 1;
+	}
+
+	if (dma_set_mask_and_coherent(&oct->pci_dev->dev, DMA_BIT_MASK(64))) {
+		dev_err(&oct->pci_dev->dev, "Unexpected DMA device capability\n");
+		pci_disable_device(oct->pci_dev);
+		return 1;
+	}
+
+	/* Enable PCI DMA Master. */
+	pci_set_master(oct->pci_dev);
+
+	return 0;
+}
+
+/**
+ * free_netbuf - Unmap and free network buffer
+ * @buf: buffer
+ */
+static void free_netbuf(void *buf)
+{
+	struct sk_buff *skb;
+	struct octnet_buf_free_info *finfo;
+	struct lio *lio;
+
+	finfo = (struct octnet_buf_free_info *)buf;
+	skb = finfo->skb;
+	lio = finfo->lio;
+
+	dma_unmap_single(&lio->oct_dev->pci_dev->dev, finfo->dptr, skb->len,
+			 DMA_TO_DEVICE);
+
+	tx_buffer_free(skb);
+}
+
+/**
+ * free_netsgbuf - Unmap and free gather buffer
+ * @buf: buffer
+ */
+static void free_netsgbuf(void *buf)
+{
+	struct octnet_buf_free_info *finfo;
+	struct sk_buff *skb;
+	struct lio *lio;
+	struct octnic_gather *g;
+	int i, frags, iq;
+
+	finfo = (struct octnet_buf_free_info *)buf;
+	skb = finfo->skb;
+	lio = finfo->lio;
+	g = finfo->g;
+	frags = skb_shinfo(skb)->nr_frags;
+
+	dma_unmap_single(&lio->oct_dev->pci_dev->dev,
+			 g->sg[0].ptr[0], (skb->len - skb->data_len),
+			 DMA_TO_DEVICE);
+
+	i = 1;
+	while (frags--) {
+		skb_frag_t *frag = &skb_shinfo(skb)->frags[i - 1];
+
+		dma_unmap_page(&lio->oct_dev->pci_dev->dev,
+			       g->sg[(i >> 2)].ptr[(i & 3)],
+			       skb_frag_size(frag), DMA_TO_DEVICE);
+		i++;
+	}
+
+	iq = skb_iq(lio->oct_dev, skb);
+	spin_lock(&lio->glist_lock[iq]);
+	list_add_tail(&g->list, &lio->glist[iq]);
+	spin_unlock(&lio->glist_lock[iq]);
+
+	tx_buffer_free(skb);
+}
+
+/**
+ * free_netsgbuf_with_resp - Unmap and free gather buffer with response
+ * @buf: buffer
+ */
+static void free_netsgbuf_with_resp(void *buf)
+{
+	struct octeon_soft_command *sc;
+	struct octnet_buf_free_info *finfo;
+	struct sk_buff *skb;
+	struct lio *lio;
+	struct octnic_gather *g;
+	int i, frags, iq;
+
+	sc = (struct octeon_soft_command *)buf;
+	skb = (struct sk_buff *)sc->callback_arg;
+	finfo = (struct octnet_buf_free_info *)&skb->cb;
+
+	lio = finfo->lio;
+	g = finfo->g;
+	frags = skb_shinfo(skb)->nr_frags;
+
+	dma_unmap_single(&lio->oct_dev->pci_dev->dev,
+			 g->sg[0].ptr[0], (skb->len - skb->data_len),
+			 DMA_TO_DEVICE);
+
+	i = 1;
+	while (frags--) {
+		skb_frag_t *frag = &skb_shinfo(skb)->frags[i - 1];
+
+		dma_unmap_page(&lio->oct_dev->pci_dev->dev,
+			       g->sg[(i >> 2)].ptr[(i & 3)],
+			       skb_frag_size(frag), DMA_TO_DEVICE);
+		i++;
+	}
+
+	iq = skb_iq(lio->oct_dev, skb);
+
+	spin_lock(&lio->glist_lock[iq]);
+	list_add_tail(&g->list, &lio->glist[iq]);
+	spin_unlock(&lio->glist_lock[iq]);
+
+	/* Don't free the skb yet */
+}
+
+/**
+ * liquidio_ptp_adjfreq - Adjust ptp frequency
+ * @ptp: PTP clock info
+ * @ppb: how much to adjust by, in parts-per-billion
+ */
+static int liquidio_ptp_adjfreq(struct ptp_clock_info *ptp, s32 ppb)
+{
+	struct lio *lio = container_of(ptp, struct lio, ptp_info);
+	struct octeon_device *oct = (struct octeon_device *)lio->oct_dev;
+	u64 comp, delta;
+	unsigned long flags;
+	bool neg_adj = false;
+
+	if (ppb < 0) {
+		neg_adj = true;
+		ppb = -ppb;
+	}
+
+	/* The hardware adds the clock compensation value to the
+	 * PTP clock on every coprocessor clock cycle, so we
+	 * compute the delta in terms of coprocessor clocks.
+	 */
+	delta = (u64)ppb << 32;
+	do_div(delta, oct->coproc_clock_rate);
+
+	spin_lock_irqsave(&lio->ptp_lock, flags);
+	comp = lio_pci_readq(oct, CN6XXX_MIO_PTP_CLOCK_COMP);
+	if (neg_adj)
+		comp -= delta;
+	else
+		comp += delta;
+	lio_pci_writeq(oct, comp, CN6XXX_MIO_PTP_CLOCK_COMP);
+	spin_unlock_irqrestore(&lio->ptp_lock, flags);
+
+	return 0;
+}
+
+/**
+ * liquidio_ptp_adjtime - Adjust ptp time
+ * @ptp: PTP clock info
+ * @delta: how much to adjust by, in nanosecs
+ */
+static int liquidio_ptp_adjtime(struct ptp_clock_info *ptp, s64 delta)
+{
+	unsigned long flags;
+	struct lio *lio = container_of(ptp, struct lio, ptp_info);
+
+	spin_lock_irqsave(&lio->ptp_lock, flags);
+	lio->ptp_adjust += delta;
+	spin_unlock_irqrestore(&lio->ptp_lock, flags);
+
+	return 0;
+}
+
+/**
+ * liquidio_ptp_gettime - Get hardware clock time, including any adjustment
+ * @ptp: PTP clock info
+ * @ts: timespec
+ */
+static int liquidio_ptp_gettime(struct ptp_clock_info *ptp,
+				struct timespec64 *ts)
+{
+	u64 ns;
+	unsigned long flags;
+	struct lio *lio = container_of(ptp, struct lio, ptp_info);
+	struct octeon_device *oct = (struct octeon_device *)lio->oct_dev;
+
+	spin_lock_irqsave(&lio->ptp_lock, flags);
+	ns = lio_pci_readq(oct, CN6XXX_MIO_PTP_CLOCK_HI);
+	ns += lio->ptp_adjust;
+	spin_unlock_irqrestore(&lio->ptp_lock, flags);
+
+	*ts = ns_to_timespec64(ns);
+
+	return 0;
+}
+
+/**
+ * liquidio_ptp_settime - Set hardware clock time. Reset adjustment
+ * @ptp: PTP clock info
+ * @ts: timespec
+ */
+static int liquidio_ptp_settime(struct ptp_clock_info *ptp,
+				const struct timespec64 *ts)
+{
+	u64 ns;
+	unsigned long flags;
+	struct lio *lio = container_of(ptp, struct lio, ptp_info);
+	struct octeon_device *oct = (struct octeon_device *)lio->oct_dev;
+
+	ns = timespec64_to_ns(ts);
+
+	spin_lock_irqsave(&lio->ptp_lock, flags);
+	lio_pci_writeq(oct, ns, CN6XXX_MIO_PTP_CLOCK_HI);
+	lio->ptp_adjust = 0;
+	spin_unlock_irqrestore(&lio->ptp_lock, flags);
+
+	return 0;
+}
+
+/**
+ * liquidio_ptp_enable - Check if PTP is enabled
+ * @ptp: PTP clock info
+ * @rq: request
+ * @on: is it on
+ */
+static int
+liquidio_ptp_enable(struct ptp_clock_info __maybe_unused *ptp,
+		    struct ptp_clock_request __maybe_unused *rq,
+		    int __maybe_unused on)
+{
+	return -EOPNOTSUPP;
+}
+
+/**
+ * oct_ptp_open - Open PTP clock source
+ * @netdev: network device
+ */
+static void oct_ptp_open(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = (struct octeon_device *)lio->oct_dev;
+
+	spin_lock_init(&lio->ptp_lock);
+
+	snprintf(lio->ptp_info.name, 16, "%s", netdev->name);
+	lio->ptp_info.owner = THIS_MODULE;
+	lio->ptp_info.max_adj = 250000000;
+	lio->ptp_info.n_alarm = 0;
+	lio->ptp_info.n_ext_ts = 0;
+	lio->ptp_info.n_per_out = 0;
+	lio->ptp_info.pps = 0;
+	lio->ptp_info.adjfreq = liquidio_ptp_adjfreq;
+	lio->ptp_info.adjtime = liquidio_ptp_adjtime;
+	lio->ptp_info.gettime64 = liquidio_ptp_gettime;
+	lio->ptp_info.settime64 = liquidio_ptp_settime;
+	lio->ptp_info.enable = liquidio_ptp_enable;
+
+	lio->ptp_adjust = 0;
+
+	lio->ptp_clock = ptp_clock_register(&lio->ptp_info,
+					     &oct->pci_dev->dev);
+
+	if (IS_ERR(lio->ptp_clock))
+		lio->ptp_clock = NULL;
+}
+
+/**
+ * liquidio_ptp_init - Init PTP clock
+ * @oct: octeon device
+ */
+static void liquidio_ptp_init(struct octeon_device *oct)
+{
+	u64 clock_comp, cfg;
+
+	clock_comp = (u64)NSEC_PER_SEC << 32;
+	do_div(clock_comp, oct->coproc_clock_rate);
+	lio_pci_writeq(oct, clock_comp, CN6XXX_MIO_PTP_CLOCK_COMP);
+
+	/* Enable */
+	cfg = lio_pci_readq(oct, CN6XXX_MIO_PTP_CLOCK_CFG);
+	lio_pci_writeq(oct, cfg | 0x01, CN6XXX_MIO_PTP_CLOCK_CFG);
+}
+
+/**
+ * load_firmware - Load firmware to device
+ * @oct: octeon device
+ *
+ * Maps device to firmware filename, requests firmware, and downloads it
+ */
+static int load_firmware(struct octeon_device *oct)
+{
+	int ret = 0;
+	const struct firmware *fw;
+	char fw_name[LIO_MAX_FW_FILENAME_LEN];
+	char *tmp_fw_type;
+
+	if (fw_type_is_auto()) {
+		tmp_fw_type = LIO_FW_NAME_TYPE_NIC;
+		strncpy(fw_type, tmp_fw_type, sizeof(fw_type));
+	} else {
+		tmp_fw_type = fw_type;
+	}
+
+	sprintf(fw_name, "%s%s%s_%s%s", LIO_FW_DIR, LIO_FW_BASE_NAME,
+		octeon_get_conf(oct)->card_name, tmp_fw_type,
+		LIO_FW_NAME_SUFFIX);
+
+	ret = request_firmware(&fw, fw_name, &oct->pci_dev->dev);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "Request firmware failed. Could not find file %s.\n",
+			fw_name);
+		release_firmware(fw);
+		return ret;
+	}
+
+	ret = octeon_download_firmware(oct, fw->data, fw->size);
+
+	release_firmware(fw);
+
+	return ret;
+}
+
+/**
+ * octnet_poll_check_txq_status - Poll routine for checking transmit queue status
+ * @work: work_struct data structure
+ */
+static void octnet_poll_check_txq_status(struct work_struct *work)
+{
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct lio *lio = (struct lio *)wk->ctxptr;
+
+	if (!ifstate_check(lio, LIO_IFSTATE_RUNNING))
+		return;
+
+	check_txq_status(lio);
+	queue_delayed_work(lio->txq_status_wq.wq,
+			   &lio->txq_status_wq.wk.work, msecs_to_jiffies(1));
+}
+
+/**
+ * setup_tx_poll_fn - Sets up the txq poll check
+ * @netdev: network device
+ */
+static inline int setup_tx_poll_fn(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+
+	lio->txq_status_wq.wq = alloc_workqueue("txq-status",
+						WQ_MEM_RECLAIM, 0);
+	if (!lio->txq_status_wq.wq) {
+		dev_err(&oct->pci_dev->dev, "unable to create cavium txq status wq\n");
+		return -1;
+	}
+	INIT_DELAYED_WORK(&lio->txq_status_wq.wk.work,
+			  octnet_poll_check_txq_status);
+	lio->txq_status_wq.wk.ctxptr = lio;
+	queue_delayed_work(lio->txq_status_wq.wq,
+			   &lio->txq_status_wq.wk.work, msecs_to_jiffies(1));
+	return 0;
+}
+
+static inline void cleanup_tx_poll_fn(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+
+	if (lio->txq_status_wq.wq) {
+		cancel_delayed_work_sync(&lio->txq_status_wq.wk.work);
+		destroy_workqueue(lio->txq_status_wq.wq);
+	}
+}
+
+/**
+ * liquidio_open - Net device open for LiquidIO
+ * @netdev: network device
+ */
+static int liquidio_open(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octeon_device_priv *oct_priv =
+		(struct octeon_device_priv *)oct->priv;
+	struct napi_struct *napi, *n;
+	int ret = 0;
+
+	if (oct->props[lio->ifidx].napi_enabled == 0) {
+		tasklet_disable(&oct_priv->droq_tasklet);
+
+		list_for_each_entry_safe(napi, n, &netdev->napi_list, dev_list)
+			napi_enable(napi);
+
+		oct->props[lio->ifidx].napi_enabled = 1;
+
+		if (OCTEON_CN23XX_PF(oct))
+			oct->droq[0]->ops.poll_mode = 1;
+	}
+
+	if (oct->ptp_enable)
+		oct_ptp_open(netdev);
+
+	ifstate_set(lio, LIO_IFSTATE_RUNNING);
+
+	if (!OCTEON_CN23XX_PF(oct) || !oct->msix_on) {
+		ret = setup_tx_poll_fn(netdev);
+		if (ret)
+			goto err_poll;
+	}
+
+	netif_tx_start_all_queues(netdev);
+
+	/* Ready for link status updates */
+	lio->intf_open = 1;
+
+	netif_info(lio, ifup, lio->netdev, "Interface Open, ready for traffic\n");
+
+	/* tell Octeon to start forwarding packets to host */
+	ret = send_rx_ctrl_cmd(lio, 1);
+	if (ret)
+		goto err_rx_ctrl;
+
+	/* start periodical statistics fetch */
+	INIT_DELAYED_WORK(&lio->stats_wk.work, lio_fetch_stats);
+	lio->stats_wk.ctxptr = lio;
+	schedule_delayed_work(&lio->stats_wk.work, msecs_to_jiffies
+					(LIQUIDIO_NDEV_STATS_POLL_TIME_MS));
+
+	dev_info(&oct->pci_dev->dev, "%s interface is opened\n",
+		 netdev->name);
+
+	return 0;
+
+err_rx_ctrl:
+	if (!OCTEON_CN23XX_PF(oct) || !oct->msix_on)
+		cleanup_tx_poll_fn(netdev);
+err_poll:
+	if (lio->ptp_clock) {
+		ptp_clock_unregister(lio->ptp_clock);
+		lio->ptp_clock = NULL;
+	}
+
+	if (oct->props[lio->ifidx].napi_enabled == 1) {
+		list_for_each_entry_safe(napi, n, &netdev->napi_list, dev_list)
+			napi_disable(napi);
+
+		oct->props[lio->ifidx].napi_enabled = 0;
+
+		if (OCTEON_CN23XX_PF(oct))
+			oct->droq[0]->ops.poll_mode = 0;
+	}
+
+	return ret;
+}
+
+/**
+ * liquidio_stop - Net device stop for LiquidIO
+ * @netdev: network device
+ */
+static int liquidio_stop(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octeon_device_priv *oct_priv =
+		(struct octeon_device_priv *)oct->priv;
+	struct napi_struct *napi, *n;
+	int ret = 0;
+
+	ifstate_reset(lio, LIO_IFSTATE_RUNNING);
+
+	/* Stop any link updates */
+	lio->intf_open = 0;
+
+	stop_txqs(netdev);
+
+	/* Inform that netif carrier is down */
+	netif_carrier_off(netdev);
+	netif_tx_disable(netdev);
+
+	lio->linfo.link.s.link_up = 0;
+	lio->link_changes++;
+
+	/* Tell Octeon that nic interface is down. */
+	ret = send_rx_ctrl_cmd(lio, 0);
+	if (ret)
+		return ret;
+
+	if (OCTEON_CN23XX_PF(oct)) {
+		if (!oct->msix_on)
+			cleanup_tx_poll_fn(netdev);
+	} else {
+		cleanup_tx_poll_fn(netdev);
+	}
+
+	cancel_delayed_work_sync(&lio->stats_wk.work);
+
+	if (lio->ptp_clock) {
+		ptp_clock_unregister(lio->ptp_clock);
+		lio->ptp_clock = NULL;
+	}
+
+	/* Wait for any pending Rx descriptors */
+	if (lio_wait_for_clean_oq(oct))
+		netif_info(lio, rx_err, lio->netdev,
+			   "Proceeding with stop interface after partial RX desc processing\n");
+
+	if (oct->props[lio->ifidx].napi_enabled == 1) {
+		list_for_each_entry_safe(napi, n, &netdev->napi_list, dev_list)
+			napi_disable(napi);
+
+		oct->props[lio->ifidx].napi_enabled = 0;
+
+		if (OCTEON_CN23XX_PF(oct))
+			oct->droq[0]->ops.poll_mode = 0;
+
+		tasklet_enable(&oct_priv->droq_tasklet);
+	}
+
+	dev_info(&oct->pci_dev->dev, "%s interface is stopped\n", netdev->name);
+
+	return ret;
+}
+
+/**
+ * get_new_flags - Converts a mask based on net device flags
+ * @netdev: network device
+ *
+ * This routine generates a octnet_ifflags mask from the net device flags
+ * received from the OS.
+ */
+static inline enum octnet_ifflags get_new_flags(struct net_device *netdev)
+{
+	enum octnet_ifflags f = OCTNET_IFFLAG_UNICAST;
+
+	if (netdev->flags & IFF_PROMISC)
+		f |= OCTNET_IFFLAG_PROMISC;
+
+	if (netdev->flags & IFF_ALLMULTI)
+		f |= OCTNET_IFFLAG_ALLMULTI;
+
+	if (netdev->flags & IFF_MULTICAST) {
+		f |= OCTNET_IFFLAG_MULTICAST;
+
+		/* Accept all multicast addresses if there are more than we
+		 * can handle
+		 */
+		if (netdev_mc_count(netdev) > MAX_OCTEON_MULTICAST_ADDR)
+			f |= OCTNET_IFFLAG_ALLMULTI;
+	}
+
+	if (netdev->flags & IFF_BROADCAST)
+		f |= OCTNET_IFFLAG_BROADCAST;
+
+	return f;
+}
+
+/**
+ * liquidio_set_mcast_list - Net device set_multicast_list
+ * @netdev: network device
+ */
+static void liquidio_set_mcast_list(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	struct netdev_hw_addr *ha;
+	u64 *mc;
+	int ret;
+	int mc_count = min(netdev_mc_count(netdev), MAX_OCTEON_MULTICAST_ADDR);
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	/* Create a ctrl pkt command to be sent to core app. */
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_SET_MULTI_LIST;
+	nctrl.ncmd.s.param1 = get_new_flags(netdev);
+	nctrl.ncmd.s.param2 = mc_count;
+	nctrl.ncmd.s.more = mc_count;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	/* copy all the addresses into the udd */
+	mc = &nctrl.udd[0];
+	netdev_for_each_mc_addr(ha, netdev) {
+		*mc = 0;
+		memcpy(((u8 *)mc) + 2, ha->addr, ETH_ALEN);
+		/* no need to swap bytes */
+
+		if (++mc > &nctrl.udd[mc_count])
+			break;
+	}
+
+	/* Apparently, any activity in this call from the kernel has to
+	 * be atomic. So we won't wait for response.
+	 */
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "DEVFLAGS change failed in core (ret: 0x%x)\n",
+			ret);
+	}
+}
+
+/**
+ * liquidio_set_mac - Net device set_mac_address
+ * @netdev: network device
+ * @p: pointer to sockaddr
+ */
+static int liquidio_set_mac(struct net_device *netdev, void *p)
+{
+	int ret = 0;
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct sockaddr *addr = (struct sockaddr *)p;
+	struct octnic_ctrl_pkt nctrl;
+
+	if (!is_valid_ether_addr(addr->sa_data))
+		return -EADDRNOTAVAIL;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_CHANGE_MACADDR;
+	nctrl.ncmd.s.param1 = 0;
+	nctrl.ncmd.s.more = 1;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+
+	nctrl.udd[0] = 0;
+	/* The MAC Address is presented in network byte order. */
+	memcpy((u8 *)&nctrl.udd[0] + 2, addr->sa_data, ETH_ALEN);
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret < 0) {
+		dev_err(&oct->pci_dev->dev, "MAC Address change failed\n");
+		return -ENOMEM;
+	}
+
+	if (nctrl.sc_status) {
+		dev_err(&oct->pci_dev->dev,
+			"%s: MAC Address change failed. sc return=%x\n",
+			 __func__, nctrl.sc_status);
+		return -EIO;
+	}
+
+	eth_hw_addr_set(netdev, addr->sa_data);
+	memcpy(((u8 *)&lio->linfo.hw_addr) + 2, addr->sa_data, ETH_ALEN);
+
+	return 0;
+}
+
+static void
+liquidio_get_stats64(struct net_device *netdev,
+		     struct rtnl_link_stats64 *lstats)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct;
+	u64 pkts = 0, drop = 0, bytes = 0;
+	struct oct_droq_stats *oq_stats;
+	struct oct_iq_stats *iq_stats;
+	int i, iq_no, oq_no;
+
+	oct = lio->oct_dev;
+
+	if (ifstate_check(lio, LIO_IFSTATE_RESETTING))
+		return;
+
+	for (i = 0; i < oct->num_iqs; i++) {
+		iq_no = lio->linfo.txpciq[i].s.q_no;
+		iq_stats = &oct->instr_queue[iq_no]->stats;
+		pkts += iq_stats->tx_done;
+		drop += iq_stats->tx_dropped;
+		bytes += iq_stats->tx_tot_bytes;
+	}
+
+	lstats->tx_packets = pkts;
+	lstats->tx_bytes = bytes;
+	lstats->tx_dropped = drop;
+
+	pkts = 0;
+	drop = 0;
+	bytes = 0;
+
+	for (i = 0; i < oct->num_oqs; i++) {
+		oq_no = lio->linfo.rxpciq[i].s.q_no;
+		oq_stats = &oct->droq[oq_no]->stats;
+		pkts += oq_stats->rx_pkts_received;
+		drop += (oq_stats->rx_dropped +
+			 oq_stats->dropped_nodispatch +
+			 oq_stats->dropped_toomany +
+			 oq_stats->dropped_nomem);
+		bytes += oq_stats->rx_bytes_received;
+	}
+
+	lstats->rx_bytes = bytes;
+	lstats->rx_packets = pkts;
+	lstats->rx_dropped = drop;
+
+	lstats->multicast = oct->link_stats.fromwire.fw_total_mcast;
+	lstats->collisions = oct->link_stats.fromhost.total_collisions;
+
+	/* detailed rx_errors: */
+	lstats->rx_length_errors = oct->link_stats.fromwire.l2_err;
+	/* recved pkt with crc error    */
+	lstats->rx_crc_errors = oct->link_stats.fromwire.fcs_err;
+	/* recv'd frame alignment error */
+	lstats->rx_frame_errors = oct->link_stats.fromwire.frame_err;
+	/* recv'r fifo overrun */
+	lstats->rx_fifo_errors = oct->link_stats.fromwire.fifo_err;
+
+	lstats->rx_errors = lstats->rx_length_errors + lstats->rx_crc_errors +
+		lstats->rx_frame_errors + lstats->rx_fifo_errors;
+
+	/* detailed tx_errors */
+	lstats->tx_aborted_errors = oct->link_stats.fromhost.fw_err_pko;
+	lstats->tx_carrier_errors = oct->link_stats.fromhost.fw_err_link;
+	lstats->tx_fifo_errors = oct->link_stats.fromhost.fifo_err;
+
+	lstats->tx_errors = lstats->tx_aborted_errors +
+		lstats->tx_carrier_errors +
+		lstats->tx_fifo_errors;
+}
+
+/**
+ * hwtstamp_ioctl - Handler for SIOCSHWTSTAMP ioctl
+ * @netdev: network device
+ * @ifr: interface request
+ */
+static int hwtstamp_ioctl(struct net_device *netdev, struct ifreq *ifr)
+{
+	struct hwtstamp_config conf;
+	struct lio *lio = GET_LIO(netdev);
+
+	if (copy_from_user(&conf, ifr->ifr_data, sizeof(conf)))
+		return -EFAULT;
+
+	switch (conf.tx_type) {
+	case HWTSTAMP_TX_ON:
+	case HWTSTAMP_TX_OFF:
+		break;
+	default:
+		return -ERANGE;
+	}
+
+	switch (conf.rx_filter) {
+	case HWTSTAMP_FILTER_NONE:
+		break;
+	case HWTSTAMP_FILTER_ALL:
+	case HWTSTAMP_FILTER_SOME:
+	case HWTSTAMP_FILTER_PTP_V1_L4_EVENT:
+	case HWTSTAMP_FILTER_PTP_V1_L4_SYNC:
+	case HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_L4_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_L4_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_L2_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_L2_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_L2_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_DELAY_REQ:
+	case HWTSTAMP_FILTER_NTP_ALL:
+		conf.rx_filter = HWTSTAMP_FILTER_ALL;
+		break;
+	default:
+		return -ERANGE;
+	}
+
+	if (conf.rx_filter == HWTSTAMP_FILTER_ALL)
+		ifstate_set(lio, LIO_IFSTATE_RX_TIMESTAMP_ENABLED);
+
+	else
+		ifstate_reset(lio, LIO_IFSTATE_RX_TIMESTAMP_ENABLED);
+
+	return copy_to_user(ifr->ifr_data, &conf, sizeof(conf)) ? -EFAULT : 0;
+}
+
+/**
+ * liquidio_ioctl - ioctl handler
+ * @netdev: network device
+ * @ifr: interface request
+ * @cmd: command
+ */
+static int liquidio_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd)
+{
+	struct lio *lio = GET_LIO(netdev);
+
+	switch (cmd) {
+	case SIOCSHWTSTAMP:
+		if (lio->oct_dev->ptp_enable)
+			return hwtstamp_ioctl(netdev, ifr);
+		fallthrough;
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+/**
+ * handle_timestamp - handle a Tx timestamp response
+ * @oct: octeon device
+ * @status: response status
+ * @buf: pointer to skb
+ */
+static void handle_timestamp(struct octeon_device *oct,
+			     u32 status,
+			     void *buf)
+{
+	struct octnet_buf_free_info *finfo;
+	struct octeon_soft_command *sc;
+	struct oct_timestamp_resp *resp;
+	struct lio *lio;
+	struct sk_buff *skb = (struct sk_buff *)buf;
+
+	finfo = (struct octnet_buf_free_info *)skb->cb;
+	lio = finfo->lio;
+	sc = finfo->sc;
+	oct = lio->oct_dev;
+	resp = (struct oct_timestamp_resp *)sc->virtrptr;
+
+	if (status != OCTEON_REQUEST_DONE) {
+		dev_err(&oct->pci_dev->dev, "Tx timestamp instruction failed. Status: %llx\n",
+			CVM_CAST64(status));
+		resp->timestamp = 0;
+	}
+
+	octeon_swap_8B_data(&resp->timestamp, 1);
+
+	if (unlikely((skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS) != 0)) {
+		struct skb_shared_hwtstamps ts;
+		u64 ns = resp->timestamp;
+
+		netif_info(lio, tx_done, lio->netdev,
+			   "Got resulting SKBTX_HW_TSTAMP skb=%p ns=%016llu\n",
+			   skb, (unsigned long long)ns);
+		ts.hwtstamp = ns_to_ktime(ns + lio->ptp_adjust);
+		skb_tstamp_tx(skb, &ts);
+	}
+
+	octeon_free_soft_command(oct, sc);
+	tx_buffer_free(skb);
+}
+
+/**
+ * send_nic_timestamp_pkt - Send a data packet that will be timestamped
+ * @oct: octeon device
+ * @ndata: pointer to network data
+ * @finfo: pointer to private network data
+ * @xmit_more: more is coming
+ */
+static inline int send_nic_timestamp_pkt(struct octeon_device *oct,
+					 struct octnic_data_pkt *ndata,
+					 struct octnet_buf_free_info *finfo,
+					 int xmit_more)
+{
+	int retval;
+	struct octeon_soft_command *sc;
+	struct lio *lio;
+	int ring_doorbell;
+	u32 len;
+
+	lio = finfo->lio;
+
+	sc = octeon_alloc_soft_command_resp(oct, &ndata->cmd,
+					    sizeof(struct oct_timestamp_resp));
+	finfo->sc = sc;
+
+	if (!sc) {
+		dev_err(&oct->pci_dev->dev, "No memory for timestamped data packet\n");
+		return IQ_SEND_FAILED;
+	}
+
+	if (ndata->reqtype == REQTYPE_NORESP_NET)
+		ndata->reqtype = REQTYPE_RESP_NET;
+	else if (ndata->reqtype == REQTYPE_NORESP_NET_SG)
+		ndata->reqtype = REQTYPE_RESP_NET_SG;
+
+	sc->callback = handle_timestamp;
+	sc->callback_arg = finfo->skb;
+	sc->iq_no = ndata->q_no;
+
+	if (OCTEON_CN23XX_PF(oct))
+		len = (u32)((struct octeon_instr_ih3 *)
+			    (&sc->cmd.cmd3.ih3))->dlengsz;
+	else
+		len = (u32)((struct octeon_instr_ih2 *)
+			    (&sc->cmd.cmd2.ih2))->dlengsz;
+
+	ring_doorbell = !xmit_more;
+
+	retval = octeon_send_command(oct, sc->iq_no, ring_doorbell, &sc->cmd,
+				     sc, len, ndata->reqtype);
+
+	if (retval == IQ_SEND_FAILED) {
+		dev_err(&oct->pci_dev->dev, "timestamp data packet failed status: %x\n",
+			retval);
+		octeon_free_soft_command(oct, sc);
+	} else {
+		netif_info(lio, tx_queued, lio->netdev, "Queued timestamp packet\n");
+	}
+
+	return retval;
+}
+
+/**
+ * liquidio_xmit - Transmit networks packets to the Octeon interface
+ * @skb: skbuff struct to be passed to network layer.
+ * @netdev: pointer to network device
+ *
+ * Return: whether the packet was transmitted to the device okay or not
+ *             (NETDEV_TX_OK or NETDEV_TX_BUSY)
+ */
+static netdev_tx_t liquidio_xmit(struct sk_buff *skb, struct net_device *netdev)
+{
+	struct lio *lio;
+	struct octnet_buf_free_info *finfo;
+	union octnic_cmd_setup cmdsetup;
+	struct octnic_data_pkt ndata;
+	struct octeon_device *oct;
+	struct oct_iq_stats *stats;
+	struct octeon_instr_irh *irh;
+	union tx_info *tx_info;
+	int status = 0;
+	int q_idx = 0, iq_no = 0;
+	int j, xmit_more = 0;
+	u64 dptr = 0;
+	u32 tag = 0;
+
+	lio = GET_LIO(netdev);
+	oct = lio->oct_dev;
+
+	q_idx = skb_iq(oct, skb);
+	tag = q_idx;
+	iq_no = lio->linfo.txpciq[q_idx].s.q_no;
+
+	stats = &oct->instr_queue[iq_no]->stats;
+
+	/* Check for all conditions in which the current packet cannot be
+	 * transmitted.
+	 */
+	if (!(atomic_read(&lio->ifstate) & LIO_IFSTATE_RUNNING) ||
+	    (!lio->linfo.link.s.link_up) ||
+	    (skb->len <= 0)) {
+		netif_info(lio, tx_err, lio->netdev,
+			   "Transmit failed link_status : %d\n",
+			   lio->linfo.link.s.link_up);
+		goto lio_xmit_failed;
+	}
+
+	/* Use space in skb->cb to store info used to unmap and
+	 * free the buffers.
+	 */
+	finfo = (struct octnet_buf_free_info *)skb->cb;
+	finfo->lio = lio;
+	finfo->skb = skb;
+	finfo->sc = NULL;
+
+	/* Prepare the attributes for the data to be passed to OSI. */
+	memset(&ndata, 0, sizeof(struct octnic_data_pkt));
+
+	ndata.buf = (void *)finfo;
+
+	ndata.q_no = iq_no;
+
+	if (octnet_iq_is_full(oct, ndata.q_no)) {
+		/* defer sending if queue is full */
+		netif_info(lio, tx_err, lio->netdev, "Transmit failed iq:%d full\n",
+			   ndata.q_no);
+		stats->tx_iq_busy++;
+		return NETDEV_TX_BUSY;
+	}
+
+	/* pr_info(" XMIT - valid Qs: %d, 1st Q no: %d, cpu:  %d, q_no:%d\n",
+	 *	lio->linfo.num_txpciq, lio->txq, cpu, ndata.q_no);
+	 */
+
+	ndata.datasize = skb->len;
+
+	cmdsetup.u64 = 0;
+	cmdsetup.s.iq_no = iq_no;
+
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
+		if (skb->encapsulation) {
+			cmdsetup.s.tnl_csum = 1;
+			stats->tx_vxlan++;
+		} else {
+			cmdsetup.s.transport_csum = 1;
+		}
+	}
+	if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)) {
+		skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
+		cmdsetup.s.timestamp = 1;
+	}
+
+	if (skb_shinfo(skb)->nr_frags == 0) {
+		cmdsetup.s.u.datasize = skb->len;
+		octnet_prepare_pci_cmd(oct, &ndata.cmd, &cmdsetup, tag);
+
+		/* Offload checksum calculation for TCP/UDP packets */
+		dptr = dma_map_single(&oct->pci_dev->dev,
+				      skb->data,
+				      skb->len,
+				      DMA_TO_DEVICE);
+		if (dma_mapping_error(&oct->pci_dev->dev, dptr)) {
+			dev_err(&oct->pci_dev->dev, "%s DMA mapping error 1\n",
+				__func__);
+			stats->tx_dmamap_fail++;
+			return NETDEV_TX_BUSY;
+		}
+
+		if (OCTEON_CN23XX_PF(oct))
+			ndata.cmd.cmd3.dptr = dptr;
+		else
+			ndata.cmd.cmd2.dptr = dptr;
+		finfo->dptr = dptr;
+		ndata.reqtype = REQTYPE_NORESP_NET;
+
+	} else {
+		int i, frags;
+		skb_frag_t *frag;
+		struct octnic_gather *g;
+
+		spin_lock(&lio->glist_lock[q_idx]);
+		g = (struct octnic_gather *)
+			lio_list_delete_head(&lio->glist[q_idx]);
+		spin_unlock(&lio->glist_lock[q_idx]);
+
+		if (!g) {
+			netif_info(lio, tx_err, lio->netdev,
+				   "Transmit scatter gather: glist null!\n");
+			goto lio_xmit_failed;
+		}
+
+		cmdsetup.s.gather = 1;
+		cmdsetup.s.u.gatherptrs = (skb_shinfo(skb)->nr_frags + 1);
+		octnet_prepare_pci_cmd(oct, &ndata.cmd, &cmdsetup, tag);
+
+		memset(g->sg, 0, g->sg_size);
+
+		g->sg[0].ptr[0] = dma_map_single(&oct->pci_dev->dev,
+						 skb->data,
+						 (skb->len - skb->data_len),
+						 DMA_TO_DEVICE);
+		if (dma_mapping_error(&oct->pci_dev->dev, g->sg[0].ptr[0])) {
+			dev_err(&oct->pci_dev->dev, "%s DMA mapping error 2\n",
+				__func__);
+			stats->tx_dmamap_fail++;
+			return NETDEV_TX_BUSY;
+		}
+		add_sg_size(&g->sg[0], (skb->len - skb->data_len), 0);
+
+		frags = skb_shinfo(skb)->nr_frags;
+		i = 1;
+		while (frags--) {
+			frag = &skb_shinfo(skb)->frags[i - 1];
+
+			g->sg[(i >> 2)].ptr[(i & 3)] =
+				skb_frag_dma_map(&oct->pci_dev->dev,
+					         frag, 0, skb_frag_size(frag),
+						 DMA_TO_DEVICE);
+
+			if (dma_mapping_error(&oct->pci_dev->dev,
+					      g->sg[i >> 2].ptr[i & 3])) {
+				dma_unmap_single(&oct->pci_dev->dev,
+						 g->sg[0].ptr[0],
+						 skb->len - skb->data_len,
+						 DMA_TO_DEVICE);
+				for (j = 1; j < i; j++) {
+					frag = &skb_shinfo(skb)->frags[j - 1];
+					dma_unmap_page(&oct->pci_dev->dev,
+						       g->sg[j >> 2].ptr[j & 3],
+						       skb_frag_size(frag),
+						       DMA_TO_DEVICE);
+				}
+				dev_err(&oct->pci_dev->dev, "%s DMA mapping error 3\n",
+					__func__);
+				return NETDEV_TX_BUSY;
+			}
+
+			add_sg_size(&g->sg[(i >> 2)], skb_frag_size(frag),
+				    (i & 3));
+			i++;
+		}
+
+		dptr = g->sg_dma_ptr;
+
+		if (OCTEON_CN23XX_PF(oct))
+			ndata.cmd.cmd3.dptr = dptr;
+		else
+			ndata.cmd.cmd2.dptr = dptr;
+		finfo->dptr = dptr;
+		finfo->g = g;
+
+		ndata.reqtype = REQTYPE_NORESP_NET_SG;
+	}
+
+	if (OCTEON_CN23XX_PF(oct)) {
+		irh = (struct octeon_instr_irh *)&ndata.cmd.cmd3.irh;
+		tx_info = (union tx_info *)&ndata.cmd.cmd3.ossp[0];
+	} else {
+		irh = (struct octeon_instr_irh *)&ndata.cmd.cmd2.irh;
+		tx_info = (union tx_info *)&ndata.cmd.cmd2.ossp[0];
+	}
+
+	if (skb_shinfo(skb)->gso_size) {
+		tx_info->s.gso_size = skb_shinfo(skb)->gso_size;
+		tx_info->s.gso_segs = skb_shinfo(skb)->gso_segs;
+		stats->tx_gso++;
+	}
+
+	/* HW insert VLAN tag */
+	if (skb_vlan_tag_present(skb)) {
+		irh->priority = skb_vlan_tag_get(skb) >> 13;
+		irh->vlan = skb_vlan_tag_get(skb) & 0xfff;
+	}
+
+	xmit_more = netdev_xmit_more();
+
+	if (unlikely(cmdsetup.s.timestamp))
+		status = send_nic_timestamp_pkt(oct, &ndata, finfo, xmit_more);
+	else
+		status = octnet_send_nic_data_pkt(oct, &ndata, xmit_more);
+	if (status == IQ_SEND_FAILED)
+		goto lio_xmit_failed;
+
+	netif_info(lio, tx_queued, lio->netdev, "Transmit queued successfully\n");
+
+	if (status == IQ_SEND_STOP)
+		netif_stop_subqueue(netdev, q_idx);
+
+	netif_trans_update(netdev);
+
+	if (tx_info->s.gso_segs)
+		stats->tx_done += tx_info->s.gso_segs;
+	else
+		stats->tx_done++;
+	stats->tx_tot_bytes += ndata.datasize;
+
+	return NETDEV_TX_OK;
+
+lio_xmit_failed:
+	stats->tx_dropped++;
+	netif_info(lio, tx_err, lio->netdev, "IQ%d Transmit dropped:%llu\n",
+		   iq_no, stats->tx_dropped);
+	if (dptr)
+		dma_unmap_single(&oct->pci_dev->dev, dptr,
+				 ndata.datasize, DMA_TO_DEVICE);
+
+	octeon_ring_doorbell_locked(oct, iq_no);
+
+	tx_buffer_free(skb);
+	return NETDEV_TX_OK;
+}
+
+/**
+ * liquidio_tx_timeout - Network device Tx timeout
+ * @netdev:    pointer to network device
+ * @txqueue: index of the hung transmit queue
+ */
+static void liquidio_tx_timeout(struct net_device *netdev, unsigned int txqueue)
+{
+	struct lio *lio;
+
+	lio = GET_LIO(netdev);
+
+	netif_info(lio, tx_err, lio->netdev,
+		   "Transmit timeout tx_dropped:%ld, waking up queues now!!\n",
+		   netdev->stats.tx_dropped);
+	netif_trans_update(netdev);
+	wake_txqs(netdev);
+}
+
+static int liquidio_vlan_rx_add_vid(struct net_device *netdev,
+				    __be16 proto __attribute__((unused)),
+				    u16 vid)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_ADD_VLAN_FILTER;
+	nctrl.ncmd.s.param1 = vid;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "Add VLAN filter failed in core (ret: 0x%x)\n",
+			ret);
+		if (ret > 0)
+			ret = -EIO;
+	}
+
+	return ret;
+}
+
+static int liquidio_vlan_rx_kill_vid(struct net_device *netdev,
+				     __be16 proto __attribute__((unused)),
+				     u16 vid)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_DEL_VLAN_FILTER;
+	nctrl.ncmd.s.param1 = vid;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "Del VLAN filter failed in core (ret: 0x%x)\n",
+			ret);
+		if (ret > 0)
+			ret = -EIO;
+	}
+	return ret;
+}
+
+/**
+ * liquidio_set_rxcsum_command - Sending command to enable/disable RX checksum offload
+ * @netdev:                pointer to network device
+ * @command:               OCTNET_CMD_TNL_RX_CSUM_CTL
+ * @rx_cmd:                OCTNET_CMD_RXCSUM_ENABLE/OCTNET_CMD_RXCSUM_DISABLE
+ * Returns:                SUCCESS or FAILURE
+ */
+static int liquidio_set_rxcsum_command(struct net_device *netdev, int command,
+				       u8 rx_cmd)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = command;
+	nctrl.ncmd.s.param1 = rx_cmd;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev,
+			"DEVFLAGS RXCSUM change failed in core(ret:0x%x)\n",
+			ret);
+		if (ret > 0)
+			ret = -EIO;
+	}
+	return ret;
+}
+
+/**
+ * liquidio_vxlan_port_command - Sending command to add/delete VxLAN UDP port to firmware
+ * @netdev:                pointer to network device
+ * @command:               OCTNET_CMD_VXLAN_PORT_CONFIG
+ * @vxlan_port:            VxLAN port to be added or deleted
+ * @vxlan_cmd_bit:         OCTNET_CMD_VXLAN_PORT_ADD,
+ *                              OCTNET_CMD_VXLAN_PORT_DEL
+ * Return:                     SUCCESS or FAILURE
+ */
+static int liquidio_vxlan_port_command(struct net_device *netdev, int command,
+				       u16 vxlan_port, u8 vxlan_cmd_bit)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = command;
+	nctrl.ncmd.s.more = vxlan_cmd_bit;
+	nctrl.ncmd.s.param1 = vxlan_port;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev,
+			"VxLAN port add/delete failed in core (ret:0x%x)\n",
+			ret);
+		if (ret > 0)
+			ret = -EIO;
+	}
+	return ret;
+}
+
+static int liquidio_udp_tunnel_set_port(struct net_device *netdev,
+					unsigned int table, unsigned int entry,
+					struct udp_tunnel_info *ti)
+{
+	return liquidio_vxlan_port_command(netdev,
+					   OCTNET_CMD_VXLAN_PORT_CONFIG,
+					   htons(ti->port),
+					   OCTNET_CMD_VXLAN_PORT_ADD);
+}
+
+static int liquidio_udp_tunnel_unset_port(struct net_device *netdev,
+					  unsigned int table,
+					  unsigned int entry,
+					  struct udp_tunnel_info *ti)
+{
+	return liquidio_vxlan_port_command(netdev,
+					   OCTNET_CMD_VXLAN_PORT_CONFIG,
+					   htons(ti->port),
+					   OCTNET_CMD_VXLAN_PORT_DEL);
+}
+
+static const struct udp_tunnel_nic_info liquidio_udp_tunnels = {
+	.set_port	= liquidio_udp_tunnel_set_port,
+	.unset_port	= liquidio_udp_tunnel_unset_port,
+	.tables		= {
+		{ .n_entries = 1024, .tunnel_types = UDP_TUNNEL_TYPE_VXLAN, },
+	},
+};
+
+/**
+ * liquidio_fix_features - Net device fix features
+ * @netdev:  pointer to network device
+ * @request: features requested
+ * Return: updated features list
+ */
+static netdev_features_t liquidio_fix_features(struct net_device *netdev,
+					       netdev_features_t request)
+{
+	struct lio *lio = netdev_priv(netdev);
+
+	if ((request & NETIF_F_RXCSUM) &&
+	    !(lio->dev_capability & NETIF_F_RXCSUM))
+		request &= ~NETIF_F_RXCSUM;
+
+	if ((request & NETIF_F_HW_CSUM) &&
+	    !(lio->dev_capability & NETIF_F_HW_CSUM))
+		request &= ~NETIF_F_HW_CSUM;
+
+	if ((request & NETIF_F_TSO) && !(lio->dev_capability & NETIF_F_TSO))
+		request &= ~NETIF_F_TSO;
+
+	if ((request & NETIF_F_TSO6) && !(lio->dev_capability & NETIF_F_TSO6))
+		request &= ~NETIF_F_TSO6;
+
+	if ((request & NETIF_F_LRO) && !(lio->dev_capability & NETIF_F_LRO))
+		request &= ~NETIF_F_LRO;
+
+	/*Disable LRO if RXCSUM is off */
+	if (!(request & NETIF_F_RXCSUM) && (netdev->features & NETIF_F_LRO) &&
+	    (lio->dev_capability & NETIF_F_LRO))
+		request &= ~NETIF_F_LRO;
+
+	if ((request & NETIF_F_HW_VLAN_CTAG_FILTER) &&
+	    !(lio->dev_capability & NETIF_F_HW_VLAN_CTAG_FILTER))
+		request &= ~NETIF_F_HW_VLAN_CTAG_FILTER;
+
+	return request;
+}
+
+/**
+ * liquidio_set_features - Net device set features
+ * @netdev:  pointer to network device
+ * @features: features to enable/disable
+ */
+static int liquidio_set_features(struct net_device *netdev,
+				 netdev_features_t features)
+{
+	struct lio *lio = netdev_priv(netdev);
+
+	if ((features & NETIF_F_LRO) &&
+	    (lio->dev_capability & NETIF_F_LRO) &&
+	    !(netdev->features & NETIF_F_LRO))
+		liquidio_set_feature(netdev, OCTNET_CMD_LRO_ENABLE,
+				     OCTNIC_LROIPV4 | OCTNIC_LROIPV6);
+	else if (!(features & NETIF_F_LRO) &&
+		 (lio->dev_capability & NETIF_F_LRO) &&
+		 (netdev->features & NETIF_F_LRO))
+		liquidio_set_feature(netdev, OCTNET_CMD_LRO_DISABLE,
+				     OCTNIC_LROIPV4 | OCTNIC_LROIPV6);
+
+	/* Sending command to firmware to enable/disable RX checksum
+	 * offload settings using ethtool
+	 */
+	if (!(netdev->features & NETIF_F_RXCSUM) &&
+	    (lio->enc_dev_capability & NETIF_F_RXCSUM) &&
+	    (features & NETIF_F_RXCSUM))
+		liquidio_set_rxcsum_command(netdev,
+					    OCTNET_CMD_TNL_RX_CSUM_CTL,
+					    OCTNET_CMD_RXCSUM_ENABLE);
+	else if ((netdev->features & NETIF_F_RXCSUM) &&
+		 (lio->enc_dev_capability & NETIF_F_RXCSUM) &&
+		 !(features & NETIF_F_RXCSUM))
+		liquidio_set_rxcsum_command(netdev, OCTNET_CMD_TNL_RX_CSUM_CTL,
+					    OCTNET_CMD_RXCSUM_DISABLE);
+
+	if ((features & NETIF_F_HW_VLAN_CTAG_FILTER) &&
+	    (lio->dev_capability & NETIF_F_HW_VLAN_CTAG_FILTER) &&
+	    !(netdev->features & NETIF_F_HW_VLAN_CTAG_FILTER))
+		liquidio_set_feature(netdev, OCTNET_CMD_VLAN_FILTER_CTL,
+				     OCTNET_CMD_VLAN_FILTER_ENABLE);
+	else if (!(features & NETIF_F_HW_VLAN_CTAG_FILTER) &&
+		 (lio->dev_capability & NETIF_F_HW_VLAN_CTAG_FILTER) &&
+		 (netdev->features & NETIF_F_HW_VLAN_CTAG_FILTER))
+		liquidio_set_feature(netdev, OCTNET_CMD_VLAN_FILTER_CTL,
+				     OCTNET_CMD_VLAN_FILTER_DISABLE);
+
+	return 0;
+}
+
+static int __liquidio_set_vf_mac(struct net_device *netdev, int vfidx,
+				 u8 *mac, bool is_admin_assigned)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	if (!is_valid_ether_addr(mac))
+		return -EINVAL;
+
+	if (vfidx < 0 || vfidx >= oct->sriov_info.max_vfs)
+		return -EINVAL;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_CHANGE_MACADDR;
+	/* vfidx is 0 based, but vf_num (param1) is 1 based */
+	nctrl.ncmd.s.param1 = vfidx + 1;
+	nctrl.ncmd.s.more = 1;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	if (is_admin_assigned) {
+		nctrl.ncmd.s.param2 = true;
+		nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+	}
+
+	nctrl.udd[0] = 0;
+	/* The MAC Address is presented in network byte order. */
+	ether_addr_copy((u8 *)&nctrl.udd[0] + 2, mac);
+
+	oct->sriov_info.vf_macaddr[vfidx] = nctrl.udd[0];
+
+	ret = octnet_send_nic_ctrl_pkt(oct, &nctrl);
+	if (ret > 0)
+		ret = -EIO;
+
+	return ret;
+}
+
+static int liquidio_set_vf_mac(struct net_device *netdev, int vfidx, u8 *mac)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	int retval;
+
+	if (vfidx < 0 || vfidx >= oct->sriov_info.num_vfs_alloced)
+		return -EINVAL;
+
+	retval = __liquidio_set_vf_mac(netdev, vfidx, mac, true);
+	if (!retval)
+		cn23xx_tell_vf_its_macaddr_changed(oct, vfidx, mac);
+
+	return retval;
+}
+
+static int liquidio_set_vf_spoofchk(struct net_device *netdev, int vfidx,
+				    bool enable)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int retval;
+
+	if (!(oct->fw_info.app_cap_flags & LIQUIDIO_SPOOFCHK_CAP)) {
+		netif_info(lio, drv, lio->netdev,
+			   "firmware does not support spoofchk\n");
+		return -EOPNOTSUPP;
+	}
+
+	if (vfidx < 0 || vfidx >= oct->sriov_info.num_vfs_alloced) {
+		netif_info(lio, drv, lio->netdev, "Invalid vfidx %d\n", vfidx);
+		return -EINVAL;
+	}
+
+	if (enable) {
+		if (oct->sriov_info.vf_spoofchk[vfidx])
+			return 0;
+	} else {
+		/* Clear */
+		if (!oct->sriov_info.vf_spoofchk[vfidx])
+			return 0;
+	}
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+	nctrl.ncmd.s.cmdgroup = OCTNET_CMD_GROUP1;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_SET_VF_SPOOFCHK;
+	nctrl.ncmd.s.param1 =
+		vfidx + 1; /* vfidx is 0 based,
+			    * but vf_num (param1) is 1 based
+			    */
+	nctrl.ncmd.s.param2 = enable;
+	nctrl.ncmd.s.more = 0;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.cb_fn = NULL;
+
+	retval = octnet_send_nic_ctrl_pkt(oct, &nctrl);
+
+	if (retval) {
+		netif_info(lio, drv, lio->netdev,
+			   "Failed to set VF %d spoofchk %s\n", vfidx,
+			enable ? "on" : "off");
+		return -1;
+	}
+
+	oct->sriov_info.vf_spoofchk[vfidx] = enable;
+	netif_info(lio, drv, lio->netdev, "VF %u spoofchk is %s\n", vfidx,
+		   enable ? "on" : "off");
+
+	return 0;
+}
+
+static int liquidio_set_vf_vlan(struct net_device *netdev, int vfidx,
+				u16 vlan, u8 qos, __be16 vlan_proto)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	u16 vlantci;
+	int ret = 0;
+
+	if (vfidx < 0 || vfidx >= oct->sriov_info.num_vfs_alloced)
+		return -EINVAL;
+
+	if (vlan_proto != htons(ETH_P_8021Q))
+		return -EPROTONOSUPPORT;
+
+	if (vlan >= VLAN_N_VID || qos > 7)
+		return -EINVAL;
+
+	if (vlan)
+		vlantci = vlan | (u16)qos << VLAN_PRIO_SHIFT;
+	else
+		vlantci = 0;
+
+	if (oct->sriov_info.vf_vlantci[vfidx] == vlantci)
+		return 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	if (vlan)
+		nctrl.ncmd.s.cmd = OCTNET_CMD_ADD_VLAN_FILTER;
+	else
+		nctrl.ncmd.s.cmd = OCTNET_CMD_DEL_VLAN_FILTER;
+
+	nctrl.ncmd.s.param1 = vlantci;
+	nctrl.ncmd.s.param2 =
+	    vfidx + 1; /* vfidx is 0 based, but vf_num (param2) is 1 based */
+	nctrl.ncmd.s.more = 0;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.cb_fn = NULL;
+
+	ret = octnet_send_nic_ctrl_pkt(oct, &nctrl);
+	if (ret) {
+		if (ret > 0)
+			ret = -EIO;
+		return ret;
+	}
+
+	oct->sriov_info.vf_vlantci[vfidx] = vlantci;
+
+	return ret;
+}
+
+static int liquidio_get_vf_config(struct net_device *netdev, int vfidx,
+				  struct ifla_vf_info *ivi)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	u8 *macaddr;
+
+	if (vfidx < 0 || vfidx >= oct->sriov_info.num_vfs_alloced)
+		return -EINVAL;
+
+	memset(ivi, 0, sizeof(struct ifla_vf_info));
+
+	ivi->vf = vfidx;
+	macaddr = 2 + (u8 *)&oct->sriov_info.vf_macaddr[vfidx];
+	ether_addr_copy(&ivi->mac[0], macaddr);
+	ivi->vlan = oct->sriov_info.vf_vlantci[vfidx] & VLAN_VID_MASK;
+	ivi->qos = oct->sriov_info.vf_vlantci[vfidx] >> VLAN_PRIO_SHIFT;
+	if (oct->sriov_info.trusted_vf.active &&
+	    oct->sriov_info.trusted_vf.id == vfidx)
+		ivi->trusted = true;
+	else
+		ivi->trusted = false;
+	ivi->linkstate = oct->sriov_info.vf_linkstate[vfidx];
+	ivi->spoofchk = oct->sriov_info.vf_spoofchk[vfidx];
+	ivi->max_tx_rate = lio->linfo.link.s.speed;
+	ivi->min_tx_rate = 0;
+
+	return 0;
+}
+
+static int liquidio_send_vf_trust_cmd(struct lio *lio, int vfidx, bool trusted)
+{
+	struct octeon_device *oct = lio->oct_dev;
+	struct octeon_soft_command *sc;
+	int retval;
+
+	sc = octeon_alloc_soft_command(oct, 0, 16, 0);
+	if (!sc)
+		return -ENOMEM;
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	/* vfidx is 0 based, but vf_num (param1) is 1 based */
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC,
+				    OPCODE_NIC_SET_TRUSTED_VF, 0, vfidx + 1,
+				    trusted);
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	retval = octeon_send_soft_command(oct, sc);
+	if (retval == IQ_SEND_FAILED) {
+		octeon_free_soft_command(oct, sc);
+		retval = -1;
+	} else {
+		/* Wait for response or timeout */
+		retval = wait_for_sc_completion_timeout(oct, sc, 0);
+		if (retval)
+			return (retval);
+
+		WRITE_ONCE(sc->caller_is_done, true);
+	}
+
+	return retval;
+}
+
+static int liquidio_set_vf_trust(struct net_device *netdev, int vfidx,
+				 bool setting)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+
+	if (strcmp(oct->fw_info.liquidio_firmware_version, "1.7.1") < 0) {
+		/* trusted vf is not supported by firmware older than 1.7.1 */
+		return -EOPNOTSUPP;
+	}
+
+	if (vfidx < 0 || vfidx >= oct->sriov_info.num_vfs_alloced) {
+		netif_info(lio, drv, lio->netdev, "Invalid vfidx %d\n", vfidx);
+		return -EINVAL;
+	}
+
+	if (setting) {
+		/* Set */
+
+		if (oct->sriov_info.trusted_vf.active &&
+		    oct->sriov_info.trusted_vf.id == vfidx)
+			return 0;
+
+		if (oct->sriov_info.trusted_vf.active) {
+			netif_info(lio, drv, lio->netdev, "More than one trusted VF is not allowed\n");
+			return -EPERM;
+		}
+	} else {
+		/* Clear */
+
+		if (!oct->sriov_info.trusted_vf.active)
+			return 0;
+	}
+
+	if (!liquidio_send_vf_trust_cmd(lio, vfidx, setting)) {
+		if (setting) {
+			oct->sriov_info.trusted_vf.id = vfidx;
+			oct->sriov_info.trusted_vf.active = true;
+		} else {
+			oct->sriov_info.trusted_vf.active = false;
+		}
+
+		netif_info(lio, drv, lio->netdev, "VF %u is %strusted\n", vfidx,
+			   setting ? "" : "not ");
+	} else {
+		netif_info(lio, drv, lio->netdev, "Failed to set VF trusted\n");
+		return -1;
+	}
+
+	return 0;
+}
+
+static int liquidio_set_vf_link_state(struct net_device *netdev, int vfidx,
+				      int linkstate)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	if (vfidx < 0 || vfidx >= oct->sriov_info.num_vfs_alloced)
+		return -EINVAL;
+
+	if (oct->sriov_info.vf_linkstate[vfidx] == linkstate)
+		return 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+	nctrl.ncmd.s.cmd = OCTNET_CMD_SET_VF_LINKSTATE;
+	nctrl.ncmd.s.param1 =
+	    vfidx + 1; /* vfidx is 0 based, but vf_num (param1) is 1 based */
+	nctrl.ncmd.s.param2 = linkstate;
+	nctrl.ncmd.s.more = 0;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.cb_fn = NULL;
+
+	ret = octnet_send_nic_ctrl_pkt(oct, &nctrl);
+
+	if (!ret)
+		oct->sriov_info.vf_linkstate[vfidx] = linkstate;
+	else if (ret > 0)
+		ret = -EIO;
+
+	return ret;
+}
+
+static int
+liquidio_eswitch_mode_get(struct devlink *devlink, u16 *mode)
+{
+	struct lio_devlink_priv *priv;
+	struct octeon_device *oct;
+
+	priv = devlink_priv(devlink);
+	oct = priv->oct;
+
+	*mode = oct->eswitch_mode;
+
+	return 0;
+}
+
+static int
+liquidio_eswitch_mode_set(struct devlink *devlink, u16 mode,
+			  struct netlink_ext_ack *extack)
+{
+	struct lio_devlink_priv *priv;
+	struct octeon_device *oct;
+	int ret = 0;
+
+	priv = devlink_priv(devlink);
+	oct = priv->oct;
+
+	if (!(oct->fw_info.app_cap_flags & LIQUIDIO_SWITCHDEV_CAP))
+		return -EINVAL;
+
+	if (oct->eswitch_mode == mode)
+		return 0;
+
+	switch (mode) {
+	case DEVLINK_ESWITCH_MODE_SWITCHDEV:
+		oct->eswitch_mode = mode;
+		ret = lio_vf_rep_create(oct);
+		break;
+
+	case DEVLINK_ESWITCH_MODE_LEGACY:
+		lio_vf_rep_destroy(oct);
+		oct->eswitch_mode = mode;
+		break;
+
+	default:
+		ret = -EINVAL;
+	}
+
+	return ret;
+}
+
+static const struct devlink_ops liquidio_devlink_ops = {
+	.eswitch_mode_get = liquidio_eswitch_mode_get,
+	.eswitch_mode_set = liquidio_eswitch_mode_set,
+};
+
+static int
+liquidio_get_port_parent_id(struct net_device *dev,
+			    struct netdev_phys_item_id *ppid)
+{
+	struct lio *lio = GET_LIO(dev);
+	struct octeon_device *oct = lio->oct_dev;
+
+	if (oct->eswitch_mode != DEVLINK_ESWITCH_MODE_SWITCHDEV)
+		return -EOPNOTSUPP;
+
+	ppid->id_len = ETH_ALEN;
+	ether_addr_copy(ppid->id, (void *)&lio->linfo.hw_addr + 2);
+
+	return 0;
+}
+
+static int liquidio_get_vf_stats(struct net_device *netdev, int vfidx,
+				 struct ifla_vf_stats *vf_stats)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct oct_vf_stats stats;
+	int ret;
+
+	if (vfidx < 0 || vfidx >= oct->sriov_info.num_vfs_alloced)
+		return -EINVAL;
+
+	memset(&stats, 0, sizeof(struct oct_vf_stats));
+	ret = cn23xx_get_vf_stats(oct, vfidx, &stats);
+	if (!ret) {
+		vf_stats->rx_packets = stats.rx_packets;
+		vf_stats->tx_packets = stats.tx_packets;
+		vf_stats->rx_bytes = stats.rx_bytes;
+		vf_stats->tx_bytes = stats.tx_bytes;
+		vf_stats->broadcast = stats.broadcast;
+		vf_stats->multicast = stats.multicast;
+	}
+
+	return ret;
+}
+
+static const struct net_device_ops lionetdevops = {
+	.ndo_open		= liquidio_open,
+	.ndo_stop		= liquidio_stop,
+	.ndo_start_xmit		= liquidio_xmit,
+	.ndo_get_stats64	= liquidio_get_stats64,
+	.ndo_set_mac_address	= liquidio_set_mac,
+	.ndo_set_rx_mode	= liquidio_set_mcast_list,
+	.ndo_tx_timeout		= liquidio_tx_timeout,
+
+	.ndo_vlan_rx_add_vid    = liquidio_vlan_rx_add_vid,
+	.ndo_vlan_rx_kill_vid   = liquidio_vlan_rx_kill_vid,
+	.ndo_change_mtu		= liquidio_change_mtu,
+	.ndo_eth_ioctl		= liquidio_ioctl,
+	.ndo_fix_features	= liquidio_fix_features,
+	.ndo_set_features	= liquidio_set_features,
+	.ndo_set_vf_mac		= liquidio_set_vf_mac,
+	.ndo_set_vf_vlan	= liquidio_set_vf_vlan,
+	.ndo_get_vf_config	= liquidio_get_vf_config,
+	.ndo_set_vf_spoofchk	= liquidio_set_vf_spoofchk,
+	.ndo_set_vf_trust	= liquidio_set_vf_trust,
+	.ndo_set_vf_link_state  = liquidio_set_vf_link_state,
+	.ndo_get_vf_stats	= liquidio_get_vf_stats,
+	.ndo_get_port_parent_id	= liquidio_get_port_parent_id,
+};
+
+/**
+ * liquidio_init - Entry point for the liquidio module
+ */
+static int __init liquidio_init(void)
+{
+	int i;
+	struct handshake *hs;
+
+	init_completion(&first_stage);
+
+	octeon_init_device_list(OCTEON_CONFIG_TYPE_DEFAULT);
+
+	if (liquidio_init_pci())
+		return -EINVAL;
+
+	wait_for_completion_timeout(&first_stage, msecs_to_jiffies(1000));
+
+	for (i = 0; i < MAX_OCTEON_DEVICES; i++) {
+		hs = &handshake[i];
+		if (hs->pci_dev) {
+			wait_for_completion(&hs->init);
+			if (!hs->init_ok) {
+				/* init handshake failed */
+				dev_err(&hs->pci_dev->dev,
+					"Failed to init device\n");
+				liquidio_deinit_pci();
+				return -EIO;
+			}
+		}
+	}
+
+	for (i = 0; i < MAX_OCTEON_DEVICES; i++) {
+		hs = &handshake[i];
+		if (hs->pci_dev) {
+			wait_for_completion_timeout(&hs->started,
+						    msecs_to_jiffies(30000));
+			if (!hs->started_ok) {
+				/* starter handshake failed */
+				dev_err(&hs->pci_dev->dev,
+					"Firmware failed to start\n");
+				liquidio_deinit_pci();
+				return -EIO;
+			}
+		}
+	}
+
+	return 0;
+}
+
+static int lio_nic_info(struct octeon_recv_info *recv_info, void *buf)
+{
+	struct octeon_device *oct = (struct octeon_device *)buf;
+	struct octeon_recv_pkt *recv_pkt = recv_info->recv_pkt;
+	int gmxport = 0;
+	union oct_link_status *ls;
+	int i;
+
+	if (recv_pkt->buffer_size[0] != (sizeof(*ls) + OCT_DROQ_INFO_SIZE)) {
+		dev_err(&oct->pci_dev->dev, "Malformed NIC_INFO, len=%d, ifidx=%d\n",
+			recv_pkt->buffer_size[0],
+			recv_pkt->rh.r_nic_info.gmxport);
+		goto nic_info_err;
+	}
+
+	gmxport = recv_pkt->rh.r_nic_info.gmxport;
+	ls = (union oct_link_status *)(get_rbd(recv_pkt->buffer_ptr[0]) +
+		OCT_DROQ_INFO_SIZE);
+
+	octeon_swap_8B_data((u64 *)ls, (sizeof(union oct_link_status)) >> 3);
+	for (i = 0; i < oct->ifcount; i++) {
+		if (oct->props[i].gmxport == gmxport) {
+			update_link_status(oct->props[i].netdev, ls);
+			break;
+		}
+	}
+
+nic_info_err:
+	for (i = 0; i < recv_pkt->buffer_count; i++)
+		recv_buffer_free(recv_pkt->buffer_ptr[i]);
+	octeon_free_recv_info(recv_info);
+	return 0;
+}
+
+/**
+ * setup_nic_devices - Setup network interfaces
+ * @octeon_dev:  octeon device
+ *
+ * Called during init time for each device. It assumes the NIC
+ * is already up and running.  The link information for each
+ * interface is passed in link_info.
+ */
+static int setup_nic_devices(struct octeon_device *octeon_dev)
+{
+	struct lio *lio = NULL;
+	struct net_device *netdev;
+	u8 mac[6], i, j, *fw_ver, *micro_ver;
+	unsigned long micro;
+	u32 cur_ver;
+	struct octeon_soft_command *sc;
+	struct liquidio_if_cfg_resp *resp;
+	struct octdev_props *props;
+	int retval, num_iqueues, num_oqueues;
+	int max_num_queues = 0;
+	union oct_nic_if_cfg if_cfg;
+	unsigned int base_queue;
+	unsigned int gmx_port_id;
+	u32 resp_size, data_size;
+	u32 ifidx_or_pfnum;
+	struct lio_version *vdata;
+	struct devlink *devlink;
+	struct lio_devlink_priv *lio_devlink;
+
+	/* This is to handle link status changes */
+	octeon_register_dispatch_fn(octeon_dev, OPCODE_NIC,
+				    OPCODE_NIC_INFO,
+				    lio_nic_info, octeon_dev);
+
+	/* REQTYPE_RESP_NET and REQTYPE_SOFT_COMMAND do not have free functions.
+	 * They are handled directly.
+	 */
+	octeon_register_reqtype_free_fn(octeon_dev, REQTYPE_NORESP_NET,
+					free_netbuf);
+
+	octeon_register_reqtype_free_fn(octeon_dev, REQTYPE_NORESP_NET_SG,
+					free_netsgbuf);
+
+	octeon_register_reqtype_free_fn(octeon_dev, REQTYPE_RESP_NET_SG,
+					free_netsgbuf_with_resp);
+
+	for (i = 0; i < octeon_dev->ifcount; i++) {
+		resp_size = sizeof(struct liquidio_if_cfg_resp);
+		data_size = sizeof(struct lio_version);
+		sc = (struct octeon_soft_command *)
+			octeon_alloc_soft_command(octeon_dev, data_size,
+						  resp_size, 0);
+		resp = (struct liquidio_if_cfg_resp *)sc->virtrptr;
+		vdata = (struct lio_version *)sc->virtdptr;
+
+		*((u64 *)vdata) = 0;
+		vdata->major = cpu_to_be16(LIQUIDIO_BASE_MAJOR_VERSION);
+		vdata->minor = cpu_to_be16(LIQUIDIO_BASE_MINOR_VERSION);
+		vdata->micro = cpu_to_be16(LIQUIDIO_BASE_MICRO_VERSION);
+
+		if (OCTEON_CN23XX_PF(octeon_dev)) {
+			num_iqueues = octeon_dev->sriov_info.num_pf_rings;
+			num_oqueues = octeon_dev->sriov_info.num_pf_rings;
+			base_queue = octeon_dev->sriov_info.pf_srn;
+
+			gmx_port_id = octeon_dev->pf_num;
+			ifidx_or_pfnum = octeon_dev->pf_num;
+		} else {
+			num_iqueues = CFG_GET_NUM_TXQS_NIC_IF(
+						octeon_get_conf(octeon_dev), i);
+			num_oqueues = CFG_GET_NUM_RXQS_NIC_IF(
+						octeon_get_conf(octeon_dev), i);
+			base_queue = CFG_GET_BASE_QUE_NIC_IF(
+						octeon_get_conf(octeon_dev), i);
+			gmx_port_id = CFG_GET_GMXID_NIC_IF(
+						octeon_get_conf(octeon_dev), i);
+			ifidx_or_pfnum = i;
+		}
+
+		dev_dbg(&octeon_dev->pci_dev->dev,
+			"requesting config for interface %d, iqs %d, oqs %d\n",
+			ifidx_or_pfnum, num_iqueues, num_oqueues);
+
+		if_cfg.u64 = 0;
+		if_cfg.s.num_iqueues = num_iqueues;
+		if_cfg.s.num_oqueues = num_oqueues;
+		if_cfg.s.base_queue = base_queue;
+		if_cfg.s.gmx_port_id = gmx_port_id;
+
+		sc->iq_no = 0;
+
+		octeon_prepare_soft_command(octeon_dev, sc, OPCODE_NIC,
+					    OPCODE_NIC_IF_CFG, 0,
+					    if_cfg.u64, 0);
+
+		init_completion(&sc->complete);
+		sc->sc_status = OCTEON_REQUEST_PENDING;
+
+		retval = octeon_send_soft_command(octeon_dev, sc);
+		if (retval == IQ_SEND_FAILED) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"iq/oq config failed status: %x\n",
+				retval);
+			/* Soft instr is freed by driver in case of failure. */
+			octeon_free_soft_command(octeon_dev, sc);
+			return(-EIO);
+		}
+
+		/* Sleep on a wait queue till the cond flag indicates that the
+		 * response arrived or timed-out.
+		 */
+		retval = wait_for_sc_completion_timeout(octeon_dev, sc, 0);
+		if (retval)
+			return retval;
+
+		retval = resp->status;
+		if (retval) {
+			dev_err(&octeon_dev->pci_dev->dev, "iq/oq config failed\n");
+			WRITE_ONCE(sc->caller_is_done, true);
+			goto setup_nic_dev_done;
+		}
+		snprintf(octeon_dev->fw_info.liquidio_firmware_version,
+			 32, "%s",
+			 resp->cfg_info.liquidio_firmware_version);
+
+		/* Verify f/w version (in case of 'auto' loading from flash) */
+		fw_ver = octeon_dev->fw_info.liquidio_firmware_version;
+		if (memcmp(LIQUIDIO_BASE_VERSION,
+			   fw_ver,
+			   strlen(LIQUIDIO_BASE_VERSION))) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"Unmatched firmware version. Expected %s.x, got %s.\n",
+				LIQUIDIO_BASE_VERSION, fw_ver);
+			WRITE_ONCE(sc->caller_is_done, true);
+			goto setup_nic_dev_done;
+		} else if (atomic_read(octeon_dev->adapter_fw_state) ==
+			   FW_IS_PRELOADED) {
+			dev_info(&octeon_dev->pci_dev->dev,
+				 "Using auto-loaded firmware version %s.\n",
+				 fw_ver);
+		}
+
+		/* extract micro version field; point past '<maj>.<min>.' */
+		micro_ver = fw_ver + strlen(LIQUIDIO_BASE_VERSION) + 1;
+		if (kstrtoul(micro_ver, 10, &micro) != 0)
+			micro = 0;
+		octeon_dev->fw_info.ver.maj = LIQUIDIO_BASE_MAJOR_VERSION;
+		octeon_dev->fw_info.ver.min = LIQUIDIO_BASE_MINOR_VERSION;
+		octeon_dev->fw_info.ver.rev = micro;
+
+		octeon_swap_8B_data((u64 *)(&resp->cfg_info),
+				    (sizeof(struct liquidio_if_cfg_info)) >> 3);
+
+		num_iqueues = hweight64(resp->cfg_info.iqmask);
+		num_oqueues = hweight64(resp->cfg_info.oqmask);
+
+		if (!(num_iqueues) || !(num_oqueues)) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"Got bad iqueues (%016llx) or oqueues (%016llx) from firmware.\n",
+				resp->cfg_info.iqmask,
+				resp->cfg_info.oqmask);
+			WRITE_ONCE(sc->caller_is_done, true);
+			goto setup_nic_dev_done;
+		}
+
+		if (OCTEON_CN6XXX(octeon_dev)) {
+			max_num_queues = CFG_GET_IQ_MAX_Q(CHIP_CONF(octeon_dev,
+								    cn6xxx));
+		} else if (OCTEON_CN23XX_PF(octeon_dev)) {
+			max_num_queues = CFG_GET_IQ_MAX_Q(CHIP_CONF(octeon_dev,
+								    cn23xx_pf));
+		}
+
+		dev_dbg(&octeon_dev->pci_dev->dev,
+			"interface %d, iqmask %016llx, oqmask %016llx, numiqueues %d, numoqueues %d max_num_queues: %d\n",
+			i, resp->cfg_info.iqmask, resp->cfg_info.oqmask,
+			num_iqueues, num_oqueues, max_num_queues);
+		netdev = alloc_etherdev_mq(LIO_SIZE, max_num_queues);
+
+		if (!netdev) {
+			dev_err(&octeon_dev->pci_dev->dev, "Device allocation failed\n");
+			WRITE_ONCE(sc->caller_is_done, true);
+			goto setup_nic_dev_done;
+		}
+
+		SET_NETDEV_DEV(netdev, &octeon_dev->pci_dev->dev);
+
+		/* Associate the routines that will handle different
+		 * netdev tasks.
+		 */
+		netdev->netdev_ops = &lionetdevops;
+
+		retval = netif_set_real_num_rx_queues(netdev, num_oqueues);
+		if (retval) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"setting real number rx failed\n");
+			WRITE_ONCE(sc->caller_is_done, true);
+			goto setup_nic_dev_free;
+		}
+
+		retval = netif_set_real_num_tx_queues(netdev, num_iqueues);
+		if (retval) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"setting real number tx failed\n");
+			WRITE_ONCE(sc->caller_is_done, true);
+			goto setup_nic_dev_free;
+		}
+
+		lio = GET_LIO(netdev);
+
+		memset(lio, 0, sizeof(struct lio));
+
+		lio->ifidx = ifidx_or_pfnum;
+
+		props = &octeon_dev->props[i];
+		props->gmxport = resp->cfg_info.linfo.gmxport;
+		props->netdev = netdev;
+
+		lio->linfo.num_rxpciq = num_oqueues;
+		lio->linfo.num_txpciq = num_iqueues;
+		for (j = 0; j < num_oqueues; j++) {
+			lio->linfo.rxpciq[j].u64 =
+				resp->cfg_info.linfo.rxpciq[j].u64;
+		}
+		for (j = 0; j < num_iqueues; j++) {
+			lio->linfo.txpciq[j].u64 =
+				resp->cfg_info.linfo.txpciq[j].u64;
+		}
+		lio->linfo.hw_addr = resp->cfg_info.linfo.hw_addr;
+		lio->linfo.gmxport = resp->cfg_info.linfo.gmxport;
+		lio->linfo.link.u64 = resp->cfg_info.linfo.link.u64;
+
+		WRITE_ONCE(sc->caller_is_done, true);
+
+		lio->msg_enable = netif_msg_init(debug, DEFAULT_MSG_ENABLE);
+
+		if (OCTEON_CN23XX_PF(octeon_dev) ||
+		    OCTEON_CN6XXX(octeon_dev)) {
+			lio->dev_capability = NETIF_F_HIGHDMA
+					      | NETIF_F_IP_CSUM
+					      | NETIF_F_IPV6_CSUM
+					      | NETIF_F_SG | NETIF_F_RXCSUM
+					      | NETIF_F_GRO
+					      | NETIF_F_TSO | NETIF_F_TSO6
+					      | NETIF_F_LRO;
+		}
+		netif_set_tso_max_size(netdev, OCTNIC_GSO_MAX_SIZE);
+
+		/*  Copy of transmit encapsulation capabilities:
+		 *  TSO, TSO6, Checksums for this device
+		 */
+		lio->enc_dev_capability = NETIF_F_IP_CSUM
+					  | NETIF_F_IPV6_CSUM
+					  | NETIF_F_GSO_UDP_TUNNEL
+					  | NETIF_F_HW_CSUM | NETIF_F_SG
+					  | NETIF_F_RXCSUM
+					  | NETIF_F_TSO | NETIF_F_TSO6
+					  | NETIF_F_LRO;
+
+		netdev->hw_enc_features = (lio->enc_dev_capability &
+					   ~NETIF_F_LRO);
+
+		netdev->udp_tunnel_nic_info = &liquidio_udp_tunnels;
+
+		lio->dev_capability |= NETIF_F_GSO_UDP_TUNNEL;
+
+		netdev->vlan_features = lio->dev_capability;
+		/* Add any unchangeable hw features */
+		lio->dev_capability |=  NETIF_F_HW_VLAN_CTAG_FILTER |
+					NETIF_F_HW_VLAN_CTAG_RX |
+					NETIF_F_HW_VLAN_CTAG_TX;
+
+		netdev->features = (lio->dev_capability & ~NETIF_F_LRO);
+
+		netdev->hw_features = lio->dev_capability;
+		/*HW_VLAN_RX and HW_VLAN_FILTER is always on*/
+		netdev->hw_features = netdev->hw_features &
+			~NETIF_F_HW_VLAN_CTAG_RX;
+
+		/* MTU range: 68 - 16000 */
+		netdev->min_mtu = LIO_MIN_MTU_SIZE;
+		netdev->max_mtu = LIO_MAX_MTU_SIZE;
+
+		/* Point to the  properties for octeon device to which this
+		 * interface belongs.
+		 */
+		lio->oct_dev = octeon_dev;
+		lio->octprops = props;
+		lio->netdev = netdev;
+
+		dev_dbg(&octeon_dev->pci_dev->dev,
+			"if%d gmx: %d hw_addr: 0x%llx\n", i,
+			lio->linfo.gmxport, CVM_CAST64(lio->linfo.hw_addr));
+
+		for (j = 0; j < octeon_dev->sriov_info.max_vfs; j++) {
+			u8 vfmac[ETH_ALEN];
+
+			eth_random_addr(vfmac);
+			if (__liquidio_set_vf_mac(netdev, j, vfmac, false)) {
+				dev_err(&octeon_dev->pci_dev->dev,
+					"Error setting VF%d MAC address\n",
+					j);
+				goto setup_nic_dev_free;
+			}
+		}
+
+		/* 64-bit swap required on LE machines */
+		octeon_swap_8B_data(&lio->linfo.hw_addr, 1);
+		for (j = 0; j < 6; j++)
+			mac[j] = *((u8 *)(((u8 *)&lio->linfo.hw_addr) + 2 + j));
+
+		/* Copy MAC Address to OS network device structure */
+
+		eth_hw_addr_set(netdev, mac);
+
+		/* By default all interfaces on a single Octeon uses the same
+		 * tx and rx queues
+		 */
+		lio->txq = lio->linfo.txpciq[0].s.q_no;
+		lio->rxq = lio->linfo.rxpciq[0].s.q_no;
+		if (liquidio_setup_io_queues(octeon_dev, i,
+					     lio->linfo.num_txpciq,
+					     lio->linfo.num_rxpciq)) {
+			dev_err(&octeon_dev->pci_dev->dev, "I/O queues creation failed\n");
+			goto setup_nic_dev_free;
+		}
+
+		ifstate_set(lio, LIO_IFSTATE_DROQ_OPS);
+
+		lio->tx_qsize = octeon_get_tx_qsize(octeon_dev, lio->txq);
+		lio->rx_qsize = octeon_get_rx_qsize(octeon_dev, lio->rxq);
+
+		if (lio_setup_glists(octeon_dev, lio, num_iqueues)) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"Gather list allocation failed\n");
+			goto setup_nic_dev_free;
+		}
+
+		/* Register ethtool support */
+		liquidio_set_ethtool_ops(netdev);
+		if (lio->oct_dev->chip_id == OCTEON_CN23XX_PF_VID)
+			octeon_dev->priv_flags = OCT_PRIV_FLAG_DEFAULT;
+		else
+			octeon_dev->priv_flags = 0x0;
+
+		if (netdev->features & NETIF_F_LRO)
+			liquidio_set_feature(netdev, OCTNET_CMD_LRO_ENABLE,
+					     OCTNIC_LROIPV4 | OCTNIC_LROIPV6);
+
+		liquidio_set_feature(netdev, OCTNET_CMD_VLAN_FILTER_CTL,
+				     OCTNET_CMD_VLAN_FILTER_ENABLE);
+
+		if ((debug != -1) && (debug & NETIF_MSG_HW))
+			liquidio_set_feature(netdev,
+					     OCTNET_CMD_VERBOSE_ENABLE, 0);
+
+		if (setup_link_status_change_wq(netdev))
+			goto setup_nic_dev_free;
+
+		if ((octeon_dev->fw_info.app_cap_flags &
+		     LIQUIDIO_TIME_SYNC_CAP) &&
+		    setup_sync_octeon_time_wq(netdev))
+			goto setup_nic_dev_free;
+
+		if (setup_rx_oom_poll_fn(netdev))
+			goto setup_nic_dev_free;
+
+		/* Register the network device with the OS */
+		if (register_netdev(netdev)) {
+			dev_err(&octeon_dev->pci_dev->dev, "Device registration failed\n");
+			goto setup_nic_dev_free;
+		}
+
+		dev_dbg(&octeon_dev->pci_dev->dev,
+			"Setup NIC ifidx:%d mac:%02x%02x%02x%02x%02x%02x\n",
+			i, mac[0], mac[1], mac[2], mac[3], mac[4], mac[5]);
+		netif_carrier_off(netdev);
+		lio->link_changes++;
+
+		ifstate_set(lio, LIO_IFSTATE_REGISTERED);
+
+		/* Sending command to firmware to enable Rx checksum offload
+		 * by default at the time of setup of Liquidio driver for
+		 * this device
+		 */
+		liquidio_set_rxcsum_command(netdev, OCTNET_CMD_TNL_RX_CSUM_CTL,
+					    OCTNET_CMD_RXCSUM_ENABLE);
+		liquidio_set_feature(netdev, OCTNET_CMD_TNL_TX_CSUM_CTL,
+				     OCTNET_CMD_TXCSUM_ENABLE);
+
+		dev_dbg(&octeon_dev->pci_dev->dev,
+			"NIC ifidx:%d Setup successful\n", i);
+
+		if (octeon_dev->subsystem_id ==
+			OCTEON_CN2350_25GB_SUBSYS_ID ||
+		    octeon_dev->subsystem_id ==
+			OCTEON_CN2360_25GB_SUBSYS_ID) {
+			cur_ver = OCT_FW_VER(octeon_dev->fw_info.ver.maj,
+					     octeon_dev->fw_info.ver.min,
+					     octeon_dev->fw_info.ver.rev);
+
+			/* speed control unsupported in f/w older than 1.7.2 */
+			if (cur_ver < OCT_FW_VER(1, 7, 2)) {
+				dev_info(&octeon_dev->pci_dev->dev,
+					 "speed setting not supported by f/w.");
+				octeon_dev->speed_setting = 25;
+				octeon_dev->no_speed_setting = 1;
+			} else {
+				liquidio_get_speed(lio);
+			}
+
+			if (octeon_dev->speed_setting == 0) {
+				octeon_dev->speed_setting = 25;
+				octeon_dev->no_speed_setting = 1;
+			}
+		} else {
+			octeon_dev->no_speed_setting = 1;
+			octeon_dev->speed_setting = 10;
+		}
+		octeon_dev->speed_boot = octeon_dev->speed_setting;
+
+		/* don't read FEC setting if unsupported by f/w (see above) */
+		if (octeon_dev->speed_boot == 25 &&
+		    !octeon_dev->no_speed_setting) {
+			liquidio_get_fec(lio);
+			octeon_dev->props[lio->ifidx].fec_boot =
+				octeon_dev->props[lio->ifidx].fec;
+		}
+	}
+
+	device_lock(&octeon_dev->pci_dev->dev);
+	devlink = devlink_alloc(&liquidio_devlink_ops,
+				sizeof(struct lio_devlink_priv),
+				&octeon_dev->pci_dev->dev);
+	if (!devlink) {
+		device_unlock(&octeon_dev->pci_dev->dev);
+		dev_err(&octeon_dev->pci_dev->dev, "devlink alloc failed\n");
+		goto setup_nic_dev_free;
+	}
+
+	lio_devlink = devlink_priv(devlink);
+	lio_devlink->oct = octeon_dev;
+
+	octeon_dev->devlink = devlink;
+	octeon_dev->eswitch_mode = DEVLINK_ESWITCH_MODE_LEGACY;
+	devlink_register(devlink);
+	device_unlock(&octeon_dev->pci_dev->dev);
+
+	return 0;
+
+setup_nic_dev_free:
+
+	while (i--) {
+		dev_err(&octeon_dev->pci_dev->dev,
+			"NIC ifidx:%d Setup failed\n", i);
+		liquidio_destroy_nic_device(octeon_dev, i);
+	}
+
+setup_nic_dev_done:
+
+	return -ENODEV;
+}
+
+#ifdef CONFIG_PCI_IOV
+static int octeon_enable_sriov(struct octeon_device *oct)
+{
+	unsigned int num_vfs_alloced = oct->sriov_info.num_vfs_alloced;
+	struct pci_dev *vfdev;
+	int err;
+	u32 u;
+
+	if (OCTEON_CN23XX_PF(oct) && num_vfs_alloced) {
+		err = pci_enable_sriov(oct->pci_dev,
+				       oct->sriov_info.num_vfs_alloced);
+		if (err) {
+			dev_err(&oct->pci_dev->dev,
+				"OCTEON: Failed to enable PCI sriov: %d\n",
+				err);
+			oct->sriov_info.num_vfs_alloced = 0;
+			return err;
+		}
+		oct->sriov_info.sriov_enabled = 1;
+
+		/* init lookup table that maps DPI ring number to VF pci_dev
+		 * struct pointer
+		 */
+		u = 0;
+		vfdev = pci_get_device(PCI_VENDOR_ID_CAVIUM,
+				       OCTEON_CN23XX_VF_VID, NULL);
+		while (vfdev) {
+			if (vfdev->is_virtfn &&
+			    (vfdev->physfn == oct->pci_dev)) {
+				oct->sriov_info.dpiring_to_vfpcidev_lut[u] =
+					vfdev;
+				u += oct->sriov_info.rings_per_vf;
+			}
+			vfdev = pci_get_device(PCI_VENDOR_ID_CAVIUM,
+					       OCTEON_CN23XX_VF_VID, vfdev);
+		}
+	}
+
+	return num_vfs_alloced;
+}
+
+static int lio_pci_sriov_disable(struct octeon_device *oct)
+{
+	int u;
+
+	if (pci_vfs_assigned(oct->pci_dev)) {
+		dev_err(&oct->pci_dev->dev, "VFs are still assigned to VMs.\n");
+		return -EPERM;
+	}
+
+	pci_disable_sriov(oct->pci_dev);
+
+	u = 0;
+	while (u < MAX_POSSIBLE_VFS) {
+		oct->sriov_info.dpiring_to_vfpcidev_lut[u] = NULL;
+		u += oct->sriov_info.rings_per_vf;
+	}
+
+	oct->sriov_info.num_vfs_alloced = 0;
+	dev_info(&oct->pci_dev->dev, "oct->pf_num:%d disabled VFs\n",
+		 oct->pf_num);
+
+	return 0;
+}
+
+static int liquidio_enable_sriov(struct pci_dev *dev, int num_vfs)
+{
+	struct octeon_device *oct = pci_get_drvdata(dev);
+	int ret = 0;
+
+	if ((num_vfs == oct->sriov_info.num_vfs_alloced) &&
+	    (oct->sriov_info.sriov_enabled)) {
+		dev_info(&oct->pci_dev->dev, "oct->pf_num:%d already enabled num_vfs:%d\n",
+			 oct->pf_num, num_vfs);
+		return 0;
+	}
+
+	if (!num_vfs) {
+		lio_vf_rep_destroy(oct);
+		ret = lio_pci_sriov_disable(oct);
+	} else if (num_vfs > oct->sriov_info.max_vfs) {
+		dev_err(&oct->pci_dev->dev,
+			"OCTEON: Max allowed VFs:%d user requested:%d",
+			oct->sriov_info.max_vfs, num_vfs);
+		ret = -EPERM;
+	} else {
+		oct->sriov_info.num_vfs_alloced = num_vfs;
+		ret = octeon_enable_sriov(oct);
+		dev_info(&oct->pci_dev->dev, "oct->pf_num:%d num_vfs:%d\n",
+			 oct->pf_num, num_vfs);
+		ret = lio_vf_rep_create(oct);
+		if (ret)
+			dev_info(&oct->pci_dev->dev,
+				 "vf representor create failed");
+	}
+
+	return ret;
+}
+#endif
+
+/**
+ * liquidio_init_nic_module - initialize the NIC
+ * @oct: octeon device
+ *
+ * This initialization routine is called once the Octeon device application is
+ * up and running
+ */
+static int liquidio_init_nic_module(struct octeon_device *oct)
+{
+	int i, retval = 0;
+	int num_nic_ports = CFG_GET_NUM_NIC_PORTS(octeon_get_conf(oct));
+
+	dev_dbg(&oct->pci_dev->dev, "Initializing network interfaces\n");
+
+	/* only default iq and oq were initialized
+	 * initialize the rest as well
+	 */
+	/* run port_config command for each port */
+	oct->ifcount = num_nic_ports;
+
+	memset(oct->props, 0, sizeof(struct octdev_props) * num_nic_ports);
+
+	for (i = 0; i < MAX_OCTEON_LINKS; i++)
+		oct->props[i].gmxport = -1;
+
+	retval = setup_nic_devices(oct);
+	if (retval) {
+		dev_err(&oct->pci_dev->dev, "Setup NIC devices failed\n");
+		goto octnet_init_failure;
+	}
+
+	/* Call vf_rep_modinit if the firmware is switchdev capable
+	 * and do it from the first liquidio function probed.
+	 */
+	if (!oct->octeon_id &&
+	    oct->fw_info.app_cap_flags & LIQUIDIO_SWITCHDEV_CAP) {
+		retval = lio_vf_rep_modinit();
+		if (retval) {
+			liquidio_stop_nic_module(oct);
+			goto octnet_init_failure;
+		}
+	}
+
+	liquidio_ptp_init(oct);
+
+	dev_dbg(&oct->pci_dev->dev, "Network interfaces ready\n");
+
+	return retval;
+
+octnet_init_failure:
+
+	oct->ifcount = 0;
+
+	return retval;
+}
+
+/**
+ * nic_starter - finish init
+ * @work:  work struct work_struct
+ *
+ * starter callback that invokes the remaining initialization work after the NIC is up and running.
+ */
+static void nic_starter(struct work_struct *work)
+{
+	struct octeon_device *oct;
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+
+	oct = (struct octeon_device *)wk->ctxptr;
+
+	if (atomic_read(&oct->status) == OCT_DEV_RUNNING)
+		return;
+
+	/* If the status of the device is CORE_OK, the core
+	 * application has reported its application type. Call
+	 * any registered handlers now and move to the RUNNING
+	 * state.
+	 */
+	if (atomic_read(&oct->status) != OCT_DEV_CORE_OK) {
+		schedule_delayed_work(&oct->nic_poll_work.work,
+				      LIQUIDIO_STARTER_POLL_INTERVAL_MS);
+		return;
+	}
+
+	atomic_set(&oct->status, OCT_DEV_RUNNING);
+
+	if (oct->app_mode && oct->app_mode == CVM_DRV_NIC_APP) {
+		dev_dbg(&oct->pci_dev->dev, "Starting NIC module\n");
+
+		if (liquidio_init_nic_module(oct))
+			dev_err(&oct->pci_dev->dev, "NIC initialization failed\n");
+		else
+			handshake[oct->octeon_id].started_ok = 1;
+	} else {
+		dev_err(&oct->pci_dev->dev,
+			"Unexpected application running on NIC (%d). Check firmware.\n",
+			oct->app_mode);
+	}
+
+	complete(&handshake[oct->octeon_id].started);
+}
+
+static int
+octeon_recv_vf_drv_notice(struct octeon_recv_info *recv_info, void *buf)
+{
+	struct octeon_device *oct = (struct octeon_device *)buf;
+	struct octeon_recv_pkt *recv_pkt = recv_info->recv_pkt;
+	int i, notice, vf_idx;
+	bool cores_crashed;
+	u64 *data, vf_num;
+
+	notice = recv_pkt->rh.r.ossp;
+	data = (u64 *)(get_rbd(recv_pkt->buffer_ptr[0]) + OCT_DROQ_INFO_SIZE);
+
+	/* the first 64-bit word of data is the vf_num */
+	vf_num = data[0];
+	octeon_swap_8B_data(&vf_num, 1);
+	vf_idx = (int)vf_num - 1;
+
+	cores_crashed = READ_ONCE(oct->cores_crashed);
+
+	if (notice == VF_DRV_LOADED) {
+		if (!(oct->sriov_info.vf_drv_loaded_mask & BIT_ULL(vf_idx))) {
+			oct->sriov_info.vf_drv_loaded_mask |= BIT_ULL(vf_idx);
+			dev_info(&oct->pci_dev->dev,
+				 "driver for VF%d was loaded\n", vf_idx);
+			if (!cores_crashed)
+				try_module_get(THIS_MODULE);
+		}
+	} else if (notice == VF_DRV_REMOVED) {
+		if (oct->sriov_info.vf_drv_loaded_mask & BIT_ULL(vf_idx)) {
+			oct->sriov_info.vf_drv_loaded_mask &= ~BIT_ULL(vf_idx);
+			dev_info(&oct->pci_dev->dev,
+				 "driver for VF%d was removed\n", vf_idx);
+			if (!cores_crashed)
+				module_put(THIS_MODULE);
+		}
+	} else if (notice == VF_DRV_MACADDR_CHANGED) {
+		u8 *b = (u8 *)&data[1];
+
+		oct->sriov_info.vf_macaddr[vf_idx] = data[1];
+		dev_info(&oct->pci_dev->dev,
+			 "VF driver changed VF%d's MAC address to %pM\n",
+			 vf_idx, b + 2);
+	}
+
+	for (i = 0; i < recv_pkt->buffer_count; i++)
+		recv_buffer_free(recv_pkt->buffer_ptr[i]);
+	octeon_free_recv_info(recv_info);
+
+	return 0;
+}
+
+/**
+ * octeon_device_init - Device initialization for each Octeon device that is probed
+ * @octeon_dev:  octeon device
+ */
+static int octeon_device_init(struct octeon_device *octeon_dev)
+{
+	int j, ret;
+	char bootcmd[] = "\n";
+	char *dbg_enb = NULL;
+	enum lio_fw_state fw_state;
+	struct octeon_device_priv *oct_priv =
+		(struct octeon_device_priv *)octeon_dev->priv;
+	atomic_set(&octeon_dev->status, OCT_DEV_BEGIN_STATE);
+
+	/* Enable access to the octeon device and make its DMA capability
+	 * known to the OS.
+	 */
+	if (octeon_pci_os_setup(octeon_dev))
+		return 1;
+
+	atomic_set(&octeon_dev->status, OCT_DEV_PCI_ENABLE_DONE);
+
+	/* Identify the Octeon type and map the BAR address space. */
+	if (octeon_chip_specific_setup(octeon_dev)) {
+		dev_err(&octeon_dev->pci_dev->dev, "Chip specific setup failed\n");
+		return 1;
+	}
+
+	atomic_set(&octeon_dev->status, OCT_DEV_PCI_MAP_DONE);
+
+	/* Only add a reference after setting status 'OCT_DEV_PCI_MAP_DONE',
+	 * since that is what is required for the reference to be removed
+	 * during de-initialization (see 'octeon_destroy_resources').
+	 */
+	octeon_register_device(octeon_dev, octeon_dev->pci_dev->bus->number,
+			       PCI_SLOT(octeon_dev->pci_dev->devfn),
+			       PCI_FUNC(octeon_dev->pci_dev->devfn),
+			       true);
+
+	octeon_dev->app_mode = CVM_DRV_INVALID_APP;
+
+	/* CN23XX supports preloaded firmware if the following is true:
+	 *
+	 * The adapter indicates that firmware is currently running AND
+	 * 'fw_type' is 'auto'.
+	 *
+	 * (default state is NEEDS_TO_BE_LOADED, override it if appropriate).
+	 */
+	if (OCTEON_CN23XX_PF(octeon_dev) &&
+	    cn23xx_fw_loaded(octeon_dev) && fw_type_is_auto()) {
+		atomic_cmpxchg(octeon_dev->adapter_fw_state,
+			       FW_NEEDS_TO_BE_LOADED, FW_IS_PRELOADED);
+	}
+
+	/* If loading firmware, only first device of adapter needs to do so. */
+	fw_state = atomic_cmpxchg(octeon_dev->adapter_fw_state,
+				  FW_NEEDS_TO_BE_LOADED,
+				  FW_IS_BEING_LOADED);
+
+	/* Here, [local variable] 'fw_state' is set to one of:
+	 *
+	 *   FW_IS_PRELOADED:       No firmware is to be loaded (see above)
+	 *   FW_NEEDS_TO_BE_LOADED: The driver's first instance will load
+	 *                          firmware to the adapter.
+	 *   FW_IS_BEING_LOADED:    The driver's second instance will not load
+	 *                          firmware to the adapter.
+	 */
+
+	/* Prior to f/w load, perform a soft reset of the Octeon device;
+	 * if error resetting, return w/error.
+	 */
+	if (fw_state == FW_NEEDS_TO_BE_LOADED)
+		if (octeon_dev->fn_list.soft_reset(octeon_dev))
+			return 1;
+
+	/* Initialize the dispatch mechanism used to push packets arriving on
+	 * Octeon Output queues.
+	 */
+	if (octeon_init_dispatch_list(octeon_dev))
+		return 1;
+
+	octeon_register_dispatch_fn(octeon_dev, OPCODE_NIC,
+				    OPCODE_NIC_CORE_DRV_ACTIVE,
+				    octeon_core_drv_init,
+				    octeon_dev);
+
+	octeon_register_dispatch_fn(octeon_dev, OPCODE_NIC,
+				    OPCODE_NIC_VF_DRV_NOTICE,
+				    octeon_recv_vf_drv_notice, octeon_dev);
+	INIT_DELAYED_WORK(&octeon_dev->nic_poll_work.work, nic_starter);
+	octeon_dev->nic_poll_work.ctxptr = (void *)octeon_dev;
+	schedule_delayed_work(&octeon_dev->nic_poll_work.work,
+			      LIQUIDIO_STARTER_POLL_INTERVAL_MS);
+
+	atomic_set(&octeon_dev->status, OCT_DEV_DISPATCH_INIT_DONE);
+
+	if (octeon_set_io_queues_off(octeon_dev)) {
+		dev_err(&octeon_dev->pci_dev->dev, "setting io queues off failed\n");
+		return 1;
+	}
+
+	if (OCTEON_CN23XX_PF(octeon_dev)) {
+		ret = octeon_dev->fn_list.setup_device_regs(octeon_dev);
+		if (ret) {
+			dev_err(&octeon_dev->pci_dev->dev, "OCTEON: Failed to configure device registers\n");
+			return ret;
+		}
+	}
+
+	/* Initialize soft command buffer pool
+	 */
+	if (octeon_setup_sc_buffer_pool(octeon_dev)) {
+		dev_err(&octeon_dev->pci_dev->dev, "sc buffer pool allocation failed\n");
+		return 1;
+	}
+	atomic_set(&octeon_dev->status, OCT_DEV_SC_BUFF_POOL_INIT_DONE);
+
+	/*  Setup the data structures that manage this Octeon's Input queues. */
+	if (octeon_setup_instr_queues(octeon_dev)) {
+		dev_err(&octeon_dev->pci_dev->dev,
+			"instruction queue initialization failed\n");
+		return 1;
+	}
+	atomic_set(&octeon_dev->status, OCT_DEV_INSTR_QUEUE_INIT_DONE);
+
+	/* Initialize lists to manage the requests of different types that
+	 * arrive from user & kernel applications for this octeon device.
+	 */
+	if (octeon_setup_response_list(octeon_dev)) {
+		dev_err(&octeon_dev->pci_dev->dev, "Response list allocation failed\n");
+		return 1;
+	}
+	atomic_set(&octeon_dev->status, OCT_DEV_RESP_LIST_INIT_DONE);
+
+	if (octeon_setup_output_queues(octeon_dev)) {
+		dev_err(&octeon_dev->pci_dev->dev, "Output queue initialization failed\n");
+		return 1;
+	}
+
+	atomic_set(&octeon_dev->status, OCT_DEV_DROQ_INIT_DONE);
+
+	if (OCTEON_CN23XX_PF(octeon_dev)) {
+		if (octeon_dev->fn_list.setup_mbox(octeon_dev)) {
+			dev_err(&octeon_dev->pci_dev->dev, "OCTEON: Mailbox setup failed\n");
+			return 1;
+		}
+		atomic_set(&octeon_dev->status, OCT_DEV_MBOX_SETUP_DONE);
+
+		if (octeon_allocate_ioq_vector
+				(octeon_dev,
+				 octeon_dev->sriov_info.num_pf_rings)) {
+			dev_err(&octeon_dev->pci_dev->dev, "OCTEON: ioq vector allocation failed\n");
+			return 1;
+		}
+		atomic_set(&octeon_dev->status, OCT_DEV_MSIX_ALLOC_VECTOR_DONE);
+
+	} else {
+		/* The input and output queue registers were setup earlier (the
+		 * queues were not enabled). Any additional registers
+		 * that need to be programmed should be done now.
+		 */
+		ret = octeon_dev->fn_list.setup_device_regs(octeon_dev);
+		if (ret) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"Failed to configure device registers\n");
+			return ret;
+		}
+	}
+
+	/* Initialize the tasklet that handles output queue packet processing.*/
+	dev_dbg(&octeon_dev->pci_dev->dev, "Initializing droq tasklet\n");
+	tasklet_setup(&oct_priv->droq_tasklet, octeon_droq_bh);
+
+	/* Setup the interrupt handler and record the INT SUM register address
+	 */
+	if (octeon_setup_interrupt(octeon_dev,
+				   octeon_dev->sriov_info.num_pf_rings))
+		return 1;
+
+	/* Enable Octeon device interrupts */
+	octeon_dev->fn_list.enable_interrupt(octeon_dev, OCTEON_ALL_INTR);
+
+	atomic_set(&octeon_dev->status, OCT_DEV_INTR_SET_DONE);
+
+	/* Send Credit for Octeon Output queues. Credits are always sent BEFORE
+	 * the output queue is enabled.
+	 * This ensures that we'll receive the f/w CORE DRV_ACTIVE message in
+	 * case we've configured CN23XX_SLI_GBL_CONTROL[NOPTR_D] = 0.
+	 * Otherwise, it is possible that the DRV_ACTIVE message will be sent
+	 * before any credits have been issued, causing the ring to be reset
+	 * (and the f/w appear to never have started).
+	 */
+	for (j = 0; j < octeon_dev->num_oqs; j++)
+		writel(octeon_dev->droq[j]->max_count,
+		       octeon_dev->droq[j]->pkts_credit_reg);
+
+	/* Enable the input and output queues for this Octeon device */
+	ret = octeon_dev->fn_list.enable_io_queues(octeon_dev);
+	if (ret) {
+		dev_err(&octeon_dev->pci_dev->dev, "Failed to enable input/output queues");
+		return ret;
+	}
+
+	atomic_set(&octeon_dev->status, OCT_DEV_IO_QUEUES_DONE);
+
+	if (fw_state == FW_NEEDS_TO_BE_LOADED) {
+		dev_dbg(&octeon_dev->pci_dev->dev, "Waiting for DDR initialization...\n");
+		if (!ddr_timeout) {
+			dev_info(&octeon_dev->pci_dev->dev,
+				 "WAITING. Set ddr_timeout to non-zero value to proceed with initialization.\n");
+		}
+
+		schedule_timeout_uninterruptible(HZ * LIO_RESET_SECS);
+
+		/* Wait for the octeon to initialize DDR after the soft-reset.*/
+		while (!ddr_timeout) {
+			set_current_state(TASK_INTERRUPTIBLE);
+			if (schedule_timeout(HZ / 10)) {
+				/* user probably pressed Control-C */
+				return 1;
+			}
+		}
+		ret = octeon_wait_for_ddr_init(octeon_dev, &ddr_timeout);
+		if (ret) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"DDR not initialized. Please confirm that board is configured to boot from Flash, ret: %d\n",
+				ret);
+			return 1;
+		}
+
+		if (octeon_wait_for_bootloader(octeon_dev, 1000)) {
+			dev_err(&octeon_dev->pci_dev->dev, "Board not responding\n");
+			return 1;
+		}
+
+		/* Divert uboot to take commands from host instead. */
+		ret = octeon_console_send_cmd(octeon_dev, bootcmd, 50);
+
+		dev_dbg(&octeon_dev->pci_dev->dev, "Initializing consoles\n");
+		ret = octeon_init_consoles(octeon_dev);
+		if (ret) {
+			dev_err(&octeon_dev->pci_dev->dev, "Could not access board consoles\n");
+			return 1;
+		}
+		/* If console debug enabled, specify empty string to use default
+		 * enablement ELSE specify NULL string for 'disabled'.
+		 */
+		dbg_enb = octeon_console_debug_enabled(0) ? "" : NULL;
+		ret = octeon_add_console(octeon_dev, 0, dbg_enb);
+		if (ret) {
+			dev_err(&octeon_dev->pci_dev->dev, "Could not access board console\n");
+			return 1;
+		} else if (octeon_console_debug_enabled(0)) {
+			/* If console was added AND we're logging console output
+			 * then set our console print function.
+			 */
+			octeon_dev->console[0].print = octeon_dbg_console_print;
+		}
+
+		atomic_set(&octeon_dev->status, OCT_DEV_CONSOLE_INIT_DONE);
+
+		dev_dbg(&octeon_dev->pci_dev->dev, "Loading firmware\n");
+		ret = load_firmware(octeon_dev);
+		if (ret) {
+			dev_err(&octeon_dev->pci_dev->dev, "Could not load firmware to board\n");
+			return 1;
+		}
+
+		atomic_set(octeon_dev->adapter_fw_state, FW_HAS_BEEN_LOADED);
+	}
+
+	handshake[octeon_dev->octeon_id].init_ok = 1;
+	complete(&handshake[octeon_dev->octeon_id].init);
+
+	atomic_set(&octeon_dev->status, OCT_DEV_HOST_OK);
+	oct_priv->dev = octeon_dev;
+
+	return 0;
+}
+
+/**
+ * octeon_dbg_console_print - Debug console print function
+ * @oct:  octeon device
+ * @console_num: console number
+ * @prefix:      first portion of line to display
+ * @suffix:      second portion of line to display
+ *
+ * The OCTEON debug console outputs entire lines (excluding '\n').
+ * Normally, the line will be passed in the 'prefix' parameter.
+ * However, due to buffering, it is possible for a line to be split into two
+ * parts, in which case they will be passed as the 'prefix' parameter and
+ * 'suffix' parameter.
+ */
+static int octeon_dbg_console_print(struct octeon_device *oct, u32 console_num,
+				    char *prefix, char *suffix)
+{
+	if (prefix && suffix)
+		dev_info(&oct->pci_dev->dev, "%u: %s%s\n", console_num, prefix,
+			 suffix);
+	else if (prefix)
+		dev_info(&oct->pci_dev->dev, "%u: %s\n", console_num, prefix);
+	else if (suffix)
+		dev_info(&oct->pci_dev->dev, "%u: %s\n", console_num, suffix);
+
+	return 0;
+}
+
+/**
+ * liquidio_exit - Exits the module
+ */
+static void __exit liquidio_exit(void)
+{
+	liquidio_deinit_pci();
+
+	pr_info("LiquidIO network module is now unloaded\n");
+}
+
+module_init(liquidio_init);
+module_exit(liquidio_exit);
diff --git a/drivers/net/ethernet/cavium/liquidio/lio_vf_main.c b/drivers/net/ethernet/cavium/liquidio/lio_vf_main.c
new file mode 100644
index 000000000..ac196883f
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/lio_vf_main.c
@@ -0,0 +1,2442 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include <net/vxlan.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_nic.h"
+#include "octeon_main.h"
+#include "octeon_network.h"
+#include "cn23xx_vf_device.h"
+
+MODULE_AUTHOR("Cavium Networks, <support@cavium.com>");
+MODULE_DESCRIPTION("Cavium LiquidIO Intelligent Server Adapter Virtual Function Driver");
+MODULE_LICENSE("GPL");
+
+static int debug = -1;
+module_param(debug, int, 0644);
+MODULE_PARM_DESC(debug, "NETIF_MSG debug bits");
+
+#define DEFAULT_MSG_ENABLE (NETIF_MSG_DRV | NETIF_MSG_PROBE | NETIF_MSG_LINK)
+
+struct oct_timestamp_resp {
+	u64 rh;
+	u64 timestamp;
+	u64 status;
+};
+
+union tx_info {
+	u64 u64;
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		u16 gso_size;
+		u16 gso_segs;
+		u32 reserved;
+#else
+		u32 reserved;
+		u16 gso_segs;
+		u16 gso_size;
+#endif
+	} s;
+};
+
+#define OCTNIC_GSO_MAX_HEADER_SIZE 128
+#define OCTNIC_GSO_MAX_SIZE \
+		(CN23XX_DEFAULT_INPUT_JABBER - OCTNIC_GSO_MAX_HEADER_SIZE)
+
+static int
+liquidio_vf_probe(struct pci_dev *pdev, const struct pci_device_id *ent);
+static void liquidio_vf_remove(struct pci_dev *pdev);
+static int octeon_device_init(struct octeon_device *oct);
+static int liquidio_stop(struct net_device *netdev);
+
+static int lio_wait_for_oq_pkts(struct octeon_device *oct)
+{
+	struct octeon_device_priv *oct_priv =
+	    (struct octeon_device_priv *)oct->priv;
+	int retry = MAX_IO_PENDING_PKT_COUNT;
+	int pkt_cnt = 0, pending_pkts;
+	int i;
+
+	do {
+		pending_pkts = 0;
+
+		for (i = 0; i < MAX_OCTEON_OUTPUT_QUEUES(oct); i++) {
+			if (!(oct->io_qmask.oq & BIT_ULL(i)))
+				continue;
+			pkt_cnt += octeon_droq_check_hw_for_pkts(oct->droq[i]);
+		}
+		if (pkt_cnt > 0) {
+			pending_pkts += pkt_cnt;
+			tasklet_schedule(&oct_priv->droq_tasklet);
+		}
+		pkt_cnt = 0;
+		schedule_timeout_uninterruptible(1);
+
+	} while (retry-- && pending_pkts);
+
+	return pkt_cnt;
+}
+
+/**
+ * pcierror_quiesce_device - Cause device to go quiet so it can be safely removed/reset/etc
+ * @oct: Pointer to Octeon device
+ */
+static void pcierror_quiesce_device(struct octeon_device *oct)
+{
+	int i;
+
+	/* Disable the input and output queues now. No more packets will
+	 * arrive from Octeon, but we should wait for all packet processing
+	 * to finish.
+	 */
+
+	/* To allow for in-flight requests */
+	schedule_timeout_uninterruptible(100);
+
+	if (wait_for_pending_requests(oct))
+		dev_err(&oct->pci_dev->dev, "There were pending requests\n");
+
+	/* Force all requests waiting to be fetched by OCTEON to complete. */
+	for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct); i++) {
+		struct octeon_instr_queue *iq;
+
+		if (!(oct->io_qmask.iq & BIT_ULL(i)))
+			continue;
+		iq = oct->instr_queue[i];
+
+		if (atomic_read(&iq->instr_pending)) {
+			spin_lock_bh(&iq->lock);
+			iq->fill_cnt = 0;
+			iq->octeon_read_index = iq->host_write_index;
+			iq->stats.instr_processed +=
+			    atomic_read(&iq->instr_pending);
+			lio_process_iq_request_list(oct, iq, 0);
+			spin_unlock_bh(&iq->lock);
+		}
+	}
+
+	/* Force all pending ordered list requests to time out. */
+	lio_process_ordered_list(oct, 1);
+
+	/* We do not need to wait for output queue packets to be processed. */
+}
+
+/**
+ * cleanup_aer_uncorrect_error_status - Cleanup PCI AER uncorrectable error status
+ * @dev: Pointer to PCI device
+ */
+static void cleanup_aer_uncorrect_error_status(struct pci_dev *dev)
+{
+	u32 status, mask;
+	int pos = 0x100;
+
+	pr_info("%s :\n", __func__);
+
+	pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_STATUS, &status);
+	pci_read_config_dword(dev, pos + PCI_ERR_UNCOR_SEVER, &mask);
+	if (dev->error_state == pci_channel_io_normal)
+		status &= ~mask; /* Clear corresponding nonfatal bits */
+	else
+		status &= mask; /* Clear corresponding fatal bits */
+	pci_write_config_dword(dev, pos + PCI_ERR_UNCOR_STATUS, status);
+}
+
+/**
+ * stop_pci_io - Stop all PCI IO to a given device
+ * @oct: Pointer to Octeon device
+ */
+static void stop_pci_io(struct octeon_device *oct)
+{
+	struct msix_entry *msix_entries;
+	int i;
+
+	/* No more instructions will be forwarded. */
+	atomic_set(&oct->status, OCT_DEV_IN_RESET);
+
+	for (i = 0; i < oct->ifcount; i++)
+		netif_device_detach(oct->props[i].netdev);
+
+	/* Disable interrupts  */
+	oct->fn_list.disable_interrupt(oct, OCTEON_ALL_INTR);
+
+	pcierror_quiesce_device(oct);
+	if (oct->msix_on) {
+		msix_entries = (struct msix_entry *)oct->msix_entries;
+		for (i = 0; i < oct->num_msix_irqs; i++) {
+			/* clear the affinity_cpumask */
+			irq_set_affinity_hint(msix_entries[i].vector,
+					      NULL);
+			free_irq(msix_entries[i].vector,
+				 &oct->ioq_vector[i]);
+		}
+		pci_disable_msix(oct->pci_dev);
+		kfree(oct->msix_entries);
+		oct->msix_entries = NULL;
+		octeon_free_ioq_vector(oct);
+	}
+	dev_dbg(&oct->pci_dev->dev, "Device state is now %s\n",
+		lio_get_state_string(&oct->status));
+
+	/* making it a common function for all OCTEON models */
+	cleanup_aer_uncorrect_error_status(oct->pci_dev);
+
+	pci_disable_device(oct->pci_dev);
+}
+
+/**
+ * liquidio_pcie_error_detected - called when PCI error is detected
+ * @pdev: Pointer to PCI device
+ * @state: The current pci connection state
+ *
+ * This function is called after a PCI bus error affecting
+ * this device has been detected.
+ */
+static pci_ers_result_t liquidio_pcie_error_detected(struct pci_dev *pdev,
+						     pci_channel_state_t state)
+{
+	struct octeon_device *oct = pci_get_drvdata(pdev);
+
+	/* Non-correctable Non-fatal errors */
+	if (state == pci_channel_io_normal) {
+		dev_err(&oct->pci_dev->dev, "Non-correctable non-fatal error reported:\n");
+		cleanup_aer_uncorrect_error_status(oct->pci_dev);
+		return PCI_ERS_RESULT_CAN_RECOVER;
+	}
+
+	/* Non-correctable Fatal errors */
+	dev_err(&oct->pci_dev->dev, "Non-correctable FATAL reported by PCI AER driver\n");
+	stop_pci_io(oct);
+
+	return PCI_ERS_RESULT_DISCONNECT;
+}
+
+/* For PCI-E Advanced Error Recovery (AER) Interface */
+static const struct pci_error_handlers liquidio_vf_err_handler = {
+	.error_detected = liquidio_pcie_error_detected,
+};
+
+static const struct pci_device_id liquidio_vf_pci_tbl[] = {
+	{
+		PCI_VENDOR_ID_CAVIUM, OCTEON_CN23XX_VF_VID,
+		PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0
+	},
+	{
+		0, 0, 0, 0, 0, 0, 0
+	}
+};
+MODULE_DEVICE_TABLE(pci, liquidio_vf_pci_tbl);
+
+static struct pci_driver liquidio_vf_pci_driver = {
+	.name		= "LiquidIO_VF",
+	.id_table	= liquidio_vf_pci_tbl,
+	.probe		= liquidio_vf_probe,
+	.remove		= liquidio_vf_remove,
+	.err_handler	= &liquidio_vf_err_handler,    /* For AER */
+};
+
+/**
+ * print_link_info - Print link information
+ * @netdev: network device
+ */
+static void print_link_info(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+
+	if (!ifstate_check(lio, LIO_IFSTATE_RESETTING) &&
+	    ifstate_check(lio, LIO_IFSTATE_REGISTERED)) {
+		struct oct_link_info *linfo = &lio->linfo;
+
+		if (linfo->link.s.link_up) {
+			netif_info(lio, link, lio->netdev, "%d Mbps %s Duplex UP\n",
+				   linfo->link.s.speed,
+				   (linfo->link.s.duplex) ? "Full" : "Half");
+		} else {
+			netif_info(lio, link, lio->netdev, "Link Down\n");
+		}
+	}
+}
+
+/**
+ * octnet_link_status_change - Routine to notify MTU change
+ * @work: work_struct data structure
+ */
+static void octnet_link_status_change(struct work_struct *work)
+{
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct lio *lio = (struct lio *)wk->ctxptr;
+
+	/* lio->linfo.link.s.mtu always contains max MTU of the lio interface.
+	 * this API is invoked only when new max-MTU of the interface is
+	 * less than current MTU.
+	 */
+	rtnl_lock();
+	dev_set_mtu(lio->netdev, lio->linfo.link.s.mtu);
+	rtnl_unlock();
+}
+
+/**
+ * setup_link_status_change_wq - Sets up the mtu status change work
+ * @netdev: network device
+ */
+static int setup_link_status_change_wq(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+
+	lio->link_status_wq.wq = alloc_workqueue("link-status",
+						 WQ_MEM_RECLAIM, 0);
+	if (!lio->link_status_wq.wq) {
+		dev_err(&oct->pci_dev->dev, "unable to create cavium link status wq\n");
+		return -1;
+	}
+	INIT_DELAYED_WORK(&lio->link_status_wq.wk.work,
+			  octnet_link_status_change);
+	lio->link_status_wq.wk.ctxptr = lio;
+
+	return 0;
+}
+
+static void cleanup_link_status_change_wq(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+
+	if (lio->link_status_wq.wq) {
+		cancel_delayed_work_sync(&lio->link_status_wq.wk.work);
+		destroy_workqueue(lio->link_status_wq.wq);
+	}
+}
+
+/**
+ * update_link_status - Update link status
+ * @netdev: network device
+ * @ls: link status structure
+ *
+ * Called on receipt of a link status response from the core application to
+ * update each interface's link status.
+ */
+static void update_link_status(struct net_device *netdev,
+			       union oct_link_status *ls)
+{
+	struct lio *lio = GET_LIO(netdev);
+	int current_max_mtu = lio->linfo.link.s.mtu;
+	struct octeon_device *oct = lio->oct_dev;
+
+	if ((lio->intf_open) && (lio->linfo.link.u64 != ls->u64)) {
+		lio->linfo.link.u64 = ls->u64;
+
+		print_link_info(netdev);
+		lio->link_changes++;
+
+		if (lio->linfo.link.s.link_up) {
+			netif_carrier_on(netdev);
+			wake_txqs(netdev);
+		} else {
+			netif_carrier_off(netdev);
+			stop_txqs(netdev);
+		}
+
+		if (lio->linfo.link.s.mtu != current_max_mtu) {
+			dev_info(&oct->pci_dev->dev,
+				 "Max MTU Changed from %d to %d\n",
+				 current_max_mtu, lio->linfo.link.s.mtu);
+			netdev->max_mtu = lio->linfo.link.s.mtu;
+		}
+
+		if (lio->linfo.link.s.mtu < netdev->mtu) {
+			dev_warn(&oct->pci_dev->dev,
+				 "Current MTU is higher than new max MTU; Reducing the current mtu from %d to %d\n",
+				 netdev->mtu, lio->linfo.link.s.mtu);
+			queue_delayed_work(lio->link_status_wq.wq,
+					   &lio->link_status_wq.wk.work, 0);
+		}
+	}
+}
+
+/**
+ * liquidio_vf_probe - PCI probe handler
+ * @pdev: PCI device structure
+ * @ent: unused
+ */
+static int
+liquidio_vf_probe(struct pci_dev *pdev,
+		  const struct pci_device_id __maybe_unused *ent)
+{
+	struct octeon_device *oct_dev = NULL;
+
+	oct_dev = octeon_allocate_device(pdev->device,
+					 sizeof(struct octeon_device_priv));
+
+	if (!oct_dev) {
+		dev_err(&pdev->dev, "Unable to allocate device\n");
+		return -ENOMEM;
+	}
+	oct_dev->msix_on = LIO_FLAG_MSIX_ENABLED;
+
+	dev_info(&pdev->dev, "Initializing device %x:%x.\n",
+		 (u32)pdev->vendor, (u32)pdev->device);
+
+	/* Assign octeon_device for this device to the private data area. */
+	pci_set_drvdata(pdev, oct_dev);
+
+	/* set linux specific device pointer */
+	oct_dev->pci_dev = pdev;
+
+	oct_dev->subsystem_id = pdev->subsystem_vendor |
+		(pdev->subsystem_device << 16);
+
+	if (octeon_device_init(oct_dev)) {
+		liquidio_vf_remove(pdev);
+		return -ENOMEM;
+	}
+
+	dev_dbg(&oct_dev->pci_dev->dev, "Device is ready\n");
+
+	return 0;
+}
+
+/**
+ * octeon_pci_flr - PCI FLR for each Octeon device.
+ * @oct: octeon device
+ */
+static void octeon_pci_flr(struct octeon_device *oct)
+{
+	pci_save_state(oct->pci_dev);
+
+	pci_cfg_access_lock(oct->pci_dev);
+
+	/* Quiesce the device completely */
+	pci_write_config_word(oct->pci_dev, PCI_COMMAND,
+			      PCI_COMMAND_INTX_DISABLE);
+
+	pcie_flr(oct->pci_dev);
+
+	pci_cfg_access_unlock(oct->pci_dev);
+
+	pci_restore_state(oct->pci_dev);
+}
+
+/**
+ * octeon_destroy_resources - Destroy resources associated with octeon device
+ * @oct: octeon device
+ */
+static void octeon_destroy_resources(struct octeon_device *oct)
+{
+	struct octeon_device_priv *oct_priv =
+		(struct octeon_device_priv *)oct->priv;
+	struct msix_entry *msix_entries;
+	int i;
+
+	switch (atomic_read(&oct->status)) {
+	case OCT_DEV_RUNNING:
+	case OCT_DEV_CORE_OK:
+		/* No more instructions will be forwarded. */
+		atomic_set(&oct->status, OCT_DEV_IN_RESET);
+
+		oct->app_mode = CVM_DRV_INVALID_APP;
+		dev_dbg(&oct->pci_dev->dev, "Device state is now %s\n",
+			lio_get_state_string(&oct->status));
+
+		schedule_timeout_uninterruptible(HZ / 10);
+
+		fallthrough;
+	case OCT_DEV_HOST_OK:
+	case OCT_DEV_IO_QUEUES_DONE:
+		if (lio_wait_for_instr_fetch(oct))
+			dev_err(&oct->pci_dev->dev, "IQ had pending instructions\n");
+
+		if (wait_for_pending_requests(oct))
+			dev_err(&oct->pci_dev->dev, "There were pending requests\n");
+
+		/* Disable the input and output queues now. No more packets will
+		 * arrive from Octeon, but we should wait for all packet
+		 * processing to finish.
+		 */
+		oct->fn_list.disable_io_queues(oct);
+
+		if (lio_wait_for_oq_pkts(oct))
+			dev_err(&oct->pci_dev->dev, "OQ had pending packets\n");
+
+		/* Force all requests waiting to be fetched by OCTEON to
+		 * complete.
+		 */
+		for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct); i++) {
+			struct octeon_instr_queue *iq;
+
+			if (!(oct->io_qmask.iq & BIT_ULL(i)))
+				continue;
+			iq = oct->instr_queue[i];
+
+			if (atomic_read(&iq->instr_pending)) {
+				spin_lock_bh(&iq->lock);
+				iq->fill_cnt = 0;
+				iq->octeon_read_index = iq->host_write_index;
+				iq->stats.instr_processed +=
+					atomic_read(&iq->instr_pending);
+				lio_process_iq_request_list(oct, iq, 0);
+				spin_unlock_bh(&iq->lock);
+			}
+		}
+
+		lio_process_ordered_list(oct, 1);
+		octeon_free_sc_done_list(oct);
+		octeon_free_sc_zombie_list(oct);
+
+		fallthrough;
+	case OCT_DEV_INTR_SET_DONE:
+		/* Disable interrupts  */
+		oct->fn_list.disable_interrupt(oct, OCTEON_ALL_INTR);
+
+		if (oct->msix_on) {
+			msix_entries = (struct msix_entry *)oct->msix_entries;
+			for (i = 0; i < oct->num_msix_irqs; i++) {
+				if (oct->ioq_vector[i].vector) {
+					irq_set_affinity_hint(
+							msix_entries[i].vector,
+							NULL);
+					free_irq(msix_entries[i].vector,
+						 &oct->ioq_vector[i]);
+					oct->ioq_vector[i].vector = 0;
+				}
+			}
+			pci_disable_msix(oct->pci_dev);
+			kfree(oct->msix_entries);
+			oct->msix_entries = NULL;
+			kfree(oct->irq_name_storage);
+			oct->irq_name_storage = NULL;
+		}
+		/* Soft reset the octeon device before exiting */
+		if (!pcie_reset_flr(oct->pci_dev, PCI_RESET_PROBE))
+			octeon_pci_flr(oct);
+		else
+			cn23xx_vf_ask_pf_to_do_flr(oct);
+
+		fallthrough;
+	case OCT_DEV_MSIX_ALLOC_VECTOR_DONE:
+		octeon_free_ioq_vector(oct);
+
+		fallthrough;
+	case OCT_DEV_MBOX_SETUP_DONE:
+		oct->fn_list.free_mbox(oct);
+
+		fallthrough;
+	case OCT_DEV_IN_RESET:
+	case OCT_DEV_DROQ_INIT_DONE:
+		mdelay(100);
+		for (i = 0; i < MAX_OCTEON_OUTPUT_QUEUES(oct); i++) {
+			if (!(oct->io_qmask.oq & BIT_ULL(i)))
+				continue;
+			octeon_delete_droq(oct, i);
+		}
+
+		fallthrough;
+	case OCT_DEV_RESP_LIST_INIT_DONE:
+		octeon_delete_response_list(oct);
+
+		fallthrough;
+	case OCT_DEV_INSTR_QUEUE_INIT_DONE:
+		for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct); i++) {
+			if (!(oct->io_qmask.iq & BIT_ULL(i)))
+				continue;
+			octeon_delete_instr_queue(oct, i);
+		}
+
+		fallthrough;
+	case OCT_DEV_SC_BUFF_POOL_INIT_DONE:
+		octeon_free_sc_buffer_pool(oct);
+
+		fallthrough;
+	case OCT_DEV_DISPATCH_INIT_DONE:
+		octeon_delete_dispatch_list(oct);
+		cancel_delayed_work_sync(&oct->nic_poll_work.work);
+
+		fallthrough;
+	case OCT_DEV_PCI_MAP_DONE:
+		octeon_unmap_pci_barx(oct, 0);
+		octeon_unmap_pci_barx(oct, 1);
+
+		fallthrough;
+	case OCT_DEV_PCI_ENABLE_DONE:
+		pci_clear_master(oct->pci_dev);
+		/* Disable the device, releasing the PCI INT */
+		pci_disable_device(oct->pci_dev);
+
+		fallthrough;
+	case OCT_DEV_BEGIN_STATE:
+		/* Nothing to be done here either */
+		break;
+	}
+
+	tasklet_kill(&oct_priv->droq_tasklet);
+}
+
+/**
+ * send_rx_ctrl_cmd - Send Rx control command
+ * @lio: per-network private data
+ * @start_stop: whether to start or stop
+ */
+static int send_rx_ctrl_cmd(struct lio *lio, int start_stop)
+{
+	struct octeon_device *oct = (struct octeon_device *)lio->oct_dev;
+	struct octeon_soft_command *sc;
+	union octnet_cmd *ncmd;
+	int retval;
+
+	if (oct->props[lio->ifidx].rx_on == start_stop)
+		return 0;
+
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct, OCTNET_CMD_SIZE,
+					  16, 0);
+	if (!sc) {
+		netif_info(lio, rx_err, lio->netdev,
+			   "Failed to allocate octeon_soft_command struct\n");
+		return -ENOMEM;
+	}
+
+	ncmd = (union octnet_cmd *)sc->virtdptr;
+
+	ncmd->u64 = 0;
+	ncmd->s.cmd = OCTNET_CMD_RX_CTL;
+	ncmd->s.param1 = start_stop;
+
+	octeon_swap_8B_data((u64 *)ncmd, (OCTNET_CMD_SIZE >> 3));
+
+	sc->iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC,
+				    OPCODE_NIC_CMD, 0, 0, 0);
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	retval = octeon_send_soft_command(oct, sc);
+	if (retval == IQ_SEND_FAILED) {
+		netif_info(lio, rx_err, lio->netdev, "Failed to send RX Control message\n");
+		octeon_free_soft_command(oct, sc);
+	} else {
+		/* Sleep on a wait queue till the cond flag indicates that the
+		 * response arrived or timed-out.
+		 */
+		retval = wait_for_sc_completion_timeout(oct, sc, 0);
+		if (retval)
+			return retval;
+
+		oct->props[lio->ifidx].rx_on = start_stop;
+		WRITE_ONCE(sc->caller_is_done, true);
+	}
+
+	return retval;
+}
+
+/**
+ * liquidio_destroy_nic_device - Destroy NIC device interface
+ * @oct: octeon device
+ * @ifidx: which interface to destroy
+ *
+ * Cleanup associated with each interface for an Octeon device  when NIC
+ * module is being unloaded or if initialization fails during load.
+ */
+static void liquidio_destroy_nic_device(struct octeon_device *oct, int ifidx)
+{
+	struct net_device *netdev = oct->props[ifidx].netdev;
+	struct octeon_device_priv *oct_priv =
+		(struct octeon_device_priv *)oct->priv;
+	struct napi_struct *napi, *n;
+	struct lio *lio;
+
+	if (!netdev) {
+		dev_err(&oct->pci_dev->dev, "%s No netdevice ptr for index %d\n",
+			__func__, ifidx);
+		return;
+	}
+
+	lio = GET_LIO(netdev);
+
+	dev_dbg(&oct->pci_dev->dev, "NIC device cleanup\n");
+
+	if (atomic_read(&lio->ifstate) & LIO_IFSTATE_RUNNING)
+		liquidio_stop(netdev);
+
+	if (oct->props[lio->ifidx].napi_enabled == 1) {
+		list_for_each_entry_safe(napi, n, &netdev->napi_list, dev_list)
+			napi_disable(napi);
+
+		oct->props[lio->ifidx].napi_enabled = 0;
+
+		oct->droq[0]->ops.poll_mode = 0;
+	}
+
+	/* Delete NAPI */
+	list_for_each_entry_safe(napi, n, &netdev->napi_list, dev_list)
+		netif_napi_del(napi);
+
+	tasklet_enable(&oct_priv->droq_tasklet);
+
+	if (atomic_read(&lio->ifstate) & LIO_IFSTATE_REGISTERED)
+		unregister_netdev(netdev);
+
+	cleanup_rx_oom_poll_fn(netdev);
+
+	cleanup_link_status_change_wq(netdev);
+
+	lio_delete_glists(lio);
+
+	free_netdev(netdev);
+
+	oct->props[ifidx].gmxport = -1;
+
+	oct->props[ifidx].netdev = NULL;
+}
+
+/**
+ * liquidio_stop_nic_module - Stop complete NIC functionality
+ * @oct: octeon device
+ */
+static int liquidio_stop_nic_module(struct octeon_device *oct)
+{
+	struct lio *lio;
+	int i, j;
+
+	dev_dbg(&oct->pci_dev->dev, "Stopping network interfaces\n");
+	if (!oct->ifcount) {
+		dev_err(&oct->pci_dev->dev, "Init for Octeon was not completed\n");
+		return 1;
+	}
+
+	spin_lock_bh(&oct->cmd_resp_wqlock);
+	oct->cmd_resp_state = OCT_DRV_OFFLINE;
+	spin_unlock_bh(&oct->cmd_resp_wqlock);
+
+	for (i = 0; i < oct->ifcount; i++) {
+		lio = GET_LIO(oct->props[i].netdev);
+		for (j = 0; j < oct->num_oqs; j++)
+			octeon_unregister_droq_ops(oct,
+						   lio->linfo.rxpciq[j].s.q_no);
+	}
+
+	for (i = 0; i < oct->ifcount; i++)
+		liquidio_destroy_nic_device(oct, i);
+
+	dev_dbg(&oct->pci_dev->dev, "Network interfaces stopped\n");
+	return 0;
+}
+
+/**
+ * liquidio_vf_remove - Cleans up resources at unload time
+ * @pdev: PCI device structure
+ */
+static void liquidio_vf_remove(struct pci_dev *pdev)
+{
+	struct octeon_device *oct_dev = pci_get_drvdata(pdev);
+
+	dev_dbg(&oct_dev->pci_dev->dev, "Stopping device\n");
+
+	if (oct_dev->app_mode == CVM_DRV_NIC_APP)
+		liquidio_stop_nic_module(oct_dev);
+
+	/* Reset the octeon device and cleanup all memory allocated for
+	 * the octeon device by driver.
+	 */
+	octeon_destroy_resources(oct_dev);
+
+	dev_info(&oct_dev->pci_dev->dev, "Device removed\n");
+
+	/* This octeon device has been removed. Update the global
+	 * data structure to reflect this. Free the device structure.
+	 */
+	octeon_free_device_mem(oct_dev);
+}
+
+/**
+ * octeon_pci_os_setup - PCI initialization for each Octeon device.
+ * @oct: octeon device
+ */
+static int octeon_pci_os_setup(struct octeon_device *oct)
+{
+#ifdef CONFIG_PCI_IOV
+	/* setup PCI stuff first */
+	if (!oct->pci_dev->physfn)
+		octeon_pci_flr(oct);
+#endif
+
+	if (pci_enable_device(oct->pci_dev)) {
+		dev_err(&oct->pci_dev->dev, "pci_enable_device failed\n");
+		return 1;
+	}
+
+	if (dma_set_mask_and_coherent(&oct->pci_dev->dev, DMA_BIT_MASK(64))) {
+		dev_err(&oct->pci_dev->dev, "Unexpected DMA device capability\n");
+		pci_disable_device(oct->pci_dev);
+		return 1;
+	}
+
+	/* Enable PCI DMA Master. */
+	pci_set_master(oct->pci_dev);
+
+	return 0;
+}
+
+/**
+ * free_netbuf - Unmap and free network buffer
+ * @buf: buffer
+ */
+static void free_netbuf(void *buf)
+{
+	struct octnet_buf_free_info *finfo;
+	struct sk_buff *skb;
+	struct lio *lio;
+
+	finfo = (struct octnet_buf_free_info *)buf;
+	skb = finfo->skb;
+	lio = finfo->lio;
+
+	dma_unmap_single(&lio->oct_dev->pci_dev->dev, finfo->dptr, skb->len,
+			 DMA_TO_DEVICE);
+
+	tx_buffer_free(skb);
+}
+
+/**
+ * free_netsgbuf - Unmap and free gather buffer
+ * @buf: buffer
+ */
+static void free_netsgbuf(void *buf)
+{
+	struct octnet_buf_free_info *finfo;
+	struct octnic_gather *g;
+	struct sk_buff *skb;
+	int i, frags, iq;
+	struct lio *lio;
+
+	finfo = (struct octnet_buf_free_info *)buf;
+	skb = finfo->skb;
+	lio = finfo->lio;
+	g = finfo->g;
+	frags = skb_shinfo(skb)->nr_frags;
+
+	dma_unmap_single(&lio->oct_dev->pci_dev->dev,
+			 g->sg[0].ptr[0], (skb->len - skb->data_len),
+			 DMA_TO_DEVICE);
+
+	i = 1;
+	while (frags--) {
+		skb_frag_t *frag = &skb_shinfo(skb)->frags[i - 1];
+
+		dma_unmap_page(&lio->oct_dev->pci_dev->dev,
+			       g->sg[(i >> 2)].ptr[(i & 3)],
+			       skb_frag_size(frag), DMA_TO_DEVICE);
+		i++;
+	}
+
+	iq = skb_iq(lio->oct_dev, skb);
+
+	spin_lock(&lio->glist_lock[iq]);
+	list_add_tail(&g->list, &lio->glist[iq]);
+	spin_unlock(&lio->glist_lock[iq]);
+
+	tx_buffer_free(skb);
+}
+
+/**
+ * free_netsgbuf_with_resp - Unmap and free gather buffer with response
+ * @buf: buffer
+ */
+static void free_netsgbuf_with_resp(void *buf)
+{
+	struct octnet_buf_free_info *finfo;
+	struct octeon_soft_command *sc;
+	struct octnic_gather *g;
+	struct sk_buff *skb;
+	int i, frags, iq;
+	struct lio *lio;
+
+	sc = (struct octeon_soft_command *)buf;
+	skb = (struct sk_buff *)sc->callback_arg;
+	finfo = (struct octnet_buf_free_info *)&skb->cb;
+
+	lio = finfo->lio;
+	g = finfo->g;
+	frags = skb_shinfo(skb)->nr_frags;
+
+	dma_unmap_single(&lio->oct_dev->pci_dev->dev,
+			 g->sg[0].ptr[0], (skb->len - skb->data_len),
+			 DMA_TO_DEVICE);
+
+	i = 1;
+	while (frags--) {
+		skb_frag_t *frag = &skb_shinfo(skb)->frags[i - 1];
+
+		dma_unmap_page(&lio->oct_dev->pci_dev->dev,
+			       g->sg[(i >> 2)].ptr[(i & 3)],
+			       skb_frag_size(frag), DMA_TO_DEVICE);
+		i++;
+	}
+
+	iq = skb_iq(lio->oct_dev, skb);
+
+	spin_lock(&lio->glist_lock[iq]);
+	list_add_tail(&g->list, &lio->glist[iq]);
+	spin_unlock(&lio->glist_lock[iq]);
+
+	/* Don't free the skb yet */
+}
+
+/**
+ * liquidio_open - Net device open for LiquidIO
+ * @netdev: network device
+ */
+static int liquidio_open(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octeon_device_priv *oct_priv =
+		(struct octeon_device_priv *)oct->priv;
+	struct napi_struct *napi, *n;
+	int ret = 0;
+
+	if (!oct->props[lio->ifidx].napi_enabled) {
+		tasklet_disable(&oct_priv->droq_tasklet);
+
+		list_for_each_entry_safe(napi, n, &netdev->napi_list, dev_list)
+			napi_enable(napi);
+
+		oct->props[lio->ifidx].napi_enabled = 1;
+
+		oct->droq[0]->ops.poll_mode = 1;
+	}
+
+	ifstate_set(lio, LIO_IFSTATE_RUNNING);
+
+	/* Ready for link status updates */
+	lio->intf_open = 1;
+
+	netif_info(lio, ifup, lio->netdev, "Interface Open, ready for traffic\n");
+	start_txqs(netdev);
+
+	INIT_DELAYED_WORK(&lio->stats_wk.work, lio_fetch_stats);
+	lio->stats_wk.ctxptr = lio;
+	schedule_delayed_work(&lio->stats_wk.work, msecs_to_jiffies
+					(LIQUIDIO_NDEV_STATS_POLL_TIME_MS));
+
+	/* tell Octeon to start forwarding packets to host */
+	ret = send_rx_ctrl_cmd(lio, 1);
+	if (ret)
+		return ret;
+
+	dev_info(&oct->pci_dev->dev, "%s interface is opened\n", netdev->name);
+
+	return ret;
+}
+
+/**
+ * liquidio_stop - jNet device stop for LiquidIO
+ * @netdev: network device
+ */
+static int liquidio_stop(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octeon_device_priv *oct_priv =
+		(struct octeon_device_priv *)oct->priv;
+	struct napi_struct *napi, *n;
+	int ret = 0;
+
+	/* tell Octeon to stop forwarding packets to host */
+	ret = send_rx_ctrl_cmd(lio, 0);
+	if (ret)
+		return ret;
+
+	netif_info(lio, ifdown, lio->netdev, "Stopping interface!\n");
+	/* Inform that netif carrier is down */
+	lio->intf_open = 0;
+	lio->linfo.link.s.link_up = 0;
+
+	netif_carrier_off(netdev);
+	lio->link_changes++;
+
+	ifstate_reset(lio, LIO_IFSTATE_RUNNING);
+
+	stop_txqs(netdev);
+
+	/* Wait for any pending Rx descriptors */
+	if (lio_wait_for_clean_oq(oct))
+		netif_info(lio, rx_err, lio->netdev,
+			   "Proceeding with stop interface after partial RX desc processing\n");
+
+	if (oct->props[lio->ifidx].napi_enabled == 1) {
+		list_for_each_entry_safe(napi, n, &netdev->napi_list, dev_list)
+			napi_disable(napi);
+
+		oct->props[lio->ifidx].napi_enabled = 0;
+
+		oct->droq[0]->ops.poll_mode = 0;
+
+		tasklet_enable(&oct_priv->droq_tasklet);
+	}
+
+	cancel_delayed_work_sync(&lio->stats_wk.work);
+
+	dev_info(&oct->pci_dev->dev, "%s interface is stopped\n", netdev->name);
+
+	return ret;
+}
+
+/**
+ * get_new_flags - Converts a mask based on net device flags
+ * @netdev: network device
+ *
+ * This routine generates a octnet_ifflags mask from the net device flags
+ * received from the OS.
+ */
+static enum octnet_ifflags get_new_flags(struct net_device *netdev)
+{
+	enum octnet_ifflags f = OCTNET_IFFLAG_UNICAST;
+
+	if (netdev->flags & IFF_PROMISC)
+		f |= OCTNET_IFFLAG_PROMISC;
+
+	if (netdev->flags & IFF_ALLMULTI)
+		f |= OCTNET_IFFLAG_ALLMULTI;
+
+	if (netdev->flags & IFF_MULTICAST) {
+		f |= OCTNET_IFFLAG_MULTICAST;
+
+		/* Accept all multicast addresses if there are more than we
+		 * can handle
+		 */
+		if (netdev_mc_count(netdev) > MAX_OCTEON_MULTICAST_ADDR)
+			f |= OCTNET_IFFLAG_ALLMULTI;
+	}
+
+	if (netdev->flags & IFF_BROADCAST)
+		f |= OCTNET_IFFLAG_BROADCAST;
+
+	return f;
+}
+
+static void liquidio_set_uc_list(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	struct netdev_hw_addr *ha;
+	u64 *mac;
+
+	if (lio->netdev_uc_count == netdev_uc_count(netdev))
+		return;
+
+	if (netdev_uc_count(netdev) > MAX_NCTRL_UDD) {
+		dev_err(&oct->pci_dev->dev, "too many MAC addresses in netdev uc list\n");
+		return;
+	}
+
+	lio->netdev_uc_count = netdev_uc_count(netdev);
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+	nctrl.ncmd.s.cmd = OCTNET_CMD_SET_UC_LIST;
+	nctrl.ncmd.s.more = lio->netdev_uc_count;
+	nctrl.ncmd.s.param1 = oct->vf_num;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	/* copy all the addresses into the udd */
+	mac = &nctrl.udd[0];
+	netdev_for_each_uc_addr(ha, netdev) {
+		ether_addr_copy(((u8 *)mac) + 2, ha->addr);
+		mac++;
+	}
+
+	octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+}
+
+/**
+ * liquidio_set_mcast_list - Net device set_multicast_list
+ * @netdev: network device
+ */
+static void liquidio_set_mcast_list(struct net_device *netdev)
+{
+	int mc_count = min(netdev_mc_count(netdev), MAX_OCTEON_MULTICAST_ADDR);
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	struct netdev_hw_addr *ha;
+	u64 *mc;
+	int ret;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	/* Create a ctrl pkt command to be sent to core app. */
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_SET_MULTI_LIST;
+	nctrl.ncmd.s.param1 = get_new_flags(netdev);
+	nctrl.ncmd.s.param2 = mc_count;
+	nctrl.ncmd.s.more = mc_count;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	/* copy all the addresses into the udd */
+	mc = &nctrl.udd[0];
+	netdev_for_each_mc_addr(ha, netdev) {
+		*mc = 0;
+		ether_addr_copy(((u8 *)mc) + 2, ha->addr);
+		/* no need to swap bytes */
+		if (++mc > &nctrl.udd[mc_count])
+			break;
+	}
+
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+
+	/* Apparently, any activity in this call from the kernel has to
+	 * be atomic. So we won't wait for response.
+	 */
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "DEVFLAGS change failed in core (ret: 0x%x)\n",
+			ret);
+	}
+
+	liquidio_set_uc_list(netdev);
+}
+
+/**
+ * liquidio_set_mac - Net device set_mac_address
+ * @netdev: network device
+ * @p: opaque pointer to sockaddr
+ */
+static int liquidio_set_mac(struct net_device *netdev, void *p)
+{
+	struct sockaddr *addr = (struct sockaddr *)p;
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	if (!is_valid_ether_addr(addr->sa_data))
+		return -EADDRNOTAVAIL;
+
+	if (ether_addr_equal(addr->sa_data, netdev->dev_addr))
+		return 0;
+
+	if (lio->linfo.macaddr_is_admin_asgnd)
+		return -EPERM;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_CHANGE_MACADDR;
+	nctrl.ncmd.s.param1 = 0;
+	nctrl.ncmd.s.more = 1;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+
+	nctrl.udd[0] = 0;
+	/* The MAC Address is presented in network byte order. */
+	ether_addr_copy((u8 *)&nctrl.udd[0] + 2, addr->sa_data);
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret < 0) {
+		dev_err(&oct->pci_dev->dev, "MAC Address change failed\n");
+		return -ENOMEM;
+	}
+
+	if (nctrl.sc_status ==
+	    FIRMWARE_STATUS_CODE(OCTEON_REQUEST_NO_PERMISSION)) {
+		dev_err(&oct->pci_dev->dev, "MAC Address change failed: no permission\n");
+		return -EPERM;
+	}
+
+	eth_hw_addr_set(netdev, addr->sa_data);
+	ether_addr_copy(((u8 *)&lio->linfo.hw_addr) + 2, addr->sa_data);
+
+	return 0;
+}
+
+static void
+liquidio_get_stats64(struct net_device *netdev,
+		     struct rtnl_link_stats64 *lstats)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct;
+	u64 pkts = 0, drop = 0, bytes = 0;
+	struct oct_droq_stats *oq_stats;
+	struct oct_iq_stats *iq_stats;
+	int i, iq_no, oq_no;
+
+	oct = lio->oct_dev;
+
+	if (ifstate_check(lio, LIO_IFSTATE_RESETTING))
+		return;
+
+	for (i = 0; i < oct->num_iqs; i++) {
+		iq_no = lio->linfo.txpciq[i].s.q_no;
+		iq_stats = &oct->instr_queue[iq_no]->stats;
+		pkts += iq_stats->tx_done;
+		drop += iq_stats->tx_dropped;
+		bytes += iq_stats->tx_tot_bytes;
+	}
+
+	lstats->tx_packets = pkts;
+	lstats->tx_bytes = bytes;
+	lstats->tx_dropped = drop;
+
+	pkts = 0;
+	drop = 0;
+	bytes = 0;
+
+	for (i = 0; i < oct->num_oqs; i++) {
+		oq_no = lio->linfo.rxpciq[i].s.q_no;
+		oq_stats = &oct->droq[oq_no]->stats;
+		pkts += oq_stats->rx_pkts_received;
+		drop += (oq_stats->rx_dropped +
+			 oq_stats->dropped_nodispatch +
+			 oq_stats->dropped_toomany +
+			 oq_stats->dropped_nomem);
+		bytes += oq_stats->rx_bytes_received;
+	}
+
+	lstats->rx_bytes = bytes;
+	lstats->rx_packets = pkts;
+	lstats->rx_dropped = drop;
+
+	lstats->multicast = oct->link_stats.fromwire.fw_total_mcast;
+
+	/* detailed rx_errors: */
+	lstats->rx_length_errors = oct->link_stats.fromwire.l2_err;
+	/* recved pkt with crc error */
+	lstats->rx_crc_errors = oct->link_stats.fromwire.fcs_err;
+	/* recv'd frame alignment error */
+	lstats->rx_frame_errors = oct->link_stats.fromwire.frame_err;
+
+	lstats->rx_errors = lstats->rx_length_errors + lstats->rx_crc_errors +
+			    lstats->rx_frame_errors;
+
+	/* detailed tx_errors */
+	lstats->tx_aborted_errors = oct->link_stats.fromhost.fw_err_pko;
+	lstats->tx_carrier_errors = oct->link_stats.fromhost.fw_err_link;
+
+	lstats->tx_errors = lstats->tx_aborted_errors +
+		lstats->tx_carrier_errors;
+}
+
+/**
+ * hwtstamp_ioctl - Handler for SIOCSHWTSTAMP ioctl
+ * @netdev: network device
+ * @ifr: interface request
+ */
+static int hwtstamp_ioctl(struct net_device *netdev, struct ifreq *ifr)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct hwtstamp_config conf;
+
+	if (copy_from_user(&conf, ifr->ifr_data, sizeof(conf)))
+		return -EFAULT;
+
+	switch (conf.tx_type) {
+	case HWTSTAMP_TX_ON:
+	case HWTSTAMP_TX_OFF:
+		break;
+	default:
+		return -ERANGE;
+	}
+
+	switch (conf.rx_filter) {
+	case HWTSTAMP_FILTER_NONE:
+		break;
+	case HWTSTAMP_FILTER_ALL:
+	case HWTSTAMP_FILTER_SOME:
+	case HWTSTAMP_FILTER_PTP_V1_L4_EVENT:
+	case HWTSTAMP_FILTER_PTP_V1_L4_SYNC:
+	case HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_L4_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_L4_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_L2_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_L2_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_L2_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_DELAY_REQ:
+	case HWTSTAMP_FILTER_NTP_ALL:
+		conf.rx_filter = HWTSTAMP_FILTER_ALL;
+		break;
+	default:
+		return -ERANGE;
+	}
+
+	if (conf.rx_filter == HWTSTAMP_FILTER_ALL)
+		ifstate_set(lio, LIO_IFSTATE_RX_TIMESTAMP_ENABLED);
+
+	else
+		ifstate_reset(lio, LIO_IFSTATE_RX_TIMESTAMP_ENABLED);
+
+	return copy_to_user(ifr->ifr_data, &conf, sizeof(conf)) ? -EFAULT : 0;
+}
+
+/**
+ * liquidio_ioctl - ioctl handler
+ * @netdev: network device
+ * @ifr: interface request
+ * @cmd: command
+ */
+static int liquidio_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd)
+{
+	switch (cmd) {
+	case SIOCSHWTSTAMP:
+		return hwtstamp_ioctl(netdev, ifr);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static void handle_timestamp(struct octeon_device *oct, u32 status, void *buf)
+{
+	struct sk_buff *skb = (struct sk_buff *)buf;
+	struct octnet_buf_free_info *finfo;
+	struct oct_timestamp_resp *resp;
+	struct octeon_soft_command *sc;
+	struct lio *lio;
+
+	finfo = (struct octnet_buf_free_info *)skb->cb;
+	lio = finfo->lio;
+	sc = finfo->sc;
+	oct = lio->oct_dev;
+	resp = (struct oct_timestamp_resp *)sc->virtrptr;
+
+	if (status != OCTEON_REQUEST_DONE) {
+		dev_err(&oct->pci_dev->dev, "Tx timestamp instruction failed. Status: %llx\n",
+			CVM_CAST64(status));
+		resp->timestamp = 0;
+	}
+
+	octeon_swap_8B_data(&resp->timestamp, 1);
+
+	if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS)) {
+		struct skb_shared_hwtstamps ts;
+		u64 ns = resp->timestamp;
+
+		netif_info(lio, tx_done, lio->netdev,
+			   "Got resulting SKBTX_HW_TSTAMP skb=%p ns=%016llu\n",
+			   skb, (unsigned long long)ns);
+		ts.hwtstamp = ns_to_ktime(ns + lio->ptp_adjust);
+		skb_tstamp_tx(skb, &ts);
+	}
+
+	octeon_free_soft_command(oct, sc);
+	tx_buffer_free(skb);
+}
+
+/* send_nic_timestamp_pkt - Send a data packet that will be timestamped
+ * @oct: octeon device
+ * @ndata: pointer to network data
+ * @finfo: pointer to private network data
+ */
+static int send_nic_timestamp_pkt(struct octeon_device *oct,
+				  struct octnic_data_pkt *ndata,
+				  struct octnet_buf_free_info *finfo,
+				  int xmit_more)
+{
+	struct octeon_soft_command *sc;
+	int ring_doorbell;
+	struct lio *lio;
+	int retval;
+	u32 len;
+
+	lio = finfo->lio;
+
+	sc = octeon_alloc_soft_command_resp(oct, &ndata->cmd,
+					    sizeof(struct oct_timestamp_resp));
+	finfo->sc = sc;
+
+	if (!sc) {
+		dev_err(&oct->pci_dev->dev, "No memory for timestamped data packet\n");
+		return IQ_SEND_FAILED;
+	}
+
+	if (ndata->reqtype == REQTYPE_NORESP_NET)
+		ndata->reqtype = REQTYPE_RESP_NET;
+	else if (ndata->reqtype == REQTYPE_NORESP_NET_SG)
+		ndata->reqtype = REQTYPE_RESP_NET_SG;
+
+	sc->callback = handle_timestamp;
+	sc->callback_arg = finfo->skb;
+	sc->iq_no = ndata->q_no;
+
+	len = (u32)((struct octeon_instr_ih3 *)(&sc->cmd.cmd3.ih3))->dlengsz;
+
+	ring_doorbell = !xmit_more;
+
+	retval = octeon_send_command(oct, sc->iq_no, ring_doorbell, &sc->cmd,
+				     sc, len, ndata->reqtype);
+
+	if (retval == IQ_SEND_FAILED) {
+		dev_err(&oct->pci_dev->dev, "timestamp data packet failed status: %x\n",
+			retval);
+		octeon_free_soft_command(oct, sc);
+	} else {
+		netif_info(lio, tx_queued, lio->netdev, "Queued timestamp packet\n");
+	}
+
+	return retval;
+}
+
+/**
+ * liquidio_xmit - Transmit networks packets to the Octeon interface
+ * @skb: skbuff struct to be passed to network layer.
+ * @netdev: pointer to network device
+ * @returns whether the packet was transmitted to the device okay or not
+ *             (NETDEV_TX_OK or NETDEV_TX_BUSY)
+ */
+static netdev_tx_t liquidio_xmit(struct sk_buff *skb, struct net_device *netdev)
+{
+	struct octnet_buf_free_info *finfo;
+	union octnic_cmd_setup cmdsetup;
+	struct octnic_data_pkt ndata;
+	struct octeon_instr_irh *irh;
+	struct oct_iq_stats *stats;
+	struct octeon_device *oct;
+	int q_idx = 0, iq_no = 0;
+	union tx_info *tx_info;
+	int xmit_more = 0;
+	struct lio *lio;
+	int status = 0;
+	u64 dptr = 0;
+	u32 tag = 0;
+	int j;
+
+	lio = GET_LIO(netdev);
+	oct = lio->oct_dev;
+
+	q_idx = skb_iq(lio->oct_dev, skb);
+	tag = q_idx;
+	iq_no = lio->linfo.txpciq[q_idx].s.q_no;
+
+	stats = &oct->instr_queue[iq_no]->stats;
+
+	/* Check for all conditions in which the current packet cannot be
+	 * transmitted.
+	 */
+	if (!(atomic_read(&lio->ifstate) & LIO_IFSTATE_RUNNING) ||
+	    (!lio->linfo.link.s.link_up) || (skb->len <= 0)) {
+		netif_info(lio, tx_err, lio->netdev, "Transmit failed link_status : %d\n",
+			   lio->linfo.link.s.link_up);
+		goto lio_xmit_failed;
+	}
+
+	/* Use space in skb->cb to store info used to unmap and
+	 * free the buffers.
+	 */
+	finfo = (struct octnet_buf_free_info *)skb->cb;
+	finfo->lio = lio;
+	finfo->skb = skb;
+	finfo->sc = NULL;
+
+	/* Prepare the attributes for the data to be passed to OSI. */
+	memset(&ndata, 0, sizeof(struct octnic_data_pkt));
+
+	ndata.buf = finfo;
+
+	ndata.q_no = iq_no;
+
+	if (octnet_iq_is_full(oct, ndata.q_no)) {
+		/* defer sending if queue is full */
+		netif_info(lio, tx_err, lio->netdev, "Transmit failed iq:%d full\n",
+			   ndata.q_no);
+		stats->tx_iq_busy++;
+		return NETDEV_TX_BUSY;
+	}
+
+	ndata.datasize = skb->len;
+
+	cmdsetup.u64 = 0;
+	cmdsetup.s.iq_no = iq_no;
+
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
+		if (skb->encapsulation) {
+			cmdsetup.s.tnl_csum = 1;
+			stats->tx_vxlan++;
+		} else {
+			cmdsetup.s.transport_csum = 1;
+		}
+	}
+	if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)) {
+		skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
+		cmdsetup.s.timestamp = 1;
+	}
+
+	if (!skb_shinfo(skb)->nr_frags) {
+		cmdsetup.s.u.datasize = skb->len;
+		octnet_prepare_pci_cmd(oct, &ndata.cmd, &cmdsetup, tag);
+		/* Offload checksum calculation for TCP/UDP packets */
+		dptr = dma_map_single(&oct->pci_dev->dev,
+				      skb->data,
+				      skb->len,
+				      DMA_TO_DEVICE);
+		if (dma_mapping_error(&oct->pci_dev->dev, dptr)) {
+			dev_err(&oct->pci_dev->dev, "%s DMA mapping error 1\n",
+				__func__);
+			return NETDEV_TX_BUSY;
+		}
+
+		ndata.cmd.cmd3.dptr = dptr;
+		finfo->dptr = dptr;
+		ndata.reqtype = REQTYPE_NORESP_NET;
+
+	} else {
+		skb_frag_t *frag;
+		struct octnic_gather *g;
+		int i, frags;
+
+		spin_lock(&lio->glist_lock[q_idx]);
+		g = (struct octnic_gather *)
+			lio_list_delete_head(&lio->glist[q_idx]);
+		spin_unlock(&lio->glist_lock[q_idx]);
+
+		if (!g) {
+			netif_info(lio, tx_err, lio->netdev,
+				   "Transmit scatter gather: glist null!\n");
+			goto lio_xmit_failed;
+		}
+
+		cmdsetup.s.gather = 1;
+		cmdsetup.s.u.gatherptrs = (skb_shinfo(skb)->nr_frags + 1);
+		octnet_prepare_pci_cmd(oct, &ndata.cmd, &cmdsetup, tag);
+
+		memset(g->sg, 0, g->sg_size);
+
+		g->sg[0].ptr[0] = dma_map_single(&oct->pci_dev->dev,
+						 skb->data,
+						 (skb->len - skb->data_len),
+						 DMA_TO_DEVICE);
+		if (dma_mapping_error(&oct->pci_dev->dev, g->sg[0].ptr[0])) {
+			dev_err(&oct->pci_dev->dev, "%s DMA mapping error 2\n",
+				__func__);
+			return NETDEV_TX_BUSY;
+		}
+		add_sg_size(&g->sg[0], (skb->len - skb->data_len), 0);
+
+		frags = skb_shinfo(skb)->nr_frags;
+		i = 1;
+		while (frags--) {
+			frag = &skb_shinfo(skb)->frags[i - 1];
+
+			g->sg[(i >> 2)].ptr[(i & 3)] =
+				skb_frag_dma_map(&oct->pci_dev->dev,
+						 frag, 0, skb_frag_size(frag),
+						 DMA_TO_DEVICE);
+			if (dma_mapping_error(&oct->pci_dev->dev,
+					      g->sg[i >> 2].ptr[i & 3])) {
+				dma_unmap_single(&oct->pci_dev->dev,
+						 g->sg[0].ptr[0],
+						 skb->len - skb->data_len,
+						 DMA_TO_DEVICE);
+				for (j = 1; j < i; j++) {
+					frag = &skb_shinfo(skb)->frags[j - 1];
+					dma_unmap_page(&oct->pci_dev->dev,
+						       g->sg[j >> 2].ptr[j & 3],
+						       skb_frag_size(frag),
+						       DMA_TO_DEVICE);
+				}
+				dev_err(&oct->pci_dev->dev, "%s DMA mapping error 3\n",
+					__func__);
+				return NETDEV_TX_BUSY;
+			}
+
+			add_sg_size(&g->sg[(i >> 2)], skb_frag_size(frag),
+				    (i & 3));
+			i++;
+		}
+
+		dptr = g->sg_dma_ptr;
+
+		ndata.cmd.cmd3.dptr = dptr;
+		finfo->dptr = dptr;
+		finfo->g = g;
+
+		ndata.reqtype = REQTYPE_NORESP_NET_SG;
+	}
+
+	irh = (struct octeon_instr_irh *)&ndata.cmd.cmd3.irh;
+	tx_info = (union tx_info *)&ndata.cmd.cmd3.ossp[0];
+
+	if (skb_shinfo(skb)->gso_size) {
+		tx_info->s.gso_size = skb_shinfo(skb)->gso_size;
+		tx_info->s.gso_segs = skb_shinfo(skb)->gso_segs;
+	}
+
+	/* HW insert VLAN tag */
+	if (skb_vlan_tag_present(skb)) {
+		irh->priority = skb_vlan_tag_get(skb) >> VLAN_PRIO_SHIFT;
+		irh->vlan = skb_vlan_tag_get(skb) & VLAN_VID_MASK;
+	}
+
+	xmit_more = netdev_xmit_more();
+
+	if (unlikely(cmdsetup.s.timestamp))
+		status = send_nic_timestamp_pkt(oct, &ndata, finfo, xmit_more);
+	else
+		status = octnet_send_nic_data_pkt(oct, &ndata, xmit_more);
+	if (status == IQ_SEND_FAILED)
+		goto lio_xmit_failed;
+
+	netif_info(lio, tx_queued, lio->netdev, "Transmit queued successfully\n");
+
+	if (status == IQ_SEND_STOP) {
+		dev_err(&oct->pci_dev->dev, "Rcvd IQ_SEND_STOP signal; stopping IQ-%d\n",
+			iq_no);
+		netif_stop_subqueue(netdev, q_idx);
+	}
+
+	netif_trans_update(netdev);
+
+	if (tx_info->s.gso_segs)
+		stats->tx_done += tx_info->s.gso_segs;
+	else
+		stats->tx_done++;
+	stats->tx_tot_bytes += ndata.datasize;
+
+	return NETDEV_TX_OK;
+
+lio_xmit_failed:
+	stats->tx_dropped++;
+	netif_info(lio, tx_err, lio->netdev, "IQ%d Transmit dropped:%llu\n",
+		   iq_no, stats->tx_dropped);
+	if (dptr)
+		dma_unmap_single(&oct->pci_dev->dev, dptr,
+				 ndata.datasize, DMA_TO_DEVICE);
+
+	octeon_ring_doorbell_locked(oct, iq_no);
+
+	tx_buffer_free(skb);
+	return NETDEV_TX_OK;
+}
+
+/**
+ * liquidio_tx_timeout - Network device Tx timeout
+ * @netdev: pointer to network device
+ * @txqueue: index of the hung transmit queue
+ */
+static void liquidio_tx_timeout(struct net_device *netdev, unsigned int txqueue)
+{
+	struct lio *lio;
+
+	lio = GET_LIO(netdev);
+
+	netif_info(lio, tx_err, lio->netdev,
+		   "Transmit timeout tx_dropped:%ld, waking up queues now!!\n",
+		   netdev->stats.tx_dropped);
+	netif_trans_update(netdev);
+	wake_txqs(netdev);
+}
+
+static int
+liquidio_vlan_rx_add_vid(struct net_device *netdev,
+			 __be16 proto __attribute__((unused)), u16 vid)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_ADD_VLAN_FILTER;
+	nctrl.ncmd.s.param1 = vid;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "Add VLAN filter failed in core (ret: 0x%x)\n",
+			ret);
+		return -EPERM;
+	}
+
+	return 0;
+}
+
+static int
+liquidio_vlan_rx_kill_vid(struct net_device *netdev,
+			  __be16 proto __attribute__((unused)), u16 vid)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = OCTNET_CMD_DEL_VLAN_FILTER;
+	nctrl.ncmd.s.param1 = vid;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "Del VLAN filter failed in core (ret: 0x%x)\n",
+			ret);
+		if (ret > 0)
+			ret = -EIO;
+	}
+	return ret;
+}
+
+/** Sending command to enable/disable RX checksum offload
+ * @param netdev                pointer to network device
+ * @param command               OCTNET_CMD_TNL_RX_CSUM_CTL
+ * @param rx_cmd_bit            OCTNET_CMD_RXCSUM_ENABLE/
+ *                              OCTNET_CMD_RXCSUM_DISABLE
+ * @returns                     SUCCESS or FAILURE
+ */
+static int liquidio_set_rxcsum_command(struct net_device *netdev, int command,
+				       u8 rx_cmd)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = command;
+	nctrl.ncmd.s.param1 = rx_cmd;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "DEVFLAGS RXCSUM change failed in core (ret:0x%x)\n",
+			ret);
+		if (ret > 0)
+			ret = -EIO;
+	}
+	return ret;
+}
+
+/** Sending command to add/delete VxLAN UDP port to firmware
+ * @param netdev                pointer to network device
+ * @param command               OCTNET_CMD_VXLAN_PORT_CONFIG
+ * @param vxlan_port            VxLAN port to be added or deleted
+ * @param vxlan_cmd_bit         OCTNET_CMD_VXLAN_PORT_ADD,
+ *                              OCTNET_CMD_VXLAN_PORT_DEL
+ * @returns                     SUCCESS or FAILURE
+ */
+static int liquidio_vxlan_port_command(struct net_device *netdev, int command,
+				       u16 vxlan_port, u8 vxlan_cmd_bit)
+{
+	struct lio *lio = GET_LIO(netdev);
+	struct octeon_device *oct = lio->oct_dev;
+	struct octnic_ctrl_pkt nctrl;
+	int ret = 0;
+
+	memset(&nctrl, 0, sizeof(struct octnic_ctrl_pkt));
+
+	nctrl.ncmd.u64 = 0;
+	nctrl.ncmd.s.cmd = command;
+	nctrl.ncmd.s.more = vxlan_cmd_bit;
+	nctrl.ncmd.s.param1 = vxlan_port;
+	nctrl.iq_no = lio->linfo.txpciq[0].s.q_no;
+	nctrl.netpndev = (u64)netdev;
+	nctrl.cb_fn = liquidio_link_ctrl_cmd_completion;
+
+	ret = octnet_send_nic_ctrl_pkt(lio->oct_dev, &nctrl);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev,
+			"DEVFLAGS VxLAN port add/delete failed in core (ret : 0x%x)\n",
+			ret);
+		if (ret > 0)
+			ret = -EIO;
+	}
+	return ret;
+}
+
+static int liquidio_udp_tunnel_set_port(struct net_device *netdev,
+					unsigned int table, unsigned int entry,
+					struct udp_tunnel_info *ti)
+{
+	return liquidio_vxlan_port_command(netdev,
+					   OCTNET_CMD_VXLAN_PORT_CONFIG,
+					   htons(ti->port),
+					   OCTNET_CMD_VXLAN_PORT_ADD);
+}
+
+static int liquidio_udp_tunnel_unset_port(struct net_device *netdev,
+					  unsigned int table,
+					  unsigned int entry,
+					  struct udp_tunnel_info *ti)
+{
+	return liquidio_vxlan_port_command(netdev,
+					   OCTNET_CMD_VXLAN_PORT_CONFIG,
+					   htons(ti->port),
+					   OCTNET_CMD_VXLAN_PORT_DEL);
+}
+
+static const struct udp_tunnel_nic_info liquidio_udp_tunnels = {
+	.set_port	= liquidio_udp_tunnel_set_port,
+	.unset_port	= liquidio_udp_tunnel_unset_port,
+	.tables		= {
+		{ .n_entries = 1024, .tunnel_types = UDP_TUNNEL_TYPE_VXLAN, },
+	},
+};
+
+/** \brief Net device fix features
+ * @param netdev  pointer to network device
+ * @param request features requested
+ * @returns updated features list
+ */
+static netdev_features_t liquidio_fix_features(struct net_device *netdev,
+					       netdev_features_t request)
+{
+	struct lio *lio = netdev_priv(netdev);
+
+	if ((request & NETIF_F_RXCSUM) &&
+	    !(lio->dev_capability & NETIF_F_RXCSUM))
+		request &= ~NETIF_F_RXCSUM;
+
+	if ((request & NETIF_F_HW_CSUM) &&
+	    !(lio->dev_capability & NETIF_F_HW_CSUM))
+		request &= ~NETIF_F_HW_CSUM;
+
+	if ((request & NETIF_F_TSO) && !(lio->dev_capability & NETIF_F_TSO))
+		request &= ~NETIF_F_TSO;
+
+	if ((request & NETIF_F_TSO6) && !(lio->dev_capability & NETIF_F_TSO6))
+		request &= ~NETIF_F_TSO6;
+
+	if ((request & NETIF_F_LRO) && !(lio->dev_capability & NETIF_F_LRO))
+		request &= ~NETIF_F_LRO;
+
+	/* Disable LRO if RXCSUM is off */
+	if (!(request & NETIF_F_RXCSUM) && (netdev->features & NETIF_F_LRO) &&
+	    (lio->dev_capability & NETIF_F_LRO))
+		request &= ~NETIF_F_LRO;
+
+	return request;
+}
+
+/** \brief Net device set features
+ * @param netdev  pointer to network device
+ * @param features features to enable/disable
+ */
+static int liquidio_set_features(struct net_device *netdev,
+				 netdev_features_t features)
+{
+	struct lio *lio = netdev_priv(netdev);
+
+	if (!((netdev->features ^ features) & NETIF_F_LRO))
+		return 0;
+
+	if ((features & NETIF_F_LRO) && (lio->dev_capability & NETIF_F_LRO))
+		liquidio_set_feature(netdev, OCTNET_CMD_LRO_ENABLE,
+				     OCTNIC_LROIPV4 | OCTNIC_LROIPV6);
+	else if (!(features & NETIF_F_LRO) &&
+		 (lio->dev_capability & NETIF_F_LRO))
+		liquidio_set_feature(netdev, OCTNET_CMD_LRO_DISABLE,
+				     OCTNIC_LROIPV4 | OCTNIC_LROIPV6);
+	if (!(netdev->features & NETIF_F_RXCSUM) &&
+	    (lio->enc_dev_capability & NETIF_F_RXCSUM) &&
+	    (features & NETIF_F_RXCSUM))
+		liquidio_set_rxcsum_command(netdev, OCTNET_CMD_TNL_RX_CSUM_CTL,
+					    OCTNET_CMD_RXCSUM_ENABLE);
+	else if ((netdev->features & NETIF_F_RXCSUM) &&
+		 (lio->enc_dev_capability & NETIF_F_RXCSUM) &&
+		 !(features & NETIF_F_RXCSUM))
+		liquidio_set_rxcsum_command(netdev, OCTNET_CMD_TNL_RX_CSUM_CTL,
+					    OCTNET_CMD_RXCSUM_DISABLE);
+
+	return 0;
+}
+
+static const struct net_device_ops lionetdevops = {
+	.ndo_open		= liquidio_open,
+	.ndo_stop		= liquidio_stop,
+	.ndo_start_xmit		= liquidio_xmit,
+	.ndo_get_stats64	= liquidio_get_stats64,
+	.ndo_set_mac_address	= liquidio_set_mac,
+	.ndo_set_rx_mode	= liquidio_set_mcast_list,
+	.ndo_tx_timeout		= liquidio_tx_timeout,
+	.ndo_vlan_rx_add_vid    = liquidio_vlan_rx_add_vid,
+	.ndo_vlan_rx_kill_vid   = liquidio_vlan_rx_kill_vid,
+	.ndo_change_mtu		= liquidio_change_mtu,
+	.ndo_eth_ioctl		= liquidio_ioctl,
+	.ndo_fix_features	= liquidio_fix_features,
+	.ndo_set_features	= liquidio_set_features,
+};
+
+static int lio_nic_info(struct octeon_recv_info *recv_info, void *buf)
+{
+	struct octeon_device *oct = (struct octeon_device *)buf;
+	struct octeon_recv_pkt *recv_pkt = recv_info->recv_pkt;
+	union oct_link_status *ls;
+	int gmxport = 0;
+	int i;
+
+	if (recv_pkt->buffer_size[0] != (sizeof(*ls) + OCT_DROQ_INFO_SIZE)) {
+		dev_err(&oct->pci_dev->dev, "Malformed NIC_INFO, len=%d, ifidx=%d\n",
+			recv_pkt->buffer_size[0],
+			recv_pkt->rh.r_nic_info.gmxport);
+		goto nic_info_err;
+	}
+
+	gmxport = recv_pkt->rh.r_nic_info.gmxport;
+	ls = (union oct_link_status *)(get_rbd(recv_pkt->buffer_ptr[0]) +
+		OCT_DROQ_INFO_SIZE);
+
+	octeon_swap_8B_data((u64 *)ls, (sizeof(union oct_link_status)) >> 3);
+
+	for (i = 0; i < oct->ifcount; i++) {
+		if (oct->props[i].gmxport == gmxport) {
+			update_link_status(oct->props[i].netdev, ls);
+			break;
+		}
+	}
+
+nic_info_err:
+	for (i = 0; i < recv_pkt->buffer_count; i++)
+		recv_buffer_free(recv_pkt->buffer_ptr[i]);
+	octeon_free_recv_info(recv_info);
+	return 0;
+}
+
+/**
+ * setup_nic_devices - Setup network interfaces
+ * @octeon_dev:  octeon device
+ *
+ * Called during init time for each device. It assumes the NIC
+ * is already up and running.  The link information for each
+ * interface is passed in link_info.
+ */
+static int setup_nic_devices(struct octeon_device *octeon_dev)
+{
+	int retval, num_iqueues, num_oqueues;
+	u32 resp_size, data_size;
+	struct liquidio_if_cfg_resp *resp;
+	struct octeon_soft_command *sc;
+	union oct_nic_if_cfg if_cfg;
+	struct octdev_props *props;
+	struct net_device *netdev;
+	struct lio_version *vdata;
+	struct lio *lio = NULL;
+	u8 mac[ETH_ALEN], i, j;
+	u32 ifidx_or_pfnum;
+
+	ifidx_or_pfnum = octeon_dev->pf_num;
+
+	/* This is to handle link status changes */
+	octeon_register_dispatch_fn(octeon_dev, OPCODE_NIC, OPCODE_NIC_INFO,
+				    lio_nic_info, octeon_dev);
+
+	/* REQTYPE_RESP_NET and REQTYPE_SOFT_COMMAND do not have free functions.
+	 * They are handled directly.
+	 */
+	octeon_register_reqtype_free_fn(octeon_dev, REQTYPE_NORESP_NET,
+					free_netbuf);
+
+	octeon_register_reqtype_free_fn(octeon_dev, REQTYPE_NORESP_NET_SG,
+					free_netsgbuf);
+
+	octeon_register_reqtype_free_fn(octeon_dev, REQTYPE_RESP_NET_SG,
+					free_netsgbuf_with_resp);
+
+	for (i = 0; i < octeon_dev->ifcount; i++) {
+		resp_size = sizeof(struct liquidio_if_cfg_resp);
+		data_size = sizeof(struct lio_version);
+		sc = (struct octeon_soft_command *)
+			octeon_alloc_soft_command(octeon_dev, data_size,
+						  resp_size, 0);
+		resp = (struct liquidio_if_cfg_resp *)sc->virtrptr;
+		vdata = (struct lio_version *)sc->virtdptr;
+
+		*((u64 *)vdata) = 0;
+		vdata->major = cpu_to_be16(LIQUIDIO_BASE_MAJOR_VERSION);
+		vdata->minor = cpu_to_be16(LIQUIDIO_BASE_MINOR_VERSION);
+		vdata->micro = cpu_to_be16(LIQUIDIO_BASE_MICRO_VERSION);
+
+		if_cfg.u64 = 0;
+
+		if_cfg.s.num_iqueues = octeon_dev->sriov_info.rings_per_vf;
+		if_cfg.s.num_oqueues = octeon_dev->sriov_info.rings_per_vf;
+		if_cfg.s.base_queue = 0;
+
+		sc->iq_no = 0;
+
+		octeon_prepare_soft_command(octeon_dev, sc, OPCODE_NIC,
+					    OPCODE_NIC_IF_CFG, 0, if_cfg.u64,
+					    0);
+
+		init_completion(&sc->complete);
+		sc->sc_status = OCTEON_REQUEST_PENDING;
+
+		retval = octeon_send_soft_command(octeon_dev, sc);
+		if (retval == IQ_SEND_FAILED) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"iq/oq config failed status: %x\n", retval);
+			/* Soft instr is freed by driver in case of failure. */
+			octeon_free_soft_command(octeon_dev, sc);
+			return(-EIO);
+		}
+
+		/* Sleep on a wait queue till the cond flag indicates that the
+		 * response arrived or timed-out.
+		 */
+		retval = wait_for_sc_completion_timeout(octeon_dev, sc, 0);
+		if (retval)
+			return retval;
+
+		retval = resp->status;
+		if (retval) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"iq/oq config failed, retval = %d\n", retval);
+			WRITE_ONCE(sc->caller_is_done, true);
+			return -EIO;
+		}
+
+		snprintf(octeon_dev->fw_info.liquidio_firmware_version,
+			 32, "%s",
+			 resp->cfg_info.liquidio_firmware_version);
+
+		octeon_swap_8B_data((u64 *)(&resp->cfg_info),
+				    (sizeof(struct liquidio_if_cfg_info)) >> 3);
+
+		num_iqueues = hweight64(resp->cfg_info.iqmask);
+		num_oqueues = hweight64(resp->cfg_info.oqmask);
+
+		if (!(num_iqueues) || !(num_oqueues)) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"Got bad iqueues (%016llx) or oqueues (%016llx) from firmware.\n",
+				resp->cfg_info.iqmask, resp->cfg_info.oqmask);
+			WRITE_ONCE(sc->caller_is_done, true);
+			goto setup_nic_dev_done;
+		}
+		dev_dbg(&octeon_dev->pci_dev->dev,
+			"interface %d, iqmask %016llx, oqmask %016llx, numiqueues %d, numoqueues %d\n",
+			i, resp->cfg_info.iqmask, resp->cfg_info.oqmask,
+			num_iqueues, num_oqueues);
+
+		netdev = alloc_etherdev_mq(LIO_SIZE, num_iqueues);
+
+		if (!netdev) {
+			dev_err(&octeon_dev->pci_dev->dev, "Device allocation failed\n");
+			WRITE_ONCE(sc->caller_is_done, true);
+			goto setup_nic_dev_done;
+		}
+
+		SET_NETDEV_DEV(netdev, &octeon_dev->pci_dev->dev);
+
+		/* Associate the routines that will handle different
+		 * netdev tasks.
+		 */
+		netdev->netdev_ops = &lionetdevops;
+
+		lio = GET_LIO(netdev);
+
+		memset(lio, 0, sizeof(struct lio));
+
+		lio->ifidx = ifidx_or_pfnum;
+
+		props = &octeon_dev->props[i];
+		props->gmxport = resp->cfg_info.linfo.gmxport;
+		props->netdev = netdev;
+
+		lio->linfo.num_rxpciq = num_oqueues;
+		lio->linfo.num_txpciq = num_iqueues;
+
+		for (j = 0; j < num_oqueues; j++) {
+			lio->linfo.rxpciq[j].u64 =
+			    resp->cfg_info.linfo.rxpciq[j].u64;
+		}
+		for (j = 0; j < num_iqueues; j++) {
+			lio->linfo.txpciq[j].u64 =
+			    resp->cfg_info.linfo.txpciq[j].u64;
+		}
+
+		lio->linfo.hw_addr = resp->cfg_info.linfo.hw_addr;
+		lio->linfo.gmxport = resp->cfg_info.linfo.gmxport;
+		lio->linfo.link.u64 = resp->cfg_info.linfo.link.u64;
+		lio->linfo.macaddr_is_admin_asgnd =
+			resp->cfg_info.linfo.macaddr_is_admin_asgnd;
+		lio->linfo.macaddr_spoofchk =
+			resp->cfg_info.linfo.macaddr_spoofchk;
+
+		lio->msg_enable = netif_msg_init(debug, DEFAULT_MSG_ENABLE);
+
+		lio->dev_capability = NETIF_F_HIGHDMA
+				      | NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM
+				      | NETIF_F_SG | NETIF_F_RXCSUM
+				      | NETIF_F_TSO | NETIF_F_TSO6
+				      | NETIF_F_GRO
+				      | NETIF_F_LRO;
+		netif_set_tso_max_size(netdev, OCTNIC_GSO_MAX_SIZE);
+
+		/* Copy of transmit encapsulation capabilities:
+		 * TSO, TSO6, Checksums for this device
+		 */
+		lio->enc_dev_capability = NETIF_F_IP_CSUM
+					  | NETIF_F_IPV6_CSUM
+					  | NETIF_F_GSO_UDP_TUNNEL
+					  | NETIF_F_HW_CSUM | NETIF_F_SG
+					  | NETIF_F_RXCSUM
+					  | NETIF_F_TSO | NETIF_F_TSO6
+					  | NETIF_F_LRO;
+
+		netdev->hw_enc_features =
+		    (lio->enc_dev_capability & ~NETIF_F_LRO);
+		netdev->udp_tunnel_nic_info = &liquidio_udp_tunnels;
+
+		netdev->vlan_features = lio->dev_capability;
+		/* Add any unchangeable hw features */
+		lio->dev_capability |= NETIF_F_HW_VLAN_CTAG_FILTER |
+				       NETIF_F_HW_VLAN_CTAG_RX |
+				       NETIF_F_HW_VLAN_CTAG_TX;
+
+		netdev->features = (lio->dev_capability & ~NETIF_F_LRO);
+
+		netdev->hw_features = lio->dev_capability;
+		netdev->hw_features &= ~NETIF_F_HW_VLAN_CTAG_RX;
+
+		/* MTU range: 68 - 16000 */
+		netdev->min_mtu = LIO_MIN_MTU_SIZE;
+		netdev->max_mtu = LIO_MAX_MTU_SIZE;
+
+		WRITE_ONCE(sc->caller_is_done, true);
+
+		/* Point to the  properties for octeon device to which this
+		 * interface belongs.
+		 */
+		lio->oct_dev = octeon_dev;
+		lio->octprops = props;
+		lio->netdev = netdev;
+
+		dev_dbg(&octeon_dev->pci_dev->dev,
+			"if%d gmx: %d hw_addr: 0x%llx\n", i,
+			lio->linfo.gmxport, CVM_CAST64(lio->linfo.hw_addr));
+
+		/* 64-bit swap required on LE machines */
+		octeon_swap_8B_data(&lio->linfo.hw_addr, 1);
+		for (j = 0; j < ETH_ALEN; j++)
+			mac[j] = *((u8 *)(((u8 *)&lio->linfo.hw_addr) + 2 + j));
+
+		/* Copy MAC Address to OS network device structure */
+		eth_hw_addr_set(netdev, mac);
+
+		if (liquidio_setup_io_queues(octeon_dev, i,
+					     lio->linfo.num_txpciq,
+					     lio->linfo.num_rxpciq)) {
+			dev_err(&octeon_dev->pci_dev->dev, "I/O queues creation failed\n");
+			goto setup_nic_dev_free;
+		}
+
+		ifstate_set(lio, LIO_IFSTATE_DROQ_OPS);
+
+		/* For VFs, enable Octeon device interrupts here,
+		 * as this is contingent upon IO queue setup
+		 */
+		octeon_dev->fn_list.enable_interrupt(octeon_dev,
+						     OCTEON_ALL_INTR);
+
+		/* By default all interfaces on a single Octeon uses the same
+		 * tx and rx queues
+		 */
+		lio->txq = lio->linfo.txpciq[0].s.q_no;
+		lio->rxq = lio->linfo.rxpciq[0].s.q_no;
+
+		lio->tx_qsize = octeon_get_tx_qsize(octeon_dev, lio->txq);
+		lio->rx_qsize = octeon_get_rx_qsize(octeon_dev, lio->rxq);
+
+		if (lio_setup_glists(octeon_dev, lio, num_iqueues)) {
+			dev_err(&octeon_dev->pci_dev->dev,
+				"Gather list allocation failed\n");
+			goto setup_nic_dev_free;
+		}
+
+		/* Register ethtool support */
+		liquidio_set_ethtool_ops(netdev);
+		if (lio->oct_dev->chip_id == OCTEON_CN23XX_VF_VID)
+			octeon_dev->priv_flags = OCT_PRIV_FLAG_DEFAULT;
+		else
+			octeon_dev->priv_flags = 0x0;
+
+		if (netdev->features & NETIF_F_LRO)
+			liquidio_set_feature(netdev, OCTNET_CMD_LRO_ENABLE,
+					     OCTNIC_LROIPV4 | OCTNIC_LROIPV6);
+
+		if (setup_link_status_change_wq(netdev))
+			goto setup_nic_dev_free;
+
+		if (setup_rx_oom_poll_fn(netdev))
+			goto setup_nic_dev_free;
+
+		/* Register the network device with the OS */
+		if (register_netdev(netdev)) {
+			dev_err(&octeon_dev->pci_dev->dev, "Device registration failed\n");
+			goto setup_nic_dev_free;
+		}
+
+		dev_dbg(&octeon_dev->pci_dev->dev,
+			"Setup NIC ifidx:%d mac:%02x%02x%02x%02x%02x%02x\n",
+			i, mac[0], mac[1], mac[2], mac[3], mac[4], mac[5]);
+		netif_carrier_off(netdev);
+		lio->link_changes++;
+
+		ifstate_set(lio, LIO_IFSTATE_REGISTERED);
+
+		/* Sending command to firmware to enable Rx checksum offload
+		 * by default at the time of setup of Liquidio driver for
+		 * this device
+		 */
+		liquidio_set_rxcsum_command(netdev, OCTNET_CMD_TNL_RX_CSUM_CTL,
+					    OCTNET_CMD_RXCSUM_ENABLE);
+		liquidio_set_feature(netdev, OCTNET_CMD_TNL_TX_CSUM_CTL,
+				     OCTNET_CMD_TXCSUM_ENABLE);
+
+		dev_dbg(&octeon_dev->pci_dev->dev,
+			"NIC ifidx:%d Setup successful\n", i);
+
+		octeon_dev->no_speed_setting = 1;
+	}
+
+	return 0;
+
+setup_nic_dev_free:
+
+	while (i--) {
+		dev_err(&octeon_dev->pci_dev->dev,
+			"NIC ifidx:%d Setup failed\n", i);
+		liquidio_destroy_nic_device(octeon_dev, i);
+	}
+
+setup_nic_dev_done:
+
+	return -ENODEV;
+}
+
+/**
+ * liquidio_init_nic_module - initialize the NIC
+ * @oct: octeon device
+ *
+ * This initialization routine is called once the Octeon device application is
+ * up and running
+ */
+static int liquidio_init_nic_module(struct octeon_device *oct)
+{
+	int num_nic_ports = 1;
+	int i, retval = 0;
+
+	dev_dbg(&oct->pci_dev->dev, "Initializing network interfaces\n");
+
+	/* only default iq and oq were initialized
+	 * initialize the rest as well run port_config command for each port
+	 */
+	oct->ifcount = num_nic_ports;
+	memset(oct->props, 0,
+	       sizeof(struct octdev_props) * num_nic_ports);
+
+	for (i = 0; i < MAX_OCTEON_LINKS; i++)
+		oct->props[i].gmxport = -1;
+
+	retval = setup_nic_devices(oct);
+	if (retval) {
+		dev_err(&oct->pci_dev->dev, "Setup NIC devices failed\n");
+		goto octnet_init_failure;
+	}
+
+	dev_dbg(&oct->pci_dev->dev, "Network interfaces ready\n");
+
+	return retval;
+
+octnet_init_failure:
+
+	oct->ifcount = 0;
+
+	return retval;
+}
+
+/**
+ * octeon_device_init - Device initialization for each Octeon device that is probed
+ * @oct:  octeon device
+ */
+static int octeon_device_init(struct octeon_device *oct)
+{
+	u32 rev_id;
+	int j;
+
+	atomic_set(&oct->status, OCT_DEV_BEGIN_STATE);
+
+	/* Enable access to the octeon device and make its DMA capability
+	 * known to the OS.
+	 */
+	if (octeon_pci_os_setup(oct))
+		return 1;
+	atomic_set(&oct->status, OCT_DEV_PCI_ENABLE_DONE);
+
+	oct->chip_id = OCTEON_CN23XX_VF_VID;
+	pci_read_config_dword(oct->pci_dev, 8, &rev_id);
+	oct->rev_id = rev_id & 0xff;
+
+	if (cn23xx_setup_octeon_vf_device(oct))
+		return 1;
+
+	atomic_set(&oct->status, OCT_DEV_PCI_MAP_DONE);
+
+	oct->app_mode = CVM_DRV_NIC_APP;
+
+	/* Initialize the dispatch mechanism used to push packets arriving on
+	 * Octeon Output queues.
+	 */
+	if (octeon_init_dispatch_list(oct))
+		return 1;
+
+	atomic_set(&oct->status, OCT_DEV_DISPATCH_INIT_DONE);
+
+	if (octeon_set_io_queues_off(oct)) {
+		dev_err(&oct->pci_dev->dev, "setting io queues off failed\n");
+		return 1;
+	}
+
+	if (oct->fn_list.setup_device_regs(oct)) {
+		dev_err(&oct->pci_dev->dev, "device registers configuration failed\n");
+		return 1;
+	}
+
+	/* Initialize soft command buffer pool */
+	if (octeon_setup_sc_buffer_pool(oct)) {
+		dev_err(&oct->pci_dev->dev, "sc buffer pool allocation failed\n");
+		return 1;
+	}
+	atomic_set(&oct->status, OCT_DEV_SC_BUFF_POOL_INIT_DONE);
+
+	/* Setup the data structures that manage this Octeon's Input queues. */
+	if (octeon_setup_instr_queues(oct)) {
+		dev_err(&oct->pci_dev->dev, "instruction queue initialization failed\n");
+		return 1;
+	}
+	atomic_set(&oct->status, OCT_DEV_INSTR_QUEUE_INIT_DONE);
+
+	/* Initialize lists to manage the requests of different types that
+	 * arrive from user & kernel applications for this octeon device.
+	 */
+	if (octeon_setup_response_list(oct)) {
+		dev_err(&oct->pci_dev->dev, "Response list allocation failed\n");
+		return 1;
+	}
+	atomic_set(&oct->status, OCT_DEV_RESP_LIST_INIT_DONE);
+
+	if (octeon_setup_output_queues(oct)) {
+		dev_err(&oct->pci_dev->dev, "Output queue initialization failed\n");
+		return 1;
+	}
+	atomic_set(&oct->status, OCT_DEV_DROQ_INIT_DONE);
+
+	if (oct->fn_list.setup_mbox(oct)) {
+		dev_err(&oct->pci_dev->dev, "Mailbox setup failed\n");
+		return 1;
+	}
+	atomic_set(&oct->status, OCT_DEV_MBOX_SETUP_DONE);
+
+	if (octeon_allocate_ioq_vector(oct, oct->sriov_info.rings_per_vf)) {
+		dev_err(&oct->pci_dev->dev, "ioq vector allocation failed\n");
+		return 1;
+	}
+	atomic_set(&oct->status, OCT_DEV_MSIX_ALLOC_VECTOR_DONE);
+
+	dev_info(&oct->pci_dev->dev, "OCTEON_CN23XX VF: %d ioqs\n",
+		 oct->sriov_info.rings_per_vf);
+
+	/* Setup the interrupt handler and record the INT SUM register address*/
+	if (octeon_setup_interrupt(oct, oct->sriov_info.rings_per_vf))
+		return 1;
+
+	atomic_set(&oct->status, OCT_DEV_INTR_SET_DONE);
+
+	/* ***************************************************************
+	 * The interrupts need to be enabled for the PF<-->VF handshake.
+	 * They are [re]-enabled after the PF<-->VF handshake so that the
+	 * correct OQ tick value is used (i.e. the value retrieved from
+	 * the PF as part of the handshake).
+	 */
+
+	/* Enable Octeon device interrupts */
+	oct->fn_list.enable_interrupt(oct, OCTEON_ALL_INTR);
+
+	if (cn23xx_octeon_pfvf_handshake(oct))
+		return 1;
+
+	/* Here we [re]-enable the interrupts so that the correct OQ tick value
+	 * is used (i.e. the value that was retrieved during the handshake)
+	 */
+
+	/* Enable Octeon device interrupts */
+	oct->fn_list.enable_interrupt(oct, OCTEON_ALL_INTR);
+	/* *************************************************************** */
+
+	/* Enable the input and output queues for this Octeon device */
+	if (oct->fn_list.enable_io_queues(oct)) {
+		dev_err(&oct->pci_dev->dev, "enabling io queues failed\n");
+		return 1;
+	}
+
+	atomic_set(&oct->status, OCT_DEV_IO_QUEUES_DONE);
+
+	atomic_set(&oct->status, OCT_DEV_HOST_OK);
+
+	/* Send Credit for Octeon Output queues. Credits are always sent after
+	 * the output queue is enabled.
+	 */
+	for (j = 0; j < oct->num_oqs; j++)
+		writel(oct->droq[j]->max_count, oct->droq[j]->pkts_credit_reg);
+
+	/* Packets can start arriving on the output queues from this point. */
+
+	atomic_set(&oct->status, OCT_DEV_CORE_OK);
+
+	atomic_set(&oct->status, OCT_DEV_RUNNING);
+
+	if (liquidio_init_nic_module(oct))
+		return 1;
+
+	return 0;
+}
+
+static int __init liquidio_vf_init(void)
+{
+	octeon_init_device_list(0);
+	return pci_register_driver(&liquidio_vf_pci_driver);
+}
+
+static void __exit liquidio_vf_exit(void)
+{
+	pci_unregister_driver(&liquidio_vf_pci_driver);
+
+	pr_info("LiquidIO_VF network module is now unloaded\n");
+}
+
+module_init(liquidio_vf_init);
+module_exit(liquidio_vf_exit);
diff --git a/drivers/net/ethernet/cavium/liquidio/lio_vf_rep.c b/drivers/net/ethernet/cavium/liquidio/lio_vf_rep.c
new file mode 100644
index 000000000..600de587d
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/lio_vf_rep.c
@@ -0,0 +1,672 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2017 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/pci.h>
+#include <linux/if_vlan.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_nic.h"
+#include "octeon_main.h"
+#include "octeon_network.h"
+#include "lio_vf_rep.h"
+
+static int lio_vf_rep_open(struct net_device *ndev);
+static int lio_vf_rep_stop(struct net_device *ndev);
+static netdev_tx_t lio_vf_rep_pkt_xmit(struct sk_buff *skb,
+				       struct net_device *ndev);
+static void lio_vf_rep_tx_timeout(struct net_device *netdev, unsigned int txqueue);
+static int lio_vf_rep_phys_port_name(struct net_device *dev,
+				     char *buf, size_t len);
+static void lio_vf_rep_get_stats64(struct net_device *dev,
+				   struct rtnl_link_stats64 *stats64);
+static int lio_vf_rep_change_mtu(struct net_device *ndev, int new_mtu);
+static int lio_vf_get_port_parent_id(struct net_device *dev,
+				     struct netdev_phys_item_id *ppid);
+
+static const struct net_device_ops lio_vf_rep_ndev_ops = {
+	.ndo_open = lio_vf_rep_open,
+	.ndo_stop = lio_vf_rep_stop,
+	.ndo_start_xmit = lio_vf_rep_pkt_xmit,
+	.ndo_tx_timeout = lio_vf_rep_tx_timeout,
+	.ndo_get_phys_port_name = lio_vf_rep_phys_port_name,
+	.ndo_get_stats64 = lio_vf_rep_get_stats64,
+	.ndo_change_mtu = lio_vf_rep_change_mtu,
+	.ndo_get_port_parent_id = lio_vf_get_port_parent_id,
+};
+
+static int
+lio_vf_rep_send_soft_command(struct octeon_device *oct,
+			     void *req, int req_size,
+			     void *resp, int resp_size)
+{
+	int tot_resp_size = sizeof(struct lio_vf_rep_resp) + resp_size;
+	struct octeon_soft_command *sc = NULL;
+	struct lio_vf_rep_resp *rep_resp;
+	void *sc_req;
+	int err;
+
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct, req_size,
+					  tot_resp_size, 0);
+	if (!sc)
+		return -ENOMEM;
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	sc_req = (struct lio_vf_rep_req *)sc->virtdptr;
+	memcpy(sc_req, req, req_size);
+
+	rep_resp = (struct lio_vf_rep_resp *)sc->virtrptr;
+	memset(rep_resp, 0, tot_resp_size);
+	WRITE_ONCE(rep_resp->status, 1);
+
+	sc->iq_no = 0;
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC,
+				    OPCODE_NIC_VF_REP_CMD, 0, 0, 0);
+
+	err = octeon_send_soft_command(oct, sc);
+	if (err == IQ_SEND_FAILED)
+		goto free_buff;
+
+	err = wait_for_sc_completion_timeout(oct, sc, 0);
+	if (err)
+		return err;
+
+	err = READ_ONCE(rep_resp->status) ? -EBUSY : 0;
+	if (err)
+		dev_err(&oct->pci_dev->dev, "VF rep send config failed\n");
+	else if (resp)
+		memcpy(resp, (rep_resp + 1), resp_size);
+
+	WRITE_ONCE(sc->caller_is_done, true);
+	return err;
+
+free_buff:
+	octeon_free_soft_command(oct, sc);
+
+	return err;
+}
+
+static int
+lio_vf_rep_open(struct net_device *ndev)
+{
+	struct lio_vf_rep_desc *vf_rep = netdev_priv(ndev);
+	struct lio_vf_rep_req rep_cfg;
+	struct octeon_device *oct;
+	int ret;
+
+	oct = vf_rep->oct;
+
+	memset(&rep_cfg, 0, sizeof(rep_cfg));
+	rep_cfg.req_type = LIO_VF_REP_REQ_STATE;
+	rep_cfg.ifidx = vf_rep->ifidx;
+	rep_cfg.rep_state.state = LIO_VF_REP_STATE_UP;
+
+	ret = lio_vf_rep_send_soft_command(oct, &rep_cfg,
+					   sizeof(rep_cfg), NULL, 0);
+
+	if (ret) {
+		dev_err(&oct->pci_dev->dev,
+			"VF_REP open failed with err %d\n", ret);
+		return -EIO;
+	}
+
+	atomic_set(&vf_rep->ifstate, (atomic_read(&vf_rep->ifstate) |
+				      LIO_IFSTATE_RUNNING));
+
+	netif_carrier_on(ndev);
+	netif_start_queue(ndev);
+
+	return 0;
+}
+
+static int
+lio_vf_rep_stop(struct net_device *ndev)
+{
+	struct lio_vf_rep_desc *vf_rep = netdev_priv(ndev);
+	struct lio_vf_rep_req rep_cfg;
+	struct octeon_device *oct;
+	int ret;
+
+	oct = vf_rep->oct;
+
+	memset(&rep_cfg, 0, sizeof(rep_cfg));
+	rep_cfg.req_type = LIO_VF_REP_REQ_STATE;
+	rep_cfg.ifidx = vf_rep->ifidx;
+	rep_cfg.rep_state.state = LIO_VF_REP_STATE_DOWN;
+
+	ret = lio_vf_rep_send_soft_command(oct, &rep_cfg,
+					   sizeof(rep_cfg), NULL, 0);
+
+	if (ret) {
+		dev_err(&oct->pci_dev->dev,
+			"VF_REP dev stop failed with err %d\n", ret);
+		return -EIO;
+	}
+
+	atomic_set(&vf_rep->ifstate, (atomic_read(&vf_rep->ifstate) &
+				      ~LIO_IFSTATE_RUNNING));
+
+	netif_tx_disable(ndev);
+	netif_carrier_off(ndev);
+
+	return 0;
+}
+
+static void
+lio_vf_rep_tx_timeout(struct net_device *ndev, unsigned int txqueue)
+{
+	netif_trans_update(ndev);
+
+	netif_wake_queue(ndev);
+}
+
+static void
+lio_vf_rep_get_stats64(struct net_device *dev,
+		       struct rtnl_link_stats64 *stats64)
+{
+	struct lio_vf_rep_desc *vf_rep = netdev_priv(dev);
+
+	/* Swap tx and rx stats as VF rep is a switch port */
+	stats64->tx_packets = vf_rep->stats.rx_packets;
+	stats64->tx_bytes   = vf_rep->stats.rx_bytes;
+	stats64->tx_dropped = vf_rep->stats.rx_dropped;
+
+	stats64->rx_packets = vf_rep->stats.tx_packets;
+	stats64->rx_bytes   = vf_rep->stats.tx_bytes;
+	stats64->rx_dropped = vf_rep->stats.tx_dropped;
+}
+
+static int
+lio_vf_rep_change_mtu(struct net_device *ndev, int new_mtu)
+{
+	struct lio_vf_rep_desc *vf_rep = netdev_priv(ndev);
+	struct lio_vf_rep_req rep_cfg;
+	struct octeon_device *oct;
+	int ret;
+
+	oct = vf_rep->oct;
+
+	memset(&rep_cfg, 0, sizeof(rep_cfg));
+	rep_cfg.req_type = LIO_VF_REP_REQ_MTU;
+	rep_cfg.ifidx = vf_rep->ifidx;
+	rep_cfg.rep_mtu.mtu = cpu_to_be32(new_mtu);
+
+	ret = lio_vf_rep_send_soft_command(oct, &rep_cfg,
+					   sizeof(rep_cfg), NULL, 0);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev,
+			"Change MTU failed with err %d\n", ret);
+		return -EIO;
+	}
+
+	ndev->mtu = new_mtu;
+
+	return 0;
+}
+
+static int
+lio_vf_rep_phys_port_name(struct net_device *dev,
+			  char *buf, size_t len)
+{
+	struct lio_vf_rep_desc *vf_rep = netdev_priv(dev);
+	struct octeon_device *oct = vf_rep->oct;
+	int ret;
+
+	ret = snprintf(buf, len, "pf%dvf%d", oct->pf_num,
+		       vf_rep->ifidx - oct->pf_num * 64 - 1);
+	if (ret >= len)
+		return -EOPNOTSUPP;
+
+	return 0;
+}
+
+static struct net_device *
+lio_vf_rep_get_ndev(struct octeon_device *oct, int ifidx)
+{
+	int vf_id, max_vfs = CN23XX_MAX_VFS_PER_PF + 1;
+	int vfid_mask = max_vfs - 1;
+
+	if (ifidx <= oct->pf_num * max_vfs ||
+	    ifidx >= oct->pf_num * max_vfs + max_vfs)
+		return NULL;
+
+	/* ifidx 1-63 for PF0 VFs
+	 * ifidx 65-127 for PF1 VFs
+	 */
+	vf_id = (ifidx & vfid_mask) - 1;
+
+	return oct->vf_rep_list.ndev[vf_id];
+}
+
+static void
+lio_vf_rep_copy_packet(struct octeon_device *oct,
+		       struct sk_buff *skb,
+		       int len)
+{
+	if (likely(len > MIN_SKB_SIZE)) {
+		struct octeon_skb_page_info *pg_info;
+		unsigned char *va;
+
+		pg_info = ((struct octeon_skb_page_info *)(skb->cb));
+		if (pg_info->page) {
+			va = page_address(pg_info->page) +
+				pg_info->page_offset;
+			memcpy(skb->data, va, MIN_SKB_SIZE);
+			skb_put(skb, MIN_SKB_SIZE);
+		}
+
+		skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags,
+				pg_info->page,
+				pg_info->page_offset + MIN_SKB_SIZE,
+				len - MIN_SKB_SIZE,
+				LIO_RXBUFFER_SZ);
+	} else {
+		struct octeon_skb_page_info *pg_info =
+			((struct octeon_skb_page_info *)(skb->cb));
+
+		skb_copy_to_linear_data(skb, page_address(pg_info->page) +
+					pg_info->page_offset, len);
+		skb_put(skb, len);
+		put_page(pg_info->page);
+	}
+}
+
+static int
+lio_vf_rep_pkt_recv(struct octeon_recv_info *recv_info, void *buf)
+{
+	struct octeon_recv_pkt *recv_pkt = recv_info->recv_pkt;
+	struct lio_vf_rep_desc *vf_rep;
+	struct net_device *vf_ndev;
+	struct octeon_device *oct;
+	union octeon_rh *rh;
+	struct sk_buff *skb;
+	int i, ifidx;
+
+	oct = lio_get_device(recv_pkt->octeon_id);
+	if (!oct)
+		goto free_buffers;
+
+	skb = recv_pkt->buffer_ptr[0];
+	rh = &recv_pkt->rh;
+	ifidx = rh->r.ossp;
+
+	vf_ndev = lio_vf_rep_get_ndev(oct, ifidx);
+	if (!vf_ndev)
+		goto free_buffers;
+
+	vf_rep = netdev_priv(vf_ndev);
+	if (!(atomic_read(&vf_rep->ifstate) & LIO_IFSTATE_RUNNING) ||
+	    recv_pkt->buffer_count > 1)
+		goto free_buffers;
+
+	skb->dev = vf_ndev;
+
+	/* Multiple buffers are not used for vf_rep packets.
+	 * So just buffer_size[0] is valid.
+	 */
+	lio_vf_rep_copy_packet(oct, skb, recv_pkt->buffer_size[0]);
+
+	skb_pull(skb, rh->r_dh.len * BYTES_PER_DHLEN_UNIT);
+	skb->protocol = eth_type_trans(skb, skb->dev);
+	skb->ip_summed = CHECKSUM_NONE;
+
+	netif_rx(skb);
+
+	octeon_free_recv_info(recv_info);
+
+	return 0;
+
+free_buffers:
+	for (i = 0; i < recv_pkt->buffer_count; i++)
+		recv_buffer_free(recv_pkt->buffer_ptr[i]);
+
+	octeon_free_recv_info(recv_info);
+
+	return 0;
+}
+
+static void
+lio_vf_rep_packet_sent_callback(struct octeon_device *oct,
+				u32 status, void *buf)
+{
+	struct octeon_soft_command *sc = (struct octeon_soft_command *)buf;
+	struct sk_buff *skb = sc->ctxptr;
+	struct net_device *ndev = skb->dev;
+	u32 iq_no;
+
+	dma_unmap_single(&oct->pci_dev->dev, sc->dmadptr,
+			 sc->datasize, DMA_TO_DEVICE);
+	dev_kfree_skb_any(skb);
+	iq_no = sc->iq_no;
+	octeon_free_soft_command(oct, sc);
+
+	if (octnet_iq_is_full(oct, iq_no))
+		return;
+
+	if (netif_queue_stopped(ndev))
+		netif_wake_queue(ndev);
+}
+
+static netdev_tx_t
+lio_vf_rep_pkt_xmit(struct sk_buff *skb, struct net_device *ndev)
+{
+	struct lio_vf_rep_desc *vf_rep = netdev_priv(ndev);
+	struct net_device *parent_ndev = vf_rep->parent_ndev;
+	struct octeon_device *oct = vf_rep->oct;
+	struct octeon_instr_pki_ih3 *pki_ih3;
+	struct octeon_soft_command *sc;
+	struct lio *parent_lio;
+	int status;
+
+	parent_lio = GET_LIO(parent_ndev);
+
+	if (!(atomic_read(&vf_rep->ifstate) & LIO_IFSTATE_RUNNING) ||
+	    skb->len <= 0)
+		goto xmit_failed;
+
+	if (octnet_iq_is_full(vf_rep->oct, parent_lio->txq)) {
+		dev_err(&oct->pci_dev->dev, "VF rep: Device IQ full\n");
+		netif_stop_queue(ndev);
+		return NETDEV_TX_BUSY;
+	}
+
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct, 0, 16, 0);
+	if (!sc) {
+		dev_err(&oct->pci_dev->dev, "VF rep: Soft command alloc failed\n");
+		goto xmit_failed;
+	}
+
+	/* Multiple buffers are not used for vf_rep packets. */
+	if (skb_shinfo(skb)->nr_frags != 0) {
+		dev_err(&oct->pci_dev->dev, "VF rep: nr_frags != 0. Dropping packet\n");
+		octeon_free_soft_command(oct, sc);
+		goto xmit_failed;
+	}
+
+	sc->dmadptr = dma_map_single(&oct->pci_dev->dev,
+				     skb->data, skb->len, DMA_TO_DEVICE);
+	if (dma_mapping_error(&oct->pci_dev->dev, sc->dmadptr)) {
+		dev_err(&oct->pci_dev->dev, "VF rep: DMA mapping failed\n");
+		octeon_free_soft_command(oct, sc);
+		goto xmit_failed;
+	}
+
+	sc->virtdptr = skb->data;
+	sc->datasize = skb->len;
+	sc->ctxptr = skb;
+	sc->iq_no = parent_lio->txq;
+
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC, OPCODE_NIC_VF_REP_PKT,
+				    vf_rep->ifidx, 0, 0);
+	pki_ih3 = (struct octeon_instr_pki_ih3 *)&sc->cmd.cmd3.pki_ih3;
+	pki_ih3->tagtype = ORDERED_TAG;
+
+	sc->callback = lio_vf_rep_packet_sent_callback;
+	sc->callback_arg = sc;
+
+	status = octeon_send_soft_command(oct, sc);
+	if (status == IQ_SEND_FAILED) {
+		dma_unmap_single(&oct->pci_dev->dev, sc->dmadptr,
+				 sc->datasize, DMA_TO_DEVICE);
+		octeon_free_soft_command(oct, sc);
+		goto xmit_failed;
+	}
+
+	if (status == IQ_SEND_STOP)
+		netif_stop_queue(ndev);
+
+	netif_trans_update(ndev);
+
+	return NETDEV_TX_OK;
+
+xmit_failed:
+	dev_kfree_skb_any(skb);
+
+	return NETDEV_TX_OK;
+}
+
+static int lio_vf_get_port_parent_id(struct net_device *dev,
+				     struct netdev_phys_item_id *ppid)
+{
+	struct lio_vf_rep_desc *vf_rep = netdev_priv(dev);
+	struct net_device *parent_ndev = vf_rep->parent_ndev;
+	struct lio *lio = GET_LIO(parent_ndev);
+
+	ppid->id_len = ETH_ALEN;
+	ether_addr_copy(ppid->id, (void *)&lio->linfo.hw_addr + 2);
+
+	return 0;
+}
+
+static void
+lio_vf_rep_fetch_stats(struct work_struct *work)
+{
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct lio_vf_rep_desc *vf_rep = wk->ctxptr;
+	struct lio_vf_rep_stats stats;
+	struct lio_vf_rep_req rep_cfg;
+	struct octeon_device *oct;
+	int ret;
+
+	oct = vf_rep->oct;
+
+	memset(&rep_cfg, 0, sizeof(rep_cfg));
+	rep_cfg.req_type = LIO_VF_REP_REQ_STATS;
+	rep_cfg.ifidx = vf_rep->ifidx;
+
+	ret = lio_vf_rep_send_soft_command(oct, &rep_cfg, sizeof(rep_cfg),
+					   &stats, sizeof(stats));
+
+	if (!ret) {
+		octeon_swap_8B_data((u64 *)&stats, (sizeof(stats) >> 3));
+		memcpy(&vf_rep->stats, &stats, sizeof(stats));
+	}
+
+	schedule_delayed_work(&vf_rep->stats_wk.work,
+			      msecs_to_jiffies(LIO_VF_REP_STATS_POLL_TIME_MS));
+}
+
+int
+lio_vf_rep_create(struct octeon_device *oct)
+{
+	struct lio_vf_rep_desc *vf_rep;
+	struct net_device *ndev;
+	int i, num_vfs;
+
+	if (oct->eswitch_mode != DEVLINK_ESWITCH_MODE_SWITCHDEV)
+		return 0;
+
+	if (!oct->sriov_info.sriov_enabled)
+		return 0;
+
+	num_vfs = oct->sriov_info.num_vfs_alloced;
+
+	oct->vf_rep_list.num_vfs = 0;
+	for (i = 0; i < num_vfs; i++) {
+		ndev = alloc_etherdev(sizeof(struct lio_vf_rep_desc));
+
+		if (!ndev) {
+			dev_err(&oct->pci_dev->dev,
+				"VF rep device %d creation failed\n", i);
+			goto cleanup;
+		}
+
+		ndev->min_mtu = LIO_MIN_MTU_SIZE;
+		ndev->max_mtu = LIO_MAX_MTU_SIZE;
+		ndev->netdev_ops = &lio_vf_rep_ndev_ops;
+
+		vf_rep = netdev_priv(ndev);
+		memset(vf_rep, 0, sizeof(*vf_rep));
+
+		vf_rep->ndev = ndev;
+		vf_rep->oct = oct;
+		vf_rep->parent_ndev = oct->props[0].netdev;
+		vf_rep->ifidx = (oct->pf_num * 64) + i + 1;
+
+		eth_hw_addr_random(ndev);
+
+		if (register_netdev(ndev)) {
+			dev_err(&oct->pci_dev->dev, "VF rep nerdev registration failed\n");
+
+			free_netdev(ndev);
+			goto cleanup;
+		}
+
+		netif_carrier_off(ndev);
+
+		INIT_DELAYED_WORK(&vf_rep->stats_wk.work,
+				  lio_vf_rep_fetch_stats);
+		vf_rep->stats_wk.ctxptr = (void *)vf_rep;
+		schedule_delayed_work(&vf_rep->stats_wk.work,
+				      msecs_to_jiffies
+				      (LIO_VF_REP_STATS_POLL_TIME_MS));
+		oct->vf_rep_list.num_vfs++;
+		oct->vf_rep_list.ndev[i] = ndev;
+	}
+
+	if (octeon_register_dispatch_fn(oct, OPCODE_NIC,
+					OPCODE_NIC_VF_REP_PKT,
+					lio_vf_rep_pkt_recv, oct)) {
+		dev_err(&oct->pci_dev->dev, "VF rep Dispatch func registration failed\n");
+
+		goto cleanup;
+	}
+
+	return 0;
+
+cleanup:
+	for (i = 0; i < oct->vf_rep_list.num_vfs; i++) {
+		ndev = oct->vf_rep_list.ndev[i];
+		oct->vf_rep_list.ndev[i] = NULL;
+		if (ndev) {
+			vf_rep = netdev_priv(ndev);
+			cancel_delayed_work_sync
+				(&vf_rep->stats_wk.work);
+			unregister_netdev(ndev);
+			free_netdev(ndev);
+		}
+	}
+
+	oct->vf_rep_list.num_vfs = 0;
+
+	return -1;
+}
+
+void
+lio_vf_rep_destroy(struct octeon_device *oct)
+{
+	struct lio_vf_rep_desc *vf_rep;
+	struct net_device *ndev;
+	int i;
+
+	if (oct->eswitch_mode != DEVLINK_ESWITCH_MODE_SWITCHDEV)
+		return;
+
+	if (!oct->sriov_info.sriov_enabled)
+		return;
+
+	for (i = 0; i < oct->vf_rep_list.num_vfs; i++) {
+		ndev = oct->vf_rep_list.ndev[i];
+		oct->vf_rep_list.ndev[i] = NULL;
+		if (ndev) {
+			vf_rep = netdev_priv(ndev);
+			cancel_delayed_work_sync
+				(&vf_rep->stats_wk.work);
+			netif_tx_disable(ndev);
+			netif_carrier_off(ndev);
+
+			unregister_netdev(ndev);
+			free_netdev(ndev);
+		}
+	}
+
+	oct->vf_rep_list.num_vfs = 0;
+}
+
+static int
+lio_vf_rep_netdev_event(struct notifier_block *nb,
+			unsigned long event, void *ptr)
+{
+	struct net_device *ndev = netdev_notifier_info_to_dev(ptr);
+	struct lio_vf_rep_desc *vf_rep;
+	struct lio_vf_rep_req rep_cfg;
+	struct octeon_device *oct;
+	int ret;
+
+	switch (event) {
+	case NETDEV_REGISTER:
+	case NETDEV_CHANGENAME:
+		break;
+
+	default:
+		return NOTIFY_DONE;
+	}
+
+	if (ndev->netdev_ops != &lio_vf_rep_ndev_ops)
+		return NOTIFY_DONE;
+
+	vf_rep = netdev_priv(ndev);
+	oct = vf_rep->oct;
+
+	if (strlen(ndev->name) > LIO_IF_NAME_SIZE) {
+		dev_err(&oct->pci_dev->dev,
+			"Device name change sync failed as the size is > %d\n",
+			LIO_IF_NAME_SIZE);
+		return NOTIFY_DONE;
+	}
+
+	memset(&rep_cfg, 0, sizeof(rep_cfg));
+	rep_cfg.req_type = LIO_VF_REP_REQ_DEVNAME;
+	rep_cfg.ifidx = vf_rep->ifidx;
+	strncpy(rep_cfg.rep_name.name, ndev->name, LIO_IF_NAME_SIZE);
+
+	ret = lio_vf_rep_send_soft_command(oct, &rep_cfg,
+					   sizeof(rep_cfg), NULL, 0);
+	if (ret)
+		dev_err(&oct->pci_dev->dev,
+			"vf_rep netdev name change failed with err %d\n", ret);
+
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block lio_vf_rep_netdev_notifier = {
+	.notifier_call = lio_vf_rep_netdev_event,
+};
+
+int
+lio_vf_rep_modinit(void)
+{
+	if (register_netdevice_notifier(&lio_vf_rep_netdev_notifier)) {
+		pr_err("netdev notifier registration failed\n");
+		return -EFAULT;
+	}
+
+	return 0;
+}
+
+void
+lio_vf_rep_modexit(void)
+{
+	if (unregister_netdevice_notifier(&lio_vf_rep_netdev_notifier))
+		pr_err("netdev notifier unregister failed\n");
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/lio_vf_rep.h b/drivers/net/ethernet/cavium/liquidio/lio_vf_rep.h
new file mode 100644
index 000000000..bb3cedc63
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/lio_vf_rep.h
@@ -0,0 +1,49 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2017 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more
+ * details.
+ *
+ * This file may also be available under a different license from Cavium.
+ * Contact Cavium, Inc. for more information
+ **********************************************************************/
+
+/*! \file octeon_vf_main.h
+ *  \brief Host Driver: This file defines vf_rep related macros and structures
+ */
+#ifndef __LIO_VF_REP_H__
+#define __LIO_VF_REP_H__
+#define LIO_VF_REP_REQ_TMO_MS 5000
+#define LIO_VF_REP_STATS_POLL_TIME_MS 200
+
+struct lio_vf_rep_desc {
+	struct net_device *parent_ndev;
+	struct net_device *ndev;
+	struct octeon_device *oct;
+	struct lio_vf_rep_stats stats;
+	struct cavium_wk stats_wk;
+	atomic_t ifstate;
+	int ifidx;
+};
+
+struct lio_vf_rep_sc_ctx {
+	struct completion complete;
+};
+
+int lio_vf_rep_create(struct octeon_device *oct);
+void lio_vf_rep_destroy(struct octeon_device *oct);
+int lio_vf_rep_modinit(void);
+void lio_vf_rep_modexit(void);
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/liquidio_common.h b/drivers/net/ethernet/cavium/liquidio/liquidio_common.h
new file mode 100644
index 000000000..4da90757c
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/liquidio_common.h
@@ -0,0 +1,1039 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*!  \file  liquidio_common.h
+ *   \brief Common: Structures and macros used in PCI-NIC package by core and
+ *   host driver.
+ */
+
+#ifndef __LIQUIDIO_COMMON_H__
+#define __LIQUIDIO_COMMON_H__
+
+#include "octeon_config.h"
+
+#define LIQUIDIO_BASE_MAJOR_VERSION 1
+#define LIQUIDIO_BASE_MINOR_VERSION 7
+#define LIQUIDIO_BASE_MICRO_VERSION 2
+#define LIQUIDIO_BASE_VERSION   __stringify(LIQUIDIO_BASE_MAJOR_VERSION) "." \
+				__stringify(LIQUIDIO_BASE_MINOR_VERSION)
+
+struct lio_version {
+	u16  major;
+	u16  minor;
+	u16  micro;
+	u16  reserved;
+};
+
+#define CONTROL_IQ 0
+/** Tag types used by Octeon cores in its work. */
+enum octeon_tag_type {
+	ORDERED_TAG = 0,
+	ATOMIC_TAG = 1,
+	NULL_TAG = 2,
+	NULL_NULL_TAG = 3
+};
+
+/* pre-defined host->NIC tag values */
+#define LIO_CONTROL  (0x11111110)
+#define LIO_DATA(i)  (0x11111111 + (i))
+
+/* Opcodes used by host driver/apps to perform operations on the core.
+ * These are used to identify the major subsystem that the operation
+ * is for.
+ */
+#define OPCODE_CORE 0           /* used for generic core operations */
+#define OPCODE_NIC  1           /* used for NIC operations */
+/* Subcodes are used by host driver/apps to identify the sub-operation
+ * for the core. They only need to by unique for a given subsystem.
+ */
+#define OPCODE_SUBCODE(op, sub)       ((((op) & 0x0f) << 8) | ((sub) & 0x7f))
+
+/** OPCODE_CORE subcodes. For future use. */
+
+/** OPCODE_NIC subcodes */
+
+/* This subcode is sent by core PCI driver to indicate cores are ready. */
+#define OPCODE_NIC_CORE_DRV_ACTIVE     0x01
+#define OPCODE_NIC_NW_DATA             0x02     /* network packet data */
+#define OPCODE_NIC_CMD                 0x03
+#define OPCODE_NIC_INFO                0x04
+#define OPCODE_NIC_PORT_STATS          0x05
+#define OPCODE_NIC_MDIO45              0x06
+#define OPCODE_NIC_TIMESTAMP           0x07
+#define OPCODE_NIC_INTRMOD_CFG         0x08
+#define OPCODE_NIC_IF_CFG              0x09
+#define OPCODE_NIC_VF_DRV_NOTICE       0x0A
+#define OPCODE_NIC_INTRMOD_PARAMS      0x0B
+#define OPCODE_NIC_QCOUNT_UPDATE       0x12
+#define OPCODE_NIC_SET_TRUSTED_VF	0x13
+#define OPCODE_NIC_SYNC_OCTEON_TIME	0x14
+#define VF_DRV_LOADED                  1
+#define VF_DRV_REMOVED                -1
+#define VF_DRV_MACADDR_CHANGED         2
+
+#define OPCODE_NIC_VF_REP_PKT          0x15
+#define OPCODE_NIC_VF_REP_CMD          0x16
+#define OPCODE_NIC_UBOOT_CTL           0x17
+
+#define CORE_DRV_TEST_SCATTER_OP    0xFFF5
+
+/* Application codes advertised by the core driver initialization packet. */
+#define CVM_DRV_APP_START           0x0
+#define CVM_DRV_NO_APP              0
+#define CVM_DRV_APP_COUNT           0x2
+#define CVM_DRV_BASE_APP            (CVM_DRV_APP_START + 0x0)
+#define CVM_DRV_NIC_APP             (CVM_DRV_APP_START + 0x1)
+#define CVM_DRV_INVALID_APP         (CVM_DRV_APP_START + 0x2)
+#define CVM_DRV_APP_END             (CVM_DRV_INVALID_APP - 1)
+
+#define BYTES_PER_DHLEN_UNIT        8
+#define MAX_REG_CNT                 2000000U
+#define INTRNAMSIZ                  32
+#define IRQ_NAME_OFF(i)             ((i) * INTRNAMSIZ)
+#define MAX_IOQ_INTERRUPTS_PER_PF   (64 * 2)
+#define MAX_IOQ_INTERRUPTS_PER_VF   (8 * 2)
+
+#define SCR2_BIT_FW_LOADED	    63
+
+/* App specific capabilities from firmware to pf driver */
+#define LIQUIDIO_TIME_SYNC_CAP 0x1
+#define LIQUIDIO_SWITCHDEV_CAP 0x2
+#define LIQUIDIO_SPOOFCHK_CAP  0x4
+
+/* error status return from firmware */
+#define OCTEON_REQUEST_NO_PERMISSION 0xc
+
+static inline u32 incr_index(u32 index, u32 count, u32 max)
+{
+	if ((index + count) >= max)
+		index = index + count - max;
+	else
+		index += count;
+
+	return index;
+}
+
+#define OCT_BOARD_NAME 32
+#define OCT_SERIAL_LEN 64
+
+/* Structure used by core driver to send indication that the Octeon
+ * application is ready.
+ */
+struct octeon_core_setup {
+	u64 corefreq;
+
+	char boardname[OCT_BOARD_NAME];
+
+	char board_serial_number[OCT_SERIAL_LEN];
+
+	u64 board_rev_major;
+
+	u64 board_rev_minor;
+
+};
+
+/*---------------------------  SCATTER GATHER ENTRY  -----------------------*/
+
+/* The Scatter-Gather List Entry. The scatter or gather component used with
+ * a Octeon input instruction has this format.
+ */
+struct octeon_sg_entry {
+	/** The first 64 bit gives the size of data in each dptr.*/
+	union {
+		u16 size[4];
+		u64 size64;
+	} u;
+
+	/** The 4 dptr pointers for this entry. */
+	u64 ptr[4];
+
+};
+
+#define OCT_SG_ENTRY_SIZE    (sizeof(struct octeon_sg_entry))
+
+/* \brief Add size to gather list
+ * @param sg_entry scatter/gather entry
+ * @param size size to add
+ * @param pos position to add it.
+ */
+static inline void add_sg_size(struct octeon_sg_entry *sg_entry,
+			       u16 size,
+			       u32 pos)
+{
+#ifdef __BIG_ENDIAN_BITFIELD
+	sg_entry->u.size[pos] = size;
+#else
+	sg_entry->u.size[3 - pos] = size;
+#endif
+}
+
+/*------------------------- End Scatter/Gather ---------------------------*/
+
+#define   OCTNET_FRM_LENGTH_SIZE      8
+
+#define   OCTNET_FRM_PTP_HEADER_SIZE  8
+
+#define   OCTNET_FRM_HEADER_SIZE     22 /* VLAN + Ethernet */
+
+#define   OCTNET_MIN_FRM_SIZE        64
+
+#define   OCTNET_MAX_FRM_SIZE        (16000 + OCTNET_FRM_HEADER_SIZE)
+
+#define   OCTNET_DEFAULT_MTU         (1500)
+#define   OCTNET_DEFAULT_FRM_SIZE  (OCTNET_DEFAULT_MTU + OCTNET_FRM_HEADER_SIZE)
+
+/** NIC Commands are sent using this Octeon Input Queue */
+#define   OCTNET_CMD_Q                0
+
+/* NIC Command types */
+#define   OCTNET_CMD_CHANGE_MTU       0x1
+#define   OCTNET_CMD_CHANGE_MACADDR   0x2
+#define   OCTNET_CMD_CHANGE_DEVFLAGS  0x3
+#define   OCTNET_CMD_RX_CTL           0x4
+
+#define	  OCTNET_CMD_SET_MULTI_LIST   0x5
+#define   OCTNET_CMD_CLEAR_STATS      0x6
+
+/* command for setting the speed, duplex & autoneg */
+#define   OCTNET_CMD_SET_SETTINGS     0x7
+#define   OCTNET_CMD_SET_FLOW_CTL     0x8
+
+#define   OCTNET_CMD_MDIO_READ_WRITE  0x9
+#define   OCTNET_CMD_GPIO_ACCESS      0xA
+#define   OCTNET_CMD_LRO_ENABLE       0xB
+#define   OCTNET_CMD_LRO_DISABLE      0xC
+#define   OCTNET_CMD_SET_RSS          0xD
+#define   OCTNET_CMD_WRITE_SA         0xE
+#define   OCTNET_CMD_DELETE_SA        0xF
+#define   OCTNET_CMD_UPDATE_SA        0x12
+
+#define   OCTNET_CMD_TNL_RX_CSUM_CTL 0x10
+#define   OCTNET_CMD_TNL_TX_CSUM_CTL 0x11
+#define   OCTNET_CMD_IPSECV2_AH_ESP_CTL 0x13
+#define   OCTNET_CMD_VERBOSE_ENABLE   0x14
+#define   OCTNET_CMD_VERBOSE_DISABLE  0x15
+
+#define   OCTNET_CMD_VLAN_FILTER_CTL 0x16
+#define   OCTNET_CMD_ADD_VLAN_FILTER  0x17
+#define   OCTNET_CMD_DEL_VLAN_FILTER  0x18
+#define   OCTNET_CMD_VXLAN_PORT_CONFIG 0x19
+
+#define   OCTNET_CMD_ID_ACTIVE         0x1a
+
+#define   OCTNET_CMD_SET_UC_LIST       0x1b
+#define   OCTNET_CMD_SET_VF_LINKSTATE  0x1c
+
+#define   OCTNET_CMD_QUEUE_COUNT_CTL	0x1f
+
+#define   OCTNET_CMD_GROUP1             1
+#define   OCTNET_CMD_SET_VF_SPOOFCHK    0x1
+#define   OCTNET_GROUP1_LAST_CMD        OCTNET_CMD_SET_VF_SPOOFCHK
+
+#define   OCTNET_CMD_VXLAN_PORT_ADD    0x0
+#define   OCTNET_CMD_VXLAN_PORT_DEL    0x1
+#define   OCTNET_CMD_RXCSUM_ENABLE     0x0
+#define   OCTNET_CMD_RXCSUM_DISABLE    0x1
+#define   OCTNET_CMD_TXCSUM_ENABLE     0x0
+#define   OCTNET_CMD_TXCSUM_DISABLE    0x1
+#define   OCTNET_CMD_VLAN_FILTER_ENABLE 0x1
+#define   OCTNET_CMD_VLAN_FILTER_DISABLE 0x0
+
+#define   OCTNET_CMD_FAIL 0x1
+
+#define   SEAPI_CMD_FEC_SET             0x0
+#define   SEAPI_CMD_FEC_SET_DISABLE       0x0
+#define   SEAPI_CMD_FEC_SET_RS            0x1
+#define   SEAPI_CMD_FEC_GET             0x1
+
+#define   SEAPI_CMD_SPEED_SET           0x2
+#define   SEAPI_CMD_SPEED_GET           0x3
+
+#define OPCODE_NIC_VF_PORT_STATS        0x22
+
+#define   LIO_CMD_WAIT_TM 100
+
+/* RX(packets coming from wire) Checksum verification flags */
+/* TCP/UDP csum */
+#define   CNNIC_L4SUM_VERIFIED             0x1
+#define   CNNIC_IPSUM_VERIFIED             0x2
+#define   CNNIC_TUN_CSUM_VERIFIED          0x4
+#define   CNNIC_CSUM_VERIFIED (CNNIC_IPSUM_VERIFIED | CNNIC_L4SUM_VERIFIED)
+
+/*LROIPV4 and LROIPV6 Flags*/
+#define   OCTNIC_LROIPV4    0x1
+#define   OCTNIC_LROIPV6    0x2
+
+/* Interface flags communicated between host driver and core app. */
+enum octnet_ifflags {
+	OCTNET_IFFLAG_PROMISC   = 0x01,
+	OCTNET_IFFLAG_ALLMULTI  = 0x02,
+	OCTNET_IFFLAG_MULTICAST = 0x04,
+	OCTNET_IFFLAG_BROADCAST = 0x08,
+	OCTNET_IFFLAG_UNICAST   = 0x10
+};
+
+/*   wqe
+ *  ---------------  0
+ * |  wqe  word0-3 |
+ *  ---------------  32
+ * |    PCI IH     |
+ *  ---------------  40
+ * |     RPTR      |
+ *  ---------------  48
+ * |    PCI IRH    |
+ *  ---------------  56
+ * |  OCT_NET_CMD  |
+ *  ---------------  64
+ * | Addtl 8-BData |
+ * |               |
+ *  ---------------
+ */
+
+union octnet_cmd {
+	u64 u64;
+
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		u64 cmd:5;
+
+		u64 more:6; /* How many udd words follow the command */
+
+		u64 cmdgroup:8;
+		u64 reserved:21;
+
+		u64 param1:16;
+
+		u64 param2:8;
+
+#else
+
+		u64 param2:8;
+
+		u64 param1:16;
+
+		u64 reserved:21;
+		u64 cmdgroup:8;
+
+		u64 more:6;
+
+		u64 cmd:5;
+
+#endif
+	} s;
+
+};
+
+#define   OCTNET_CMD_SIZE     (sizeof(union octnet_cmd))
+
+/*pkiih3 + irh + ossp[0] + ossp[1] + rdp + rptr = 40 bytes */
+#define LIO_SOFTCMDRESP_IH2       40
+#define LIO_SOFTCMDRESP_IH3       (40 + 8)
+
+#define LIO_PCICMD_O2             24
+#define LIO_PCICMD_O3             (24 + 8)
+
+/* Instruction Header(DPI) - for OCTEON-III models */
+struct  octeon_instr_ih3 {
+#ifdef __BIG_ENDIAN_BITFIELD
+
+	/** Reserved3 */
+	u64     reserved3:1;
+
+	/** Gather indicator 1=gather*/
+	u64     gather:1;
+
+	/** Data length OR no. of entries in gather list */
+	u64     dlengsz:14;
+
+	/** Front Data size */
+	u64     fsz:6;
+
+	/** Reserved2 */
+	u64     reserved2:4;
+
+	/** PKI port kind - PKIND */
+	u64     pkind:6;
+
+	/** Reserved1 */
+	u64     reserved1:32;
+
+#else
+	/** Reserved1 */
+	u64     reserved1:32;
+
+	/** PKI port kind - PKIND */
+	u64     pkind:6;
+
+	/** Reserved2 */
+	u64     reserved2:4;
+
+	/** Front Data size */
+	u64     fsz:6;
+
+	/** Data length OR no. of entries in gather list */
+	u64     dlengsz:14;
+
+	/** Gather indicator 1=gather*/
+	u64     gather:1;
+
+	/** Reserved3 */
+	u64     reserved3:1;
+
+#endif
+};
+
+/* Optional PKI Instruction Header(PKI IH) - for OCTEON-III models */
+/** BIG ENDIAN format.   */
+struct  octeon_instr_pki_ih3 {
+#ifdef __BIG_ENDIAN_BITFIELD
+
+	/** Wider bit */
+	u64     w:1;
+
+	/** Raw mode indicator 1 = RAW */
+	u64     raw:1;
+
+	/** Use Tag */
+	u64     utag:1;
+
+	/** Use QPG */
+	u64     uqpg:1;
+
+	/** Reserved2 */
+	u64     reserved2:1;
+
+	/** Parse Mode */
+	u64     pm:3;
+
+	/** Skip Length */
+	u64     sl:8;
+
+	/** Use Tag Type */
+	u64     utt:1;
+
+	/** Tag type */
+	u64     tagtype:2;
+
+	/** Reserved1 */
+	u64     reserved1:2;
+
+	/** QPG Value */
+	u64     qpg:11;
+
+	/** Tag Value */
+	u64     tag:32;
+
+#else
+
+	/** Tag Value */
+	u64     tag:32;
+
+	/** QPG Value */
+	u64     qpg:11;
+
+	/** Reserved1 */
+	u64     reserved1:2;
+
+	/** Tag type */
+	u64     tagtype:2;
+
+	/** Use Tag Type */
+	u64     utt:1;
+
+	/** Skip Length */
+	u64     sl:8;
+
+	/** Parse Mode */
+	u64     pm:3;
+
+	/** Reserved2 */
+	u64     reserved2:1;
+
+	/** Use QPG */
+	u64     uqpg:1;
+
+	/** Use Tag */
+	u64     utag:1;
+
+	/** Raw mode indicator 1 = RAW */
+	u64     raw:1;
+
+	/** Wider bit */
+	u64     w:1;
+#endif
+
+};
+
+/** Instruction Header */
+struct octeon_instr_ih2 {
+#ifdef __BIG_ENDIAN_BITFIELD
+	/** Raw mode indicator 1 = RAW */
+	u64 raw:1;
+
+	/** Gather indicator 1=gather*/
+	u64 gather:1;
+
+	/** Data length OR no. of entries in gather list */
+	u64 dlengsz:14;
+
+	/** Front Data size */
+	u64 fsz:6;
+
+	/** Packet Order / Work Unit selection (1 of 8)*/
+	u64 qos:3;
+
+	/** Core group selection (1 of 16) */
+	u64 grp:4;
+
+	/** Short Raw Packet Indicator 1=short raw pkt */
+	u64 rs:1;
+
+	/** Tag type */
+	u64 tagtype:2;
+
+	/** Tag Value */
+	u64 tag:32;
+#else
+	/** Tag Value */
+	u64 tag:32;
+
+	/** Tag type */
+	u64 tagtype:2;
+
+	/** Short Raw Packet Indicator 1=short raw pkt */
+	u64 rs:1;
+
+	/** Core group selection (1 of 16) */
+	u64 grp:4;
+
+	/** Packet Order / Work Unit selection (1 of 8)*/
+	u64 qos:3;
+
+	/** Front Data size */
+	u64 fsz:6;
+
+	/** Data length OR no. of entries in gather list */
+	u64 dlengsz:14;
+
+	/** Gather indicator 1=gather*/
+	u64 gather:1;
+
+	/** Raw mode indicator 1 = RAW */
+	u64 raw:1;
+#endif
+};
+
+/** Input Request Header */
+struct octeon_instr_irh {
+#ifdef __BIG_ENDIAN_BITFIELD
+	u64 opcode:4;
+	u64 rflag:1;
+	u64 subcode:7;
+	u64 vlan:12;
+	u64 priority:3;
+	u64 reserved:5;
+	u64 ossp:32;             /* opcode/subcode specific parameters */
+#else
+	u64 ossp:32;             /* opcode/subcode specific parameters */
+	u64 reserved:5;
+	u64 priority:3;
+	u64 vlan:12;
+	u64 subcode:7;
+	u64 rflag:1;
+	u64 opcode:4;
+#endif
+};
+
+/** Return Data Parameters */
+struct octeon_instr_rdp {
+#ifdef __BIG_ENDIAN_BITFIELD
+	u64 reserved:49;
+	u64 pcie_port:3;
+	u64 rlen:12;
+#else
+	u64 rlen:12;
+	u64 pcie_port:3;
+	u64 reserved:49;
+#endif
+};
+
+/** Receive Header */
+union octeon_rh {
+#ifdef __BIG_ENDIAN_BITFIELD
+	u64 u64;
+	struct {
+		u64 opcode:4;
+		u64 subcode:8;
+		u64 len:3;     /** additional 64-bit words */
+		u64 reserved:17;
+		u64 ossp:32;   /** opcode/subcode specific parameters */
+	} r;
+	struct {
+		u64 opcode:4;
+		u64 subcode:8;
+		u64 len:3;     /** additional 64-bit words */
+		u64 extra:28;
+		u64 vlan:12;
+		u64 priority:3;
+		u64 csum_verified:3;     /** checksum verified. */
+		u64 has_hwtstamp:1;      /** Has hardware timestamp. 1 = yes. */
+		u64 encap_on:1;
+		u64 has_hash:1;          /** Has hash (rth or rss). 1 = yes. */
+	} r_dh;
+	struct {
+		u64 opcode:4;
+		u64 subcode:8;
+		u64 len:3;     /** additional 64-bit words */
+		u64 reserved:11;
+		u64 num_gmx_ports:8;
+		u64 max_nic_ports:10;
+		u64 app_cap_flags:4;
+		u64 app_mode:8;
+		u64 pkind:8;
+	} r_core_drv_init;
+	struct {
+		u64 opcode:4;
+		u64 subcode:8;
+		u64 len:3;       /** additional 64-bit words */
+		u64 reserved:8;
+		u64 extra:25;
+		u64 gmxport:16;
+	} r_nic_info;
+#else
+	u64 u64;
+	struct {
+		u64 ossp:32;  /** opcode/subcode specific parameters */
+		u64 reserved:17;
+		u64 len:3;    /** additional 64-bit words */
+		u64 subcode:8;
+		u64 opcode:4;
+	} r;
+	struct {
+		u64 has_hash:1;          /** Has hash (rth or rss). 1 = yes. */
+		u64 encap_on:1;
+		u64 has_hwtstamp:1;      /** 1 = has hwtstamp */
+		u64 csum_verified:3;     /** checksum verified. */
+		u64 priority:3;
+		u64 vlan:12;
+		u64 extra:28;
+		u64 len:3;    /** additional 64-bit words */
+		u64 subcode:8;
+		u64 opcode:4;
+	} r_dh;
+	struct {
+		u64 pkind:8;
+		u64 app_mode:8;
+		u64 app_cap_flags:4;
+		u64 max_nic_ports:10;
+		u64 num_gmx_ports:8;
+		u64 reserved:11;
+		u64 len:3;       /** additional 64-bit words */
+		u64 subcode:8;
+		u64 opcode:4;
+	} r_core_drv_init;
+	struct {
+		u64 gmxport:16;
+		u64 extra:25;
+		u64 reserved:8;
+		u64 len:3;       /** additional 64-bit words */
+		u64 subcode:8;
+		u64 opcode:4;
+	} r_nic_info;
+#endif
+};
+
+#define  OCT_RH_SIZE   (sizeof(union  octeon_rh))
+
+union octnic_packet_params {
+	u32 u32;
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		u32 reserved:24;
+		u32 ip_csum:1;		/* Perform IP header checksum(s) */
+		/* Perform Outer transport header checksum */
+		u32 transport_csum:1;
+		/* Find tunnel, and perform transport csum. */
+		u32 tnl_csum:1;
+		u32 tsflag:1;		/* Timestamp this packet */
+		u32 ipsec_ops:4;	/* IPsec operation */
+#else
+		u32 ipsec_ops:4;
+		u32 tsflag:1;
+		u32 tnl_csum:1;
+		u32 transport_csum:1;
+		u32 ip_csum:1;
+		u32 reserved:24;
+#endif
+	} s;
+};
+
+/** Status of a RGMII Link on Octeon as seen by core driver. */
+union oct_link_status {
+	u64 u64;
+
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		u64 duplex:8;
+		u64 mtu:16;
+		u64 speed:16;
+		u64 link_up:1;
+		u64 autoneg:1;
+		u64 if_mode:5;
+		u64 pause:1;
+		u64 flashing:1;
+		u64 phy_type:5;
+		u64 reserved:10;
+#else
+		u64 reserved:10;
+		u64 phy_type:5;
+		u64 flashing:1;
+		u64 pause:1;
+		u64 if_mode:5;
+		u64 autoneg:1;
+		u64 link_up:1;
+		u64 speed:16;
+		u64 mtu:16;
+		u64 duplex:8;
+#endif
+	} s;
+};
+
+enum lio_phy_type {
+	LIO_PHY_PORT_TP = 0x0,
+	LIO_PHY_PORT_FIBRE = 0x1,
+	LIO_PHY_PORT_UNKNOWN,
+};
+
+/** The txpciq info passed to host from the firmware */
+
+union oct_txpciq {
+	u64 u64;
+
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		u64 q_no:8;
+		u64 port:8;
+		u64 pkind:6;
+		u64 use_qpg:1;
+		u64 qpg:11;
+		u64 reserved0:10;
+		u64 ctrl_qpg:11;
+		u64 reserved:9;
+#else
+		u64 reserved:9;
+		u64 ctrl_qpg:11;
+		u64 reserved0:10;
+		u64 qpg:11;
+		u64 use_qpg:1;
+		u64 pkind:6;
+		u64 port:8;
+		u64 q_no:8;
+#endif
+	} s;
+};
+
+/** The rxpciq info passed to host from the firmware */
+
+union oct_rxpciq {
+	u64 u64;
+
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		u64 q_no:8;
+		u64 reserved:56;
+#else
+		u64 reserved:56;
+		u64 q_no:8;
+#endif
+	} s;
+};
+
+/** Information for a OCTEON ethernet interface shared between core & host. */
+struct oct_link_info {
+	union oct_link_status link;
+	u64 hw_addr;
+
+#ifdef __BIG_ENDIAN_BITFIELD
+	u64 gmxport:16;
+	u64 macaddr_is_admin_asgnd:1;
+	u64 rsvd:13;
+	u64 macaddr_spoofchk:1;
+	u64 rsvd1:17;
+	u64 num_txpciq:8;
+	u64 num_rxpciq:8;
+#else
+	u64 num_rxpciq:8;
+	u64 num_txpciq:8;
+	u64 rsvd1:17;
+	u64 macaddr_spoofchk:1;
+	u64 rsvd:13;
+	u64 macaddr_is_admin_asgnd:1;
+	u64 gmxport:16;
+#endif
+
+	union oct_txpciq txpciq[MAX_IOQS_PER_NICIF];
+	union oct_rxpciq rxpciq[MAX_IOQS_PER_NICIF];
+};
+
+#define OCT_LINK_INFO_SIZE   (sizeof(struct oct_link_info))
+
+struct liquidio_if_cfg_info {
+	u64 iqmask; /** mask for IQs enabled for  the port */
+	u64 oqmask; /** mask for OQs enabled for the port */
+	struct oct_link_info linfo; /** initial link information */
+	char   liquidio_firmware_version[32];
+};
+
+/** Stats for each NIC port in RX direction. */
+struct nic_rx_stats {
+	/* link-level stats */
+	u64 total_rcvd;		/* Received packets */
+	u64 bytes_rcvd;		/* Octets of received packets */
+	u64 total_bcst;		/* Number of non-dropped L2 broadcast packets */
+	u64 total_mcst;		/* Number of non-dropped L2 multicast packets */
+	u64 runts;		/* Packets shorter than allowed */
+	u64 ctl_rcvd;		/* Received PAUSE packets */
+	u64 fifo_err;		/* Packets dropped due to RX FIFO full */
+	u64 dmac_drop;		/* Packets dropped by the DMAC filter */
+	u64 fcs_err;		/* Sum of fragment, overrun, and FCS errors */
+	u64 jabber_err;		/* Packets larger than allowed */
+	u64 l2_err;		/* Sum of DMA, parity, PCAM access, no memory,
+				 * buffer overflow, malformed L2 header or
+				 * length, oversize errors
+				 **/
+	u64 frame_err;		/* Sum of IPv4 and L4 checksum errors */
+	u64 red_drops;		/* Packets dropped by RED due to buffer
+				 * exhaustion
+				 **/
+
+	/* firmware stats */
+	u64 fw_total_rcvd;
+	u64 fw_total_fwd;
+	u64 fw_total_fwd_bytes;
+	u64 fw_total_mcast;
+	u64 fw_total_bcast;
+
+	u64 fw_err_pko;
+	u64 fw_err_link;
+	u64 fw_err_drop;
+	u64 fw_rx_vxlan;
+	u64 fw_rx_vxlan_err;
+
+	/* LRO */
+	u64 fw_lro_pkts;   /* Number of packets that are LROed      */
+	u64 fw_lro_octs;   /* Number of octets that are LROed       */
+	u64 fw_total_lro;  /* Number of LRO packets formed          */
+	u64 fw_lro_aborts; /* Number of times LRO of packet aborted */
+	u64 fw_lro_aborts_port;
+	u64 fw_lro_aborts_seq;
+	u64 fw_lro_aborts_tsval;
+	u64 fw_lro_aborts_timer;	/* Timer setting error */
+	/* intrmod: packet forward rate */
+	u64 fwd_rate;
+};
+
+/** Stats for each NIC port in RX direction. */
+struct nic_tx_stats {
+	/* link-level stats */
+	u64 total_pkts_sent;		/* Total frames sent on the interface */
+	u64 total_bytes_sent;		/* Total octets sent on the interface */
+	u64 mcast_pkts_sent;		/* Packets sent to the multicast DMAC */
+	u64 bcast_pkts_sent;		/* Packets sent to a broadcast DMAC */
+	u64 ctl_sent;			/* Control/PAUSE packets sent */
+	u64 one_collision_sent;		/* Packets sent that experienced a
+					 * single collision before successful
+					 * transmission
+					 **/
+	u64 multi_collision_sent;	/* Packets sent that experienced
+					 * multiple collisions before successful
+					 * transmission
+					 **/
+	u64 max_collision_fail;		/* Packets dropped due to excessive
+					 * collisions
+					 **/
+	u64 max_deferral_fail;		/* Packets not sent due to max
+					 * deferrals
+					 **/
+	u64 fifo_err;			/* Packets sent that experienced a
+					 * transmit underflow and were
+					 * truncated
+					 **/
+	u64 runts;			/* Packets sent with an octet count
+					 * lessthan 64
+					 **/
+	u64 total_collisions;		/* Packets dropped due to excessive
+					 * collisions
+					 **/
+
+	/* firmware stats */
+	u64 fw_total_sent;
+	u64 fw_total_fwd;
+	u64 fw_total_fwd_bytes;
+	u64 fw_total_mcast_sent;
+	u64 fw_total_bcast_sent;
+	u64 fw_err_pko;
+	u64 fw_err_link;
+	u64 fw_err_drop;
+	u64 fw_err_tso;
+	u64 fw_tso;		/* number of tso requests */
+	u64 fw_tso_fwd;		/* number of packets segmented in tso */
+	u64 fw_tx_vxlan;
+	u64 fw_err_pki;
+};
+
+struct oct_link_stats {
+	struct nic_rx_stats fromwire;
+	struct nic_tx_stats fromhost;
+
+};
+
+static inline int opcode_slow_path(union octeon_rh *rh)
+{
+	u16 subcode1, subcode2;
+
+	subcode1 = OPCODE_SUBCODE((rh)->r.opcode, (rh)->r.subcode);
+	subcode2 = OPCODE_SUBCODE(OPCODE_NIC, OPCODE_NIC_NW_DATA);
+
+	return (subcode2 != subcode1);
+}
+
+#define LIO68XX_LED_CTRL_ADDR     0x3501
+#define LIO68XX_LED_CTRL_CFGON    0x1f
+#define LIO68XX_LED_CTRL_CFGOFF   0x100
+#define LIO68XX_LED_BEACON_ADDR   0x3508
+#define LIO68XX_LED_BEACON_CFGON  0x47fd
+#define LIO68XX_LED_BEACON_CFGOFF 0x11fc
+#define VITESSE_PHY_GPIO_DRIVEON  0x1
+#define VITESSE_PHY_GPIO_CFG      0x8
+#define VITESSE_PHY_GPIO_DRIVEOFF 0x4
+#define VITESSE_PHY_GPIO_HIGH     0x2
+#define VITESSE_PHY_GPIO_LOW      0x3
+#define LED_IDENTIFICATION_ON     0x1
+#define LED_IDENTIFICATION_OFF    0x0
+#define LIO23XX_COPPERHEAD_LED_GPIO 0x2
+
+struct oct_mdio_cmd {
+	u64 op;
+	u64 mdio_addr;
+	u64 value1;
+	u64 value2;
+	u64 value3;
+};
+
+#define OCT_LINK_STATS_SIZE   (sizeof(struct oct_link_stats))
+
+struct oct_intrmod_cfg {
+	u64 rx_enable;
+	u64 tx_enable;
+	u64 check_intrvl;
+	u64 maxpkt_ratethr;
+	u64 minpkt_ratethr;
+	u64 rx_maxcnt_trigger;
+	u64 rx_mincnt_trigger;
+	u64 rx_maxtmr_trigger;
+	u64 rx_mintmr_trigger;
+	u64 tx_mincnt_trigger;
+	u64 tx_maxcnt_trigger;
+	u64 rx_frames;
+	u64 tx_frames;
+	u64 rx_usecs;
+};
+
+#define BASE_QUEUE_NOT_REQUESTED 65535
+
+union oct_nic_if_cfg {
+	u64 u64;
+	struct {
+#ifdef __BIG_ENDIAN_BITFIELD
+		u64 base_queue:16;
+		u64 num_iqueues:16;
+		u64 num_oqueues:16;
+		u64 gmx_port_id:8;
+		u64 vf_id:8;
+#else
+		u64 vf_id:8;
+		u64 gmx_port_id:8;
+		u64 num_oqueues:16;
+		u64 num_iqueues:16;
+		u64 base_queue:16;
+#endif
+	} s;
+};
+
+struct lio_trusted_vf {
+	uint64_t active: 1;
+	uint64_t id : 8;
+	uint64_t reserved: 55;
+};
+
+struct lio_time {
+	s64 sec;   /* seconds */
+	s64 nsec;  /* nanoseconds */
+};
+
+struct lio_vf_rep_stats {
+	u64 tx_packets;
+	u64 tx_bytes;
+	u64 tx_dropped;
+
+	u64 rx_packets;
+	u64 rx_bytes;
+	u64 rx_dropped;
+};
+
+enum lio_vf_rep_req_type {
+	LIO_VF_REP_REQ_NONE,
+	LIO_VF_REP_REQ_STATE,
+	LIO_VF_REP_REQ_MTU,
+	LIO_VF_REP_REQ_STATS,
+	LIO_VF_REP_REQ_DEVNAME
+};
+
+enum {
+	LIO_VF_REP_STATE_DOWN,
+	LIO_VF_REP_STATE_UP
+};
+
+#define LIO_IF_NAME_SIZE 16
+struct lio_vf_rep_req {
+	u8 req_type;
+	u8 ifidx;
+	u8 rsvd[6];
+
+	union {
+		struct lio_vf_rep_name {
+			char name[LIO_IF_NAME_SIZE];
+		} rep_name;
+
+		struct lio_vf_rep_mtu {
+			u32 mtu;
+			u32 rsvd;
+		} rep_mtu;
+
+		struct lio_vf_rep_state {
+			u8 state;
+			u8 rsvd[7];
+		} rep_state;
+	};
+};
+
+struct lio_vf_rep_resp {
+	u64 rh;
+	u8  status;
+	u8  rsvd[7];
+};
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/liquidio_image.h b/drivers/net/ethernet/cavium/liquidio/liquidio_image.h
new file mode 100644
index 000000000..5bf5e8791
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/liquidio_image.h
@@ -0,0 +1,54 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#ifndef _LIQUIDIO_IMAGE_H_
+#define _LIQUIDIO_IMAGE_H_
+
+#define LIO_MAX_FW_TYPE_LEN     (8)
+#define LIO_MAX_FW_FILENAME_LEN (256)
+#define LIO_FW_DIR              "liquidio/"
+#define LIO_FW_BASE_NAME        "lio_"
+#define LIO_FW_NAME_SUFFIX      ".bin"
+#define LIO_FW_NAME_TYPE_NIC    "nic"
+#define LIO_FW_NAME_TYPE_AUTO   "auto"
+#define LIO_FW_NAME_TYPE_NONE   "none"
+#define LIO_MAX_FIRMWARE_VERSION_LEN 16
+
+#define LIO_MAX_BOOTCMD_LEN 1024
+#define LIO_MAX_IMAGES 16
+#define LIO_NIC_MAGIC 0x434E4943     /* "CNIC" */
+struct octeon_firmware_desc {
+	__be64 addr;
+	__be32 len;
+	__be32 crc32;         /* crc32 of image */
+};
+
+/* Following the header is a list of 64-bit aligned binary images,
+ * as described by the desc field.
+ * Numeric fields are in network byte order.
+ */
+struct octeon_firmware_file_header {
+	__be32 magic;
+	char version[LIO_MAX_FIRMWARE_VERSION_LEN];
+	char bootcmd[LIO_MAX_BOOTCMD_LEN];
+	__be32 num_images;
+	struct octeon_firmware_desc desc[LIO_MAX_IMAGES];
+	__be32 pad;
+	__be32 crc32;         /* header checksum */
+};
+
+#endif /* _LIQUIDIO_IMAGE_H_ */
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_config.h b/drivers/net/ethernet/cavium/liquidio/octeon_config.h
new file mode 100644
index 000000000..24c212001
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_config.h
@@ -0,0 +1,472 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*! \file  octeon_config.h
+ *  \brief Host Driver: Configuration data structures for the host driver.
+ */
+
+#ifndef __OCTEON_CONFIG_H__
+#define __OCTEON_CONFIG_H__
+
+/*--------------------------CONFIG VALUES------------------------*/
+
+/* The following macros affect the way the driver data structures
+ * are generated for Octeon devices.
+ * They can be modified.
+ */
+
+/* Maximum octeon devices defined as MAX_OCTEON_NICIF to support
+ * multiple(<= MAX_OCTEON_NICIF) Miniports
+ */
+#define   MAX_OCTEON_NICIF             128
+#define   MAX_OCTEON_DEVICES           MAX_OCTEON_NICIF
+#define   MAX_OCTEON_LINKS	       MAX_OCTEON_NICIF
+#define   MAX_OCTEON_MULTICAST_ADDR    32
+
+#define   MAX_OCTEON_FILL_COUNT        8
+
+/* CN6xxx IQ configuration macros */
+#define   CN6XXX_MAX_INPUT_QUEUES      32
+#define   CN6XXX_MAX_IQ_DESCRIPTORS    2048
+#define   CN6XXX_DB_MIN                1
+#define   CN6XXX_DB_MAX                8
+#define   CN6XXX_DB_TIMEOUT            1
+
+/* CN6xxx OQ configuration macros */
+#define   CN6XXX_MAX_OUTPUT_QUEUES     32
+#define   CN6XXX_MAX_OQ_DESCRIPTORS    2048
+#define   CN6XXX_OQ_BUF_SIZE           1664
+#define   CN6XXX_OQ_PKTSPER_INTR       ((CN6XXX_MAX_OQ_DESCRIPTORS < 512) ? \
+					(CN6XXX_MAX_OQ_DESCRIPTORS / 4) : 128)
+#define   CN6XXX_OQ_REFIL_THRESHOLD    ((CN6XXX_MAX_OQ_DESCRIPTORS < 512) ? \
+					(CN6XXX_MAX_OQ_DESCRIPTORS / 4) : 128)
+
+#define   CN6XXX_OQ_INTR_PKT           64
+#define   CN6XXX_OQ_INTR_TIME          100
+#define   DEFAULT_NUM_NIC_PORTS_66XX   2
+#define   DEFAULT_NUM_NIC_PORTS_68XX   4
+#define   DEFAULT_NUM_NIC_PORTS_68XX_210NV  2
+
+/* CN23xx  IQ configuration macros */
+#define   CN23XX_MAX_VFS_PER_PF_PASS_1_0 8
+#define   CN23XX_MAX_VFS_PER_PF_PASS_1_1 31
+#define   CN23XX_MAX_VFS_PER_PF          63
+#define   CN23XX_MAX_RINGS_PER_VF        8
+
+#define   CN23XX_MAX_RINGS_PER_PF_PASS_1_0 12
+#define   CN23XX_MAX_RINGS_PER_PF_PASS_1_1 32
+#define   CN23XX_MAX_RINGS_PER_PF          64
+#define   CN23XX_MAX_RINGS_PER_VF          8
+
+#define   CN23XX_MAX_INPUT_QUEUES	CN23XX_MAX_RINGS_PER_PF
+#define   CN23XX_MAX_IQ_DESCRIPTORS	2048
+#define   CN23XX_DEFAULT_IQ_DESCRIPTORS	512
+#define   CN23XX_MIN_IQ_DESCRIPTORS	128
+#define   CN23XX_DB_MIN                 1
+#define   CN23XX_DB_MAX                 8
+#define   CN23XX_DB_TIMEOUT             1
+
+#define   CN23XX_MAX_OUTPUT_QUEUES	CN23XX_MAX_RINGS_PER_PF
+#define   CN23XX_MAX_OQ_DESCRIPTORS	2048
+#define   CN23XX_DEFAULT_OQ_DESCRIPTORS	512
+#define   CN23XX_MIN_OQ_DESCRIPTORS	128
+#define   CN23XX_OQ_BUF_SIZE		1664
+#define   CN23XX_OQ_PKTSPER_INTR	128
+/*#define CAVIUM_ONLY_CN23XX_RX_PERF*/
+#define   CN23XX_OQ_REFIL_THRESHOLD	16
+
+#define   CN23XX_OQ_INTR_PKT		64
+#define   CN23XX_OQ_INTR_TIME		100
+#define   DEFAULT_NUM_NIC_PORTS_23XX	1
+
+#define   CN23XX_CFG_IO_QUEUES		CN23XX_MAX_RINGS_PER_PF
+/* PEMs count */
+#define   CN23XX_MAX_MACS		4
+
+#define   CN23XX_DEF_IQ_INTR_THRESHOLD	32
+#define   CN23XX_DEF_IQ_INTR_BYTE_THRESHOLD   (64 * 1024)
+/* common OCTEON configuration macros */
+#define   CN6XXX_CFG_IO_QUEUES         32
+#define   OCTEON_32BYTE_INSTR          32
+#define   OCTEON_64BYTE_INSTR          64
+#define   OCTEON_MAX_BASE_IOQ          4
+
+#define   OCTEON_DMA_INTR_PKT          64
+#define   OCTEON_DMA_INTR_TIME         1000
+
+#define MAX_TXQS_PER_INTF  8
+#define MAX_RXQS_PER_INTF  8
+#define DEF_TXQS_PER_INTF  4
+#define DEF_RXQS_PER_INTF  4
+
+#define INVALID_IOQ_NO          0xff
+
+#define   DEFAULT_POW_GRP       0
+
+/* Macros to get octeon config params */
+#define CFG_GET_IQ_CFG(cfg)                      ((cfg)->iq)
+#define CFG_GET_IQ_MAX_Q(cfg)                    ((cfg)->iq.max_iqs)
+#define CFG_GET_IQ_PENDING_LIST_SIZE(cfg)        ((cfg)->iq.pending_list_size)
+#define CFG_GET_IQ_INSTR_TYPE(cfg)               ((cfg)->iq.instr_type)
+#define CFG_GET_IQ_DB_MIN(cfg)                   ((cfg)->iq.db_min)
+#define CFG_GET_IQ_DB_TIMEOUT(cfg)               ((cfg)->iq.db_timeout)
+
+#define CFG_GET_IQ_INTR_PKT(cfg)                 ((cfg)->iq.iq_intr_pkt)
+#define CFG_SET_IQ_INTR_PKT(cfg, val)            (cfg)->iq.iq_intr_pkt = val
+
+#define CFG_GET_OQ_MAX_Q(cfg)                    ((cfg)->oq.max_oqs)
+#define CFG_GET_OQ_PKTS_PER_INTR(cfg)            ((cfg)->oq.pkts_per_intr)
+#define CFG_GET_OQ_REFILL_THRESHOLD(cfg)         ((cfg)->oq.refill_threshold)
+#define CFG_GET_OQ_INTR_PKT(cfg)                 ((cfg)->oq.oq_intr_pkt)
+#define CFG_GET_OQ_INTR_TIME(cfg)                ((cfg)->oq.oq_intr_time)
+#define CFG_SET_OQ_INTR_PKT(cfg, val)            (cfg)->oq.oq_intr_pkt = val
+#define CFG_SET_OQ_INTR_TIME(cfg, val)           (cfg)->oq.oq_intr_time = val
+
+#define CFG_GET_DMA_INTR_PKT(cfg)                ((cfg)->dma.dma_intr_pkt)
+#define CFG_GET_DMA_INTR_TIME(cfg)               ((cfg)->dma.dma_intr_time)
+#define CFG_GET_NUM_NIC_PORTS(cfg)               ((cfg)->num_nic_ports)
+#define CFG_GET_NUM_DEF_TX_DESCS(cfg)            ((cfg)->num_def_tx_descs)
+#define CFG_GET_NUM_DEF_RX_DESCS(cfg)            ((cfg)->num_def_rx_descs)
+#define CFG_GET_DEF_RX_BUF_SIZE(cfg)             ((cfg)->def_rx_buf_size)
+
+#define CFG_GET_MAX_TXQS_NIC_IF(cfg, idx) \
+				((cfg)->nic_if_cfg[idx].max_txqs)
+#define CFG_GET_NUM_TXQS_NIC_IF(cfg, idx) \
+				((cfg)->nic_if_cfg[idx].num_txqs)
+#define CFG_GET_MAX_RXQS_NIC_IF(cfg, idx) \
+				((cfg)->nic_if_cfg[idx].max_rxqs)
+#define CFG_GET_NUM_RXQS_NIC_IF(cfg, idx) \
+				((cfg)->nic_if_cfg[idx].num_rxqs)
+#define CFG_GET_NUM_RX_DESCS_NIC_IF(cfg, idx) \
+				((cfg)->nic_if_cfg[idx].num_rx_descs)
+#define CFG_GET_NUM_TX_DESCS_NIC_IF(cfg, idx) \
+				((cfg)->nic_if_cfg[idx].num_tx_descs)
+#define CFG_GET_NUM_RX_BUF_SIZE_NIC_IF(cfg, idx) \
+				((cfg)->nic_if_cfg[idx].rx_buf_size)
+#define CFG_GET_BASE_QUE_NIC_IF(cfg, idx) \
+				((cfg)->nic_if_cfg[idx].base_queue)
+#define CFG_GET_GMXID_NIC_IF(cfg, idx) \
+				((cfg)->nic_if_cfg[idx].gmx_port_id)
+
+#define CFG_GET_CTRL_Q_GRP(cfg)                  ((cfg)->misc.ctrlq_grp)
+#define CFG_GET_HOST_LINK_QUERY_INTERVAL(cfg) \
+				((cfg)->misc.host_link_query_interval)
+#define CFG_GET_OCT_LINK_QUERY_INTERVAL(cfg) \
+				((cfg)->misc.oct_link_query_interval)
+#define CFG_GET_IS_SLI_BP_ON(cfg)                ((cfg)->misc.enable_sli_oq_bp)
+
+#define CFG_SET_NUM_RX_DESCS_NIC_IF(cfg, idx, value) \
+				((cfg)->nic_if_cfg[idx].num_rx_descs = value)
+#define CFG_SET_NUM_TX_DESCS_NIC_IF(cfg, idx, value) \
+				((cfg)->nic_if_cfg[idx].num_tx_descs = value)
+
+/* Max IOQs per OCTEON Link */
+#define MAX_IOQS_PER_NICIF              64
+
+enum lio_card_type {
+	LIO_210SV = 0, /* Two port, 66xx */
+	LIO_210NV,     /* Two port, 68xx */
+	LIO_410NV,     /* Four port, 68xx */
+	LIO_23XX       /* 23xx */
+};
+
+#define LIO_210SV_NAME "210sv"
+#define LIO_210NV_NAME "210nv"
+#define LIO_410NV_NAME "410nv"
+#define LIO_23XX_NAME  "23xx"
+
+/** Structure to define the configuration attributes for each Input queue.
+ *  Applicable to all Octeon processors
+ **/
+struct octeon_iq_config {
+#ifdef __BIG_ENDIAN_BITFIELD
+	u64 reserved:16;
+
+	/** Tx interrupt packets. Applicable to 23xx only */
+	u64 iq_intr_pkt:16;
+
+	/** Minimum ticks to wait before checking for pending instructions. */
+	u64 db_timeout:16;
+
+	/** Minimum number of commands pending to be posted to Octeon
+	 *  before driver hits the Input queue doorbell.
+	 */
+	u64 db_min:8;
+
+	/** Command size - 32 or 64 bytes */
+	u64 instr_type:32;
+
+	/** Pending list size (usually set to the sum of the size of all Input
+	 *  queues)
+	 */
+	u64 pending_list_size:32;
+
+	/* Max number of IQs available */
+	u64 max_iqs:8;
+#else
+	/* Max number of IQs available */
+	u64 max_iqs:8;
+
+	/** Pending list size (usually set to the sum of the size of all Input
+	 *  queues)
+	 */
+	u64 pending_list_size:32;
+
+	/** Command size - 32 or 64 bytes */
+	u64 instr_type:32;
+
+	/** Minimum number of commands pending to be posted to Octeon
+	 *  before driver hits the Input queue doorbell.
+	 */
+	u64 db_min:8;
+
+	/** Minimum ticks to wait before checking for pending instructions. */
+	u64 db_timeout:16;
+
+	/** Tx interrupt packets. Applicable to 23xx only */
+	u64 iq_intr_pkt:16;
+
+	u64 reserved:16;
+#endif
+};
+
+/** Structure to define the configuration attributes for each Output queue.
+ *  Applicable to all Octeon processors
+ **/
+struct octeon_oq_config {
+#ifdef __BIG_ENDIAN_BITFIELD
+	u64 reserved:16;
+
+	u64 pkts_per_intr:16;
+
+	/** Interrupt Coalescing (Time Interval). Octeon will interrupt the
+	 *  host if atleast one packet was sent in the time interval specified
+	 *  by this field. The driver uses time interval interrupt coalescing
+	 *  by default. The time is specified in microseconds.
+	 */
+	u64 oq_intr_time:16;
+
+	/** Interrupt Coalescing (Packet Count). Octeon will interrupt the host
+	 *  only if it sent as many packets as specified by this field.
+	 *  The driver
+	 *  usually does not use packet count interrupt coalescing.
+	 */
+	u64 oq_intr_pkt:16;
+
+	/** The number of buffers that were consumed during packet processing by
+	 *   the driver on this Output queue before the driver attempts to
+	 *   replenish
+	 *   the descriptor ring with new buffers.
+	 */
+	u64 refill_threshold:16;
+
+	/* Max number of OQs available */
+	u64 max_oqs:8;
+
+#else
+	/* Max number of OQs available */
+	u64 max_oqs:8;
+
+	/** The number of buffers that were consumed during packet processing by
+	 *   the driver on this Output queue before the driver attempts to
+	 *   replenish
+	 *   the descriptor ring with new buffers.
+	 */
+	u64 refill_threshold:16;
+
+	/** Interrupt Coalescing (Packet Count). Octeon will interrupt the host
+	 *  only if it sent as many packets as specified by this field.
+	 *  The driver
+	 *  usually does not use packet count interrupt coalescing.
+	 */
+	u64 oq_intr_pkt:16;
+
+	/** Interrupt Coalescing (Time Interval). Octeon will interrupt the
+	 *  host if atleast one packet was sent in the time interval specified
+	 *  by this field. The driver uses time interval interrupt coalescing
+	 *  by default.  The time is specified in microseconds.
+	 */
+	u64 oq_intr_time:16;
+
+	u64 pkts_per_intr:16;
+
+	u64 reserved:16;
+#endif
+
+};
+
+/** This structure conatins the NIC link configuration attributes,
+ *  common for all the OCTEON Modles.
+ */
+struct octeon_nic_if_config {
+#ifdef __BIG_ENDIAN_BITFIELD
+	u64 reserved:56;
+
+	u64 base_queue:16;
+
+	u64 gmx_port_id:8;
+
+	/* SKB size, We need not change buf size even for Jumbo frames.
+	 * Octeon can send jumbo frames in 4 consecutive descriptors,
+	 */
+	u64 rx_buf_size:16;
+
+	/* Num of desc for tx rings */
+	u64 num_tx_descs:16;
+
+	/* Num of desc for rx rings */
+	u64 num_rx_descs:16;
+
+	/* Actual configured value. Range could be: 1...max_rxqs */
+	u64 num_rxqs:16;
+
+	/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+	u64 max_rxqs:16;
+
+	/* Actual configured value. Range could be: 1...max_txqs */
+	u64 num_txqs:16;
+
+	/* Max Txqs: Half for each of the two ports :max_iq/2 */
+	u64 max_txqs:16;
+#else
+	/* Max Txqs: Half for each of the two ports :max_iq/2 */
+	u64 max_txqs:16;
+
+	/* Actual configured value. Range could be: 1...max_txqs */
+	u64 num_txqs:16;
+
+	/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+	u64 max_rxqs:16;
+
+	/* Actual configured value. Range could be: 1...max_rxqs */
+	u64 num_rxqs:16;
+
+	/* Num of desc for rx rings */
+	u64 num_rx_descs:16;
+
+	/* Num of desc for tx rings */
+	u64 num_tx_descs:16;
+
+	/* SKB size, We need not change buf size even for Jumbo frames.
+	 * Octeon can send jumbo frames in 4 consecutive descriptors,
+	 */
+	u64 rx_buf_size:16;
+
+	u64 gmx_port_id:8;
+
+	u64 base_queue:16;
+
+	u64 reserved:56;
+#endif
+
+};
+
+/** Structure to define the configuration attributes for meta data.
+ *  Applicable to all Octeon processors.
+ */
+
+struct octeon_misc_config {
+#ifdef __BIG_ENDIAN_BITFIELD
+	/** Host link status polling period */
+	u64 host_link_query_interval:32;
+	/** Oct link status polling period */
+	u64 oct_link_query_interval:32;
+
+	u64 enable_sli_oq_bp:1;
+	/** Control IQ Group */
+	u64 ctrlq_grp:4;
+#else
+	/** Control IQ Group */
+	u64 ctrlq_grp:4;
+	/** BP for SLI OQ */
+	u64 enable_sli_oq_bp:1;
+	/** Host link status polling period */
+	u64 oct_link_query_interval:32;
+	/** Oct link status polling period */
+	u64 host_link_query_interval:32;
+#endif
+};
+
+/** Structure to define the configuration for all OCTEON processors. */
+struct octeon_config {
+	u16 card_type;
+	char *card_name;
+
+	/** Input Queue attributes. */
+	struct octeon_iq_config iq;
+
+	/** Output Queue attributes. */
+	struct octeon_oq_config oq;
+
+	/** NIC Port Configuration */
+	struct octeon_nic_if_config nic_if_cfg[MAX_OCTEON_NICIF];
+
+	/** Miscellaneous attributes */
+	struct octeon_misc_config misc;
+
+	int num_nic_ports;
+
+	int num_def_tx_descs;
+
+	/* Num of desc for rx rings */
+	int num_def_rx_descs;
+
+	int def_rx_buf_size;
+
+};
+
+/* The following config values are fixed and should not be modified. */
+
+#define  BAR1_INDEX_DYNAMIC_MAP          2
+#define  BAR1_INDEX_STATIC_MAP          15
+#define  OCTEON_BAR1_ENTRY_SIZE         (4 * 1024 * 1024)
+
+#define  MAX_BAR1_IOREMAP_SIZE  (16 * OCTEON_BAR1_ENTRY_SIZE)
+
+/* Response lists - 1 ordered, 1 unordered-blocking, 1 unordered-nonblocking
+ *                  1 process done list, 1 zombie lists(timeouted sc list)
+ * NoResponse Lists are now maintained with each IQ. (Dec' 2007).
+ */
+#define MAX_RESPONSE_LISTS           6
+
+/* Opcode hash bits. The opcode is hashed on the lower 6-bits to lookup the
+ * dispatch table.
+ */
+#define OPCODE_MASK_BITS             6
+
+/* Mask for the 6-bit lookup hash */
+#define OCTEON_OPCODE_MASK           0x3f
+
+/* Size of the dispatch table. The 6-bit hash can index into 2^6 entries */
+#define DISPATCH_LIST_SIZE                      BIT(OPCODE_MASK_BITS)
+
+/* Maximum number of Octeon Instruction (command) queues */
+#define MAX_OCTEON_INSTR_QUEUES(oct)		\
+		(OCTEON_CN23XX_PF(oct) ? CN23XX_MAX_INPUT_QUEUES : \
+					CN6XXX_MAX_INPUT_QUEUES)
+
+/* Maximum number of Octeon Instruction (command) queues */
+#define MAX_OCTEON_OUTPUT_QUEUES(oct)		\
+		(OCTEON_CN23XX_PF(oct) ? CN23XX_MAX_OUTPUT_QUEUES : \
+					CN6XXX_MAX_OUTPUT_QUEUES)
+
+#define MAX_POSSIBLE_OCTEON_INSTR_QUEUES	CN23XX_MAX_INPUT_QUEUES
+#define MAX_POSSIBLE_OCTEON_OUTPUT_QUEUES	CN23XX_MAX_OUTPUT_QUEUES
+
+#define MAX_POSSIBLE_VFS			64
+
+#endif /* __OCTEON_CONFIG_H__  */
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_console.c b/drivers/net/ethernet/cavium/liquidio/octeon_console.c
new file mode 100644
index 000000000..28feabec8
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_console.c
@@ -0,0 +1,920 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*
+ * @file octeon_console.c
+ */
+#include <linux/moduleparam.h>
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include <linux/crc32.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "liquidio_image.h"
+#include "octeon_mem_ops.h"
+
+static void octeon_remote_lock(void);
+static void octeon_remote_unlock(void);
+static u64 cvmx_bootmem_phy_named_block_find(struct octeon_device *oct,
+					     const char *name,
+					     u32 flags);
+static int octeon_console_read(struct octeon_device *oct, u32 console_num,
+			       char *buffer, u32 buf_size);
+
+#define BOOTLOADER_PCI_READ_BUFFER_DATA_ADDR    0x0006c008
+#define BOOTLOADER_PCI_READ_BUFFER_LEN_ADDR     0x0006c004
+#define BOOTLOADER_PCI_READ_BUFFER_OWNER_ADDR   0x0006c000
+#define BOOTLOADER_PCI_READ_DESC_ADDR           0x0006c100
+#define BOOTLOADER_PCI_WRITE_BUFFER_STR_LEN     248
+
+#define OCTEON_PCI_IO_BUF_OWNER_OCTEON    0x00000001
+#define OCTEON_PCI_IO_BUF_OWNER_HOST      0x00000002
+
+/** Can change without breaking ABI */
+#define CVMX_BOOTMEM_NUM_NAMED_BLOCKS 64
+
+/** minimum alignment of bootmem alloced blocks */
+#define CVMX_BOOTMEM_ALIGNMENT_SIZE     (16ull)
+
+/** CVMX bootmem descriptor major version */
+#define CVMX_BOOTMEM_DESC_MAJ_VER   3
+/* CVMX bootmem descriptor minor version */
+#define CVMX_BOOTMEM_DESC_MIN_VER   0
+
+/* Current versions */
+#define OCTEON_PCI_CONSOLE_MAJOR_VERSION    1
+#define OCTEON_PCI_CONSOLE_MINOR_VERSION    0
+#define OCTEON_PCI_CONSOLE_BLOCK_NAME   "__pci_console"
+#define OCTEON_CONSOLE_POLL_INTERVAL_MS  100    /* 10 times per second */
+
+/* First three members of cvmx_bootmem_desc are left in original
+ * positions for backwards compatibility.
+ * Assumes big endian target
+ */
+struct cvmx_bootmem_desc {
+	/** spinlock to control access to list */
+	u32 lock;
+
+	/** flags for indicating various conditions */
+	u32 flags;
+
+	u64 head_addr;
+
+	/** incremented changed when incompatible changes made */
+	u32 major_version;
+
+	/** incremented changed when compatible changes made,
+	 *  reset to zero when major incremented
+	 */
+	u32 minor_version;
+
+	u64 app_data_addr;
+	u64 app_data_size;
+
+	/** number of elements in named blocks array */
+	u32 nb_num_blocks;
+
+	/** length of name array in bootmem blocks */
+	u32 named_block_name_len;
+
+	/** address of named memory block descriptors */
+	u64 named_block_array_addr;
+};
+
+/* Structure that defines a single console.
+ *
+ * Note: when read_index == write_index, the buffer is empty.
+ * The actual usable size of each console is console_buf_size -1;
+ */
+struct octeon_pci_console {
+	u64 input_base_addr;
+	u32 input_read_index;
+	u32 input_write_index;
+	u64 output_base_addr;
+	u32 output_read_index;
+	u32 output_write_index;
+	u32 lock;
+	u32 buf_size;
+};
+
+/* This is the main container structure that contains all the information
+ * about all PCI consoles.  The address of this structure is passed to various
+ * routines that operation on PCI consoles.
+ */
+struct octeon_pci_console_desc {
+	u32 major_version;
+	u32 minor_version;
+	u32 lock;
+	u32 flags;
+	u32 num_consoles;
+	u32 pad;
+	/* must be 64 bit aligned here... */
+	/* Array of addresses of octeon_pci_console structures */
+	u64 console_addr_array[];
+	/* Implicit storage for console_addr_array */
+};
+
+/*
+ * This function is the implementation of the get macros defined
+ * for individual structure members. The argument are generated
+ * by the macros inorder to read only the needed memory.
+ *
+ * @param oct    Pointer to current octeon device
+ * @param base   64bit physical address of the complete structure
+ * @param offset Offset from the beginning of the structure to the member being
+ *               accessed.
+ * @param size   Size of the structure member.
+ *
+ * @return Value of the structure member promoted into a u64.
+ */
+static inline u64 __cvmx_bootmem_desc_get(struct octeon_device *oct,
+					  u64 base,
+					  u32 offset,
+					  u32 size)
+{
+	base = (1ull << 63) | (base + offset);
+	switch (size) {
+	case 4:
+		return octeon_read_device_mem32(oct, base);
+	case 8:
+		return octeon_read_device_mem64(oct, base);
+	default:
+		return 0;
+	}
+}
+
+/*
+ * This function retrieves the string name of a named block. It is
+ * more complicated than a simple memcpy() since the named block
+ * descriptor may not be directly accessible.
+ *
+ * @param addr   Physical address of the named block descriptor
+ * @param str    String to receive the named block string name
+ * @param len    Length of the string buffer, which must match the length
+ *               stored in the bootmem descriptor.
+ */
+static void CVMX_BOOTMEM_NAMED_GET_NAME(struct octeon_device *oct,
+					u64 addr,
+					char *str,
+					u32 len)
+{
+	addr += offsetof(struct cvmx_bootmem_named_block_desc, name);
+	octeon_pci_read_core_mem(oct, addr, (u8 *)str, len);
+	str[len] = 0;
+}
+
+/* See header file for descriptions of functions */
+
+/*
+ * Check the version information on the bootmem descriptor
+ *
+ * @param exact_match
+ *               Exact major version to check against. A zero means
+ *               check that the version supports named blocks.
+ *
+ * @return Zero if the version is correct. Negative if the version is
+ *         incorrect. Failures also cause a message to be displayed.
+ */
+static int __cvmx_bootmem_check_version(struct octeon_device *oct,
+					u32 exact_match)
+{
+	u32 major_version;
+	u32 minor_version;
+
+	if (!oct->bootmem_desc_addr)
+		oct->bootmem_desc_addr =
+			octeon_read_device_mem64(oct,
+						 BOOTLOADER_PCI_READ_DESC_ADDR);
+	major_version = (u32)__cvmx_bootmem_desc_get(
+			oct, oct->bootmem_desc_addr,
+			offsetof(struct cvmx_bootmem_desc, major_version),
+			sizeof_field(struct cvmx_bootmem_desc, major_version));
+	minor_version = (u32)__cvmx_bootmem_desc_get(
+			oct, oct->bootmem_desc_addr,
+			offsetof(struct cvmx_bootmem_desc, minor_version),
+			sizeof_field(struct cvmx_bootmem_desc, minor_version));
+
+	dev_dbg(&oct->pci_dev->dev, "%s: major_version=%d\n", __func__,
+		major_version);
+	if ((major_version > 3) ||
+	    (exact_match && major_version != exact_match)) {
+		dev_err(&oct->pci_dev->dev, "bootmem ver mismatch %d.%d addr:0x%llx\n",
+			major_version, minor_version,
+			(long long)oct->bootmem_desc_addr);
+		return -1;
+	} else {
+		return 0;
+	}
+}
+
+static const struct cvmx_bootmem_named_block_desc
+*__cvmx_bootmem_find_named_block_flags(struct octeon_device *oct,
+					const char *name, u32 flags)
+{
+	struct cvmx_bootmem_named_block_desc *desc =
+		&oct->bootmem_named_block_desc;
+	u64 named_addr = cvmx_bootmem_phy_named_block_find(oct, name, flags);
+
+	if (named_addr) {
+		desc->base_addr = __cvmx_bootmem_desc_get(
+				oct, named_addr,
+				offsetof(struct cvmx_bootmem_named_block_desc,
+					 base_addr),
+				sizeof_field(
+					struct cvmx_bootmem_named_block_desc,
+					base_addr));
+		desc->size = __cvmx_bootmem_desc_get(oct, named_addr,
+				offsetof(struct cvmx_bootmem_named_block_desc,
+					 size),
+				sizeof_field(
+					struct cvmx_bootmem_named_block_desc,
+					size));
+
+		strncpy(desc->name, name, sizeof(desc->name));
+		desc->name[sizeof(desc->name) - 1] = 0;
+		return &oct->bootmem_named_block_desc;
+	} else {
+		return NULL;
+	}
+}
+
+static u64 cvmx_bootmem_phy_named_block_find(struct octeon_device *oct,
+					     const char *name,
+					     u32 flags)
+{
+	u64 result = 0;
+
+	if (!__cvmx_bootmem_check_version(oct, 3)) {
+		u32 i;
+
+		u64 named_block_array_addr = __cvmx_bootmem_desc_get(
+					oct, oct->bootmem_desc_addr,
+					offsetof(struct cvmx_bootmem_desc,
+						 named_block_array_addr),
+					sizeof_field(struct cvmx_bootmem_desc,
+						     named_block_array_addr));
+		u32 num_blocks = (u32)__cvmx_bootmem_desc_get(
+					oct, oct->bootmem_desc_addr,
+					offsetof(struct cvmx_bootmem_desc,
+						 nb_num_blocks),
+					sizeof_field(struct cvmx_bootmem_desc,
+						     nb_num_blocks));
+
+		u32 name_length = (u32)__cvmx_bootmem_desc_get(
+					oct, oct->bootmem_desc_addr,
+					offsetof(struct cvmx_bootmem_desc,
+						 named_block_name_len),
+					sizeof_field(struct cvmx_bootmem_desc,
+						     named_block_name_len));
+
+		u64 named_addr = named_block_array_addr;
+
+		for (i = 0; i < num_blocks; i++) {
+			u64 named_size = __cvmx_bootmem_desc_get(
+					oct, named_addr,
+					 offsetof(
+					struct cvmx_bootmem_named_block_desc,
+					size),
+					 sizeof_field(
+					struct cvmx_bootmem_named_block_desc,
+					size));
+
+			if (name && named_size) {
+				char *name_tmp =
+					kmalloc(name_length + 1, GFP_KERNEL);
+				if (!name_tmp)
+					break;
+
+				CVMX_BOOTMEM_NAMED_GET_NAME(oct, named_addr,
+							    name_tmp,
+							    name_length);
+				if (!strncmp(name, name_tmp, name_length)) {
+					result = named_addr;
+					kfree(name_tmp);
+					break;
+				}
+				kfree(name_tmp);
+			} else if (!name && !named_size) {
+				result = named_addr;
+				break;
+			}
+
+			named_addr +=
+				sizeof(struct cvmx_bootmem_named_block_desc);
+		}
+	}
+	return result;
+}
+
+/*
+ * Find a named block on the remote Octeon
+ *
+ * @param name      Name of block to find
+ * @param base_addr Address the block is at (OUTPUT)
+ * @param size      The size of the block (OUTPUT)
+ *
+ * @return Zero on success, One on failure.
+ */
+static int octeon_named_block_find(struct octeon_device *oct, const char *name,
+				   u64 *base_addr, u64 *size)
+{
+	const struct cvmx_bootmem_named_block_desc *named_block;
+
+	octeon_remote_lock();
+	named_block = __cvmx_bootmem_find_named_block_flags(oct, name, 0);
+	octeon_remote_unlock();
+	if (named_block) {
+		*base_addr = named_block->base_addr;
+		*size = named_block->size;
+		return 0;
+	}
+	return 1;
+}
+
+static void octeon_remote_lock(void)
+{
+	/* fill this in if any sharing is needed */
+}
+
+static void octeon_remote_unlock(void)
+{
+	/* fill this in if any sharing is needed */
+}
+
+int octeon_console_send_cmd(struct octeon_device *oct, char *cmd_str,
+			    u32 wait_hundredths)
+{
+	u32 len = (u32)strlen(cmd_str);
+
+	dev_dbg(&oct->pci_dev->dev, "sending \"%s\" to bootloader\n", cmd_str);
+
+	if (len > BOOTLOADER_PCI_WRITE_BUFFER_STR_LEN - 1) {
+		dev_err(&oct->pci_dev->dev, "Command string too long, max length is: %d\n",
+			BOOTLOADER_PCI_WRITE_BUFFER_STR_LEN - 1);
+		return -1;
+	}
+
+	if (octeon_wait_for_bootloader(oct, wait_hundredths) != 0) {
+		dev_err(&oct->pci_dev->dev, "Bootloader not ready for command.\n");
+		return -1;
+	}
+
+	/* Write command to bootloader */
+	octeon_remote_lock();
+	octeon_pci_write_core_mem(oct, BOOTLOADER_PCI_READ_BUFFER_DATA_ADDR,
+				  (u8 *)cmd_str, len);
+	octeon_write_device_mem32(oct, BOOTLOADER_PCI_READ_BUFFER_LEN_ADDR,
+				  len);
+	octeon_write_device_mem32(oct, BOOTLOADER_PCI_READ_BUFFER_OWNER_ADDR,
+				  OCTEON_PCI_IO_BUF_OWNER_OCTEON);
+
+	/* Bootloader should accept command very quickly
+	 * if it really was ready
+	 */
+	if (octeon_wait_for_bootloader(oct, 200) != 0) {
+		octeon_remote_unlock();
+		dev_err(&oct->pci_dev->dev, "Bootloader did not accept command.\n");
+		return -1;
+	}
+	octeon_remote_unlock();
+	return 0;
+}
+
+int octeon_wait_for_bootloader(struct octeon_device *oct,
+			       u32 wait_time_hundredths)
+{
+	dev_dbg(&oct->pci_dev->dev, "waiting %d0 ms for bootloader\n",
+		wait_time_hundredths);
+
+	if (octeon_mem_access_ok(oct))
+		return -1;
+
+	while (wait_time_hundredths > 0 &&
+	       octeon_read_device_mem32(oct,
+					BOOTLOADER_PCI_READ_BUFFER_OWNER_ADDR)
+	       != OCTEON_PCI_IO_BUF_OWNER_HOST) {
+		if (--wait_time_hundredths <= 0)
+			return -1;
+		schedule_timeout_uninterruptible(HZ / 100);
+	}
+	return 0;
+}
+
+static void octeon_console_handle_result(struct octeon_device *oct,
+					 size_t console_num)
+{
+	struct octeon_console *console;
+
+	console = &oct->console[console_num];
+
+	console->waiting = 0;
+}
+
+static char console_buffer[OCTEON_CONSOLE_MAX_READ_BYTES];
+
+static void output_console_line(struct octeon_device *oct,
+				struct octeon_console *console,
+				size_t console_num,
+				char *console_buffer,
+				s32 bytes_read)
+{
+	char *line;
+	s32 i;
+	size_t len;
+
+	line = console_buffer;
+	for (i = 0; i < bytes_read; i++) {
+		/* Output a line at a time, prefixed */
+		if (console_buffer[i] == '\n') {
+			console_buffer[i] = '\0';
+			/* We need to output 'line', prefaced by 'leftover'.
+			 * However, it is possible we're being called to
+			 * output 'leftover' by itself (in the case of nothing
+			 * having been read from the console).
+			 *
+			 * To avoid duplication, check for this condition.
+			 */
+			if (console->leftover[0] &&
+			    (line != console->leftover)) {
+				if (console->print)
+					(*console->print)(oct, (u32)console_num,
+							  console->leftover,
+							  line);
+				console->leftover[0] = '\0';
+			} else {
+				if (console->print)
+					(*console->print)(oct, (u32)console_num,
+							  line, NULL);
+			}
+			line = &console_buffer[i + 1];
+		}
+	}
+
+	/* Save off any leftovers */
+	if (line != &console_buffer[bytes_read]) {
+		console_buffer[bytes_read] = '\0';
+		len = strlen(console->leftover);
+		strncpy(&console->leftover[len], line,
+			sizeof(console->leftover) - len);
+	}
+}
+
+static void check_console(struct work_struct *work)
+{
+	s32 bytes_read, tries, total_read;
+	size_t len;
+	struct octeon_console *console;
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct octeon_device *oct = (struct octeon_device *)wk->ctxptr;
+	u32 console_num = (u32)wk->ctxul;
+	u32 delay;
+
+	console = &oct->console[console_num];
+	tries = 0;
+	total_read = 0;
+
+	do {
+		/* Take console output regardless of whether it will
+		 * be logged
+		 */
+		bytes_read =
+			octeon_console_read(oct, console_num, console_buffer,
+					    sizeof(console_buffer) - 1);
+		if (bytes_read > 0) {
+			total_read += bytes_read;
+			if (console->waiting)
+				octeon_console_handle_result(oct, console_num);
+			if (console->print) {
+				output_console_line(oct, console, console_num,
+						    console_buffer, bytes_read);
+			}
+		} else if (bytes_read < 0) {
+			dev_err(&oct->pci_dev->dev, "Error reading console %u, ret=%d\n",
+				console_num, bytes_read);
+		}
+
+		tries++;
+	} while ((bytes_read > 0) && (tries < 16));
+
+	/* If nothing is read after polling the console,
+	 * output any leftovers if any
+	 */
+	if (console->print && (total_read == 0) &&
+	    (console->leftover[0])) {
+		/* append '\n' as terminator for 'output_console_line' */
+		len = strlen(console->leftover);
+		console->leftover[len] = '\n';
+		output_console_line(oct, console, console_num,
+				    console->leftover, (s32)(len + 1));
+		console->leftover[0] = '\0';
+	}
+
+	delay = OCTEON_CONSOLE_POLL_INTERVAL_MS;
+
+	schedule_delayed_work(&wk->work, msecs_to_jiffies(delay));
+}
+
+int octeon_init_consoles(struct octeon_device *oct)
+{
+	int ret = 0;
+	u64 addr, size;
+
+	ret = octeon_mem_access_ok(oct);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "Memory access not okay'\n");
+		return ret;
+	}
+
+	ret = octeon_named_block_find(oct, OCTEON_PCI_CONSOLE_BLOCK_NAME, &addr,
+				      &size);
+	if (ret) {
+		dev_err(&oct->pci_dev->dev, "Could not find console '%s'\n",
+			OCTEON_PCI_CONSOLE_BLOCK_NAME);
+		return ret;
+	}
+
+	/* Dedicate one of Octeon's BAR1 index registers to create a static
+	 * mapping to a region of Octeon DRAM that contains the PCI console
+	 * named block.
+	 */
+	oct->console_nb_info.bar1_index = BAR1_INDEX_STATIC_MAP;
+	oct->fn_list.bar1_idx_setup(oct, addr, oct->console_nb_info.bar1_index,
+				    true);
+	oct->console_nb_info.dram_region_base = addr
+		& ~(OCTEON_BAR1_ENTRY_SIZE - 1ULL);
+
+	/* num_consoles > 0, is an indication that the consoles
+	 * are accessible
+	 */
+	oct->num_consoles = octeon_read_device_mem32(oct,
+		addr + offsetof(struct octeon_pci_console_desc,
+			num_consoles));
+	oct->console_desc_addr = addr;
+
+	dev_dbg(&oct->pci_dev->dev, "Initialized consoles. %d available\n",
+		oct->num_consoles);
+
+	return ret;
+}
+
+static void octeon_get_uboot_version(struct octeon_device *oct)
+{
+	s32 bytes_read, tries, total_read;
+	struct octeon_console *console;
+	u32 console_num = 0;
+	char *uboot_ver;
+	char *buf;
+	char *p;
+
+#define OCTEON_UBOOT_VER_BUF_SIZE 512
+	buf = kmalloc(OCTEON_UBOOT_VER_BUF_SIZE, GFP_KERNEL);
+	if (!buf)
+		return;
+
+	if (octeon_console_send_cmd(oct, "setenv stdout pci\n", 50)) {
+		kfree(buf);
+		return;
+	}
+
+	if (octeon_console_send_cmd(oct, "version\n", 1)) {
+		kfree(buf);
+		return;
+	}
+
+	console = &oct->console[console_num];
+	tries = 0;
+	total_read = 0;
+
+	do {
+		/* Take console output regardless of whether it will
+		 * be logged
+		 */
+		bytes_read =
+			octeon_console_read(oct,
+					    console_num, buf + total_read,
+					    OCTEON_UBOOT_VER_BUF_SIZE - 1 -
+					    total_read);
+		if (bytes_read > 0) {
+			buf[bytes_read] = '\0';
+
+			total_read += bytes_read;
+			if (console->waiting)
+				octeon_console_handle_result(oct, console_num);
+		} else if (bytes_read < 0) {
+			dev_err(&oct->pci_dev->dev, "Error reading console %u, ret=%d\n",
+				console_num, bytes_read);
+		}
+
+		tries++;
+	} while ((bytes_read > 0) && (tries < 16));
+
+	/* If nothing is read after polling the console,
+	 * output any leftovers if any
+	 */
+	if ((total_read == 0) && (console->leftover[0])) {
+		dev_dbg(&oct->pci_dev->dev, "%u: %s\n",
+			console_num, console->leftover);
+		console->leftover[0] = '\0';
+	}
+
+	buf[OCTEON_UBOOT_VER_BUF_SIZE - 1] = '\0';
+
+	uboot_ver = strstr(buf, "U-Boot");
+	if (uboot_ver) {
+		p = strstr(uboot_ver, "mips");
+		if (p) {
+			p--;
+			*p = '\0';
+			dev_info(&oct->pci_dev->dev, "%s\n", uboot_ver);
+		}
+	}
+
+	kfree(buf);
+	octeon_console_send_cmd(oct, "setenv stdout serial\n", 50);
+}
+
+int octeon_add_console(struct octeon_device *oct, u32 console_num,
+		       char *dbg_enb)
+{
+	int ret = 0;
+	u32 delay;
+	u64 coreaddr;
+	struct delayed_work *work;
+	struct octeon_console *console;
+
+	if (console_num >= oct->num_consoles) {
+		dev_err(&oct->pci_dev->dev,
+			"trying to read from console number %d when only 0 to %d exist\n",
+			console_num, oct->num_consoles);
+	} else {
+		console = &oct->console[console_num];
+
+		console->waiting = 0;
+
+		coreaddr = oct->console_desc_addr + console_num * 8 +
+			offsetof(struct octeon_pci_console_desc,
+				 console_addr_array);
+		console->addr = octeon_read_device_mem64(oct, coreaddr);
+		coreaddr = console->addr + offsetof(struct octeon_pci_console,
+						    buf_size);
+		console->buffer_size = octeon_read_device_mem32(oct, coreaddr);
+		coreaddr = console->addr + offsetof(struct octeon_pci_console,
+						    input_base_addr);
+		console->input_base_addr =
+			octeon_read_device_mem64(oct, coreaddr);
+		coreaddr = console->addr + offsetof(struct octeon_pci_console,
+						    output_base_addr);
+		console->output_base_addr =
+			octeon_read_device_mem64(oct, coreaddr);
+		console->leftover[0] = '\0';
+
+		work = &oct->console_poll_work[console_num].work;
+
+		octeon_get_uboot_version(oct);
+
+		INIT_DELAYED_WORK(work, check_console);
+		oct->console_poll_work[console_num].ctxptr = (void *)oct;
+		oct->console_poll_work[console_num].ctxul = console_num;
+		delay = OCTEON_CONSOLE_POLL_INTERVAL_MS;
+		schedule_delayed_work(work, msecs_to_jiffies(delay));
+
+		/* an empty string means use default debug console enablement */
+		if (dbg_enb && !dbg_enb[0])
+			dbg_enb = "setenv pci_console_active 1";
+		if (dbg_enb)
+			ret = octeon_console_send_cmd(oct, dbg_enb, 2000);
+
+		console->active = 1;
+	}
+
+	return ret;
+}
+
+/*
+ * Removes all consoles
+ *
+ * @param oct         octeon device
+ */
+void octeon_remove_consoles(struct octeon_device *oct)
+{
+	u32 i;
+	struct octeon_console *console;
+
+	for (i = 0; i < oct->num_consoles; i++) {
+		console = &oct->console[i];
+
+		if (!console->active)
+			continue;
+
+		cancel_delayed_work_sync(&oct->console_poll_work[i].
+						work);
+		console->addr = 0;
+		console->buffer_size = 0;
+		console->input_base_addr = 0;
+		console->output_base_addr = 0;
+	}
+
+	oct->num_consoles = 0;
+}
+
+static inline int octeon_console_free_bytes(u32 buffer_size,
+					    u32 wr_idx,
+					    u32 rd_idx)
+{
+	if (rd_idx >= buffer_size || wr_idx >= buffer_size)
+		return -1;
+
+	return ((buffer_size - 1) - (wr_idx - rd_idx)) % buffer_size;
+}
+
+static inline int octeon_console_avail_bytes(u32 buffer_size,
+					     u32 wr_idx,
+					     u32 rd_idx)
+{
+	if (rd_idx >= buffer_size || wr_idx >= buffer_size)
+		return -1;
+
+	return buffer_size - 1 -
+	       octeon_console_free_bytes(buffer_size, wr_idx, rd_idx);
+}
+
+static int octeon_console_read(struct octeon_device *oct, u32 console_num,
+			       char *buffer, u32 buf_size)
+{
+	int bytes_to_read;
+	u32 rd_idx, wr_idx;
+	struct octeon_console *console;
+
+	if (console_num >= oct->num_consoles) {
+		dev_err(&oct->pci_dev->dev, "Attempted to read from disabled console %d\n",
+			console_num);
+		return 0;
+	}
+
+	console = &oct->console[console_num];
+
+	/* Check to see if any data is available.
+	 * Maybe optimize this with 64-bit read.
+	 */
+	rd_idx = octeon_read_device_mem32(oct, console->addr +
+		offsetof(struct octeon_pci_console, output_read_index));
+	wr_idx = octeon_read_device_mem32(oct, console->addr +
+		offsetof(struct octeon_pci_console, output_write_index));
+
+	bytes_to_read = octeon_console_avail_bytes(console->buffer_size,
+						   wr_idx, rd_idx);
+	if (bytes_to_read <= 0)
+		return bytes_to_read;
+
+	bytes_to_read = min_t(s32, bytes_to_read, buf_size);
+
+	/* Check to see if what we want to read is not contiguous, and limit
+	 * ourselves to the contiguous block
+	 */
+	if (rd_idx + bytes_to_read >= console->buffer_size)
+		bytes_to_read = console->buffer_size - rd_idx;
+
+	octeon_pci_read_core_mem(oct, console->output_base_addr + rd_idx,
+				 (u8 *)buffer, bytes_to_read);
+	octeon_write_device_mem32(oct, console->addr +
+				  offsetof(struct octeon_pci_console,
+					   output_read_index),
+				  (rd_idx + bytes_to_read) %
+				  console->buffer_size);
+
+	return bytes_to_read;
+}
+
+#define FBUF_SIZE	(4 * 1024 * 1024)
+#define MAX_BOOTTIME_SIZE    80
+
+int octeon_download_firmware(struct octeon_device *oct, const u8 *data,
+			     size_t size)
+{
+	struct octeon_firmware_file_header *h;
+	char boottime[MAX_BOOTTIME_SIZE];
+	struct timespec64 ts;
+	u32 crc32_result;
+	u64 load_addr;
+	u32 image_len;
+	int ret = 0;
+	u32 i, rem;
+
+	if (size < sizeof(struct octeon_firmware_file_header)) {
+		dev_err(&oct->pci_dev->dev, "Firmware file too small (%d < %d).\n",
+			(u32)size,
+			(u32)sizeof(struct octeon_firmware_file_header));
+		return -EINVAL;
+	}
+
+	h = (struct octeon_firmware_file_header *)data;
+
+	if (be32_to_cpu(h->magic) != LIO_NIC_MAGIC) {
+		dev_err(&oct->pci_dev->dev, "Unrecognized firmware file.\n");
+		return -EINVAL;
+	}
+
+	crc32_result = crc32((unsigned int)~0, data,
+			     sizeof(struct octeon_firmware_file_header) -
+			     sizeof(u32)) ^ ~0U;
+	if (crc32_result != be32_to_cpu(h->crc32)) {
+		dev_err(&oct->pci_dev->dev, "Firmware CRC mismatch (0x%08x != 0x%08x).\n",
+			crc32_result, be32_to_cpu(h->crc32));
+		return -EINVAL;
+	}
+
+	if (memcmp(LIQUIDIO_BASE_VERSION, h->version,
+		   strlen(LIQUIDIO_BASE_VERSION))) {
+		dev_err(&oct->pci_dev->dev, "Unmatched firmware version. Expected %s.x, got %s.\n",
+			LIQUIDIO_BASE_VERSION,
+			h->version);
+		return -EINVAL;
+	}
+
+	if (be32_to_cpu(h->num_images) > LIO_MAX_IMAGES) {
+		dev_err(&oct->pci_dev->dev, "Too many images in firmware file (%d).\n",
+			be32_to_cpu(h->num_images));
+		return -EINVAL;
+	}
+
+	dev_info(&oct->pci_dev->dev, "Firmware version: %s\n", h->version);
+	snprintf(oct->fw_info.liquidio_firmware_version, 32, "LIQUIDIO: %s",
+		 h->version);
+
+	data += sizeof(struct octeon_firmware_file_header);
+
+	dev_info(&oct->pci_dev->dev, "%s: Loading %d images\n", __func__,
+		 be32_to_cpu(h->num_images));
+	/* load all images */
+	for (i = 0; i < be32_to_cpu(h->num_images); i++) {
+		load_addr = be64_to_cpu(h->desc[i].addr);
+		image_len = be32_to_cpu(h->desc[i].len);
+
+		dev_info(&oct->pci_dev->dev, "Loading firmware %d at %llx\n",
+			 image_len, load_addr);
+
+		/* Write in 4MB chunks*/
+		rem = image_len;
+
+		while (rem) {
+			if (rem < FBUF_SIZE)
+				size = rem;
+			else
+				size = FBUF_SIZE;
+
+			/* download the image */
+			octeon_pci_write_core_mem(oct, load_addr, data, (u32)size);
+
+			data += size;
+			rem -= (u32)size;
+			load_addr += size;
+		}
+	}
+
+	/* Pass date and time information to NIC at the time of loading
+	 * firmware and periodically update the host time to NIC firmware.
+	 * This is to make NIC firmware use the same time reference as Host,
+	 * so that it is easy to correlate logs from firmware and host for
+	 * debugging.
+	 *
+	 * Octeon always uses UTC time. so timezone information is not sent.
+	 */
+	ktime_get_real_ts64(&ts);
+	ret = snprintf(boottime, MAX_BOOTTIME_SIZE,
+		       " time_sec=%lld time_nsec=%ld",
+		       (s64)ts.tv_sec, ts.tv_nsec);
+	if ((sizeof(h->bootcmd) - strnlen(h->bootcmd, sizeof(h->bootcmd))) <
+		ret) {
+		dev_err(&oct->pci_dev->dev, "Boot command buffer too small\n");
+		return -EINVAL;
+	}
+	strncat(h->bootcmd, boottime,
+		sizeof(h->bootcmd) - strnlen(h->bootcmd, sizeof(h->bootcmd)));
+
+	dev_info(&oct->pci_dev->dev, "Writing boot command: %s\n",
+		 h->bootcmd);
+
+	/* Invoke the bootcmd */
+	ret = octeon_console_send_cmd(oct, h->bootcmd, 50);
+	if (ret)
+		dev_info(&oct->pci_dev->dev, "Boot command send failed\n");
+
+	return ret;
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_device.c b/drivers/net/ethernet/cavium/liquidio/octeon_device.c
new file mode 100644
index 000000000..e159194d0
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_device.c
@@ -0,0 +1,1464 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include <linux/vmalloc.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_main.h"
+#include "octeon_network.h"
+#include "cn66xx_regs.h"
+#include "cn66xx_device.h"
+#include "cn23xx_pf_device.h"
+#include "cn23xx_vf_device.h"
+
+/** Default configuration
+ *  for CN66XX OCTEON Models.
+ */
+static struct octeon_config default_cn66xx_conf = {
+	.card_type                              = LIO_210SV,
+	.card_name                              = LIO_210SV_NAME,
+
+	/** IQ attributes */
+	.iq					= {
+		.max_iqs			= CN6XXX_CFG_IO_QUEUES,
+		.pending_list_size		=
+			(CN6XXX_MAX_IQ_DESCRIPTORS * CN6XXX_CFG_IO_QUEUES),
+		.instr_type			= OCTEON_64BYTE_INSTR,
+		.db_min				= CN6XXX_DB_MIN,
+		.db_timeout			= CN6XXX_DB_TIMEOUT,
+	}
+	,
+
+	/** OQ attributes */
+	.oq					= {
+		.max_oqs			= CN6XXX_CFG_IO_QUEUES,
+		.refill_threshold		= CN6XXX_OQ_REFIL_THRESHOLD,
+		.oq_intr_pkt			= CN6XXX_OQ_INTR_PKT,
+		.oq_intr_time			= CN6XXX_OQ_INTR_TIME,
+		.pkts_per_intr			= CN6XXX_OQ_PKTSPER_INTR,
+	}
+	,
+
+	.num_nic_ports				= DEFAULT_NUM_NIC_PORTS_66XX,
+	.num_def_rx_descs			= CN6XXX_MAX_OQ_DESCRIPTORS,
+	.num_def_tx_descs			= CN6XXX_MAX_IQ_DESCRIPTORS,
+	.def_rx_buf_size			= CN6XXX_OQ_BUF_SIZE,
+
+	/* For ethernet interface 0:  Port cfg Attributes */
+	.nic_if_cfg[0] = {
+		/* Max Txqs: Half for each of the two ports :max_iq/2 */
+		.max_txqs			= MAX_TXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_txqs */
+		.num_txqs			= DEF_TXQS_PER_INTF,
+
+		/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+		.max_rxqs			= MAX_RXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_rxqs */
+		.num_rxqs			= DEF_RXQS_PER_INTF,
+
+		/* Num of desc for rx rings */
+		.num_rx_descs			= CN6XXX_MAX_OQ_DESCRIPTORS,
+
+		/* Num of desc for tx rings */
+		.num_tx_descs			= CN6XXX_MAX_IQ_DESCRIPTORS,
+
+		/* SKB size, We need not change buf size even for Jumbo frames.
+		 * Octeon can send jumbo frames in 4 consecutive descriptors,
+		 */
+		.rx_buf_size			= CN6XXX_OQ_BUF_SIZE,
+
+		.base_queue			= BASE_QUEUE_NOT_REQUESTED,
+
+		.gmx_port_id			= 0,
+	},
+
+	.nic_if_cfg[1] = {
+		/* Max Txqs: Half for each of the two ports :max_iq/2 */
+		.max_txqs			= MAX_TXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_txqs */
+		.num_txqs			= DEF_TXQS_PER_INTF,
+
+		/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+		.max_rxqs			= MAX_RXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_rxqs */
+		.num_rxqs			= DEF_RXQS_PER_INTF,
+
+		/* Num of desc for rx rings */
+		.num_rx_descs			= CN6XXX_MAX_OQ_DESCRIPTORS,
+
+		/* Num of desc for tx rings */
+		.num_tx_descs			= CN6XXX_MAX_IQ_DESCRIPTORS,
+
+		/* SKB size, We need not change buf size even for Jumbo frames.
+		 * Octeon can send jumbo frames in 4 consecutive descriptors,
+		 */
+		.rx_buf_size			= CN6XXX_OQ_BUF_SIZE,
+
+		.base_queue			= BASE_QUEUE_NOT_REQUESTED,
+
+		.gmx_port_id			= 1,
+	},
+
+	/** Miscellaneous attributes */
+	.misc					= {
+		/* Host driver link query interval */
+		.oct_link_query_interval	= 100,
+
+		/* Octeon link query interval */
+		.host_link_query_interval	= 500,
+
+		.enable_sli_oq_bp		= 0,
+
+		/* Control queue group */
+		.ctrlq_grp			= 1,
+	}
+	,
+};
+
+/** Default configuration
+ *  for CN68XX OCTEON Model.
+ */
+
+static struct octeon_config default_cn68xx_conf = {
+	.card_type                              = LIO_410NV,
+	.card_name                              = LIO_410NV_NAME,
+
+	/** IQ attributes */
+	.iq					= {
+		.max_iqs			= CN6XXX_CFG_IO_QUEUES,
+		.pending_list_size		=
+			(CN6XXX_MAX_IQ_DESCRIPTORS * CN6XXX_CFG_IO_QUEUES),
+		.instr_type			= OCTEON_64BYTE_INSTR,
+		.db_min				= CN6XXX_DB_MIN,
+		.db_timeout			= CN6XXX_DB_TIMEOUT,
+	}
+	,
+
+	/** OQ attributes */
+	.oq					= {
+		.max_oqs			= CN6XXX_CFG_IO_QUEUES,
+		.refill_threshold		= CN6XXX_OQ_REFIL_THRESHOLD,
+		.oq_intr_pkt			= CN6XXX_OQ_INTR_PKT,
+		.oq_intr_time			= CN6XXX_OQ_INTR_TIME,
+		.pkts_per_intr			= CN6XXX_OQ_PKTSPER_INTR,
+	}
+	,
+
+	.num_nic_ports				= DEFAULT_NUM_NIC_PORTS_68XX,
+	.num_def_rx_descs			= CN6XXX_MAX_OQ_DESCRIPTORS,
+	.num_def_tx_descs			= CN6XXX_MAX_IQ_DESCRIPTORS,
+	.def_rx_buf_size			= CN6XXX_OQ_BUF_SIZE,
+
+	.nic_if_cfg[0] = {
+		/* Max Txqs: Half for each of the two ports :max_iq/2 */
+		.max_txqs			= MAX_TXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_txqs */
+		.num_txqs			= DEF_TXQS_PER_INTF,
+
+		/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+		.max_rxqs			= MAX_RXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_rxqs */
+		.num_rxqs			= DEF_RXQS_PER_INTF,
+
+		/* Num of desc for rx rings */
+		.num_rx_descs			= CN6XXX_MAX_OQ_DESCRIPTORS,
+
+		/* Num of desc for tx rings */
+		.num_tx_descs			= CN6XXX_MAX_IQ_DESCRIPTORS,
+
+		/* SKB size, We need not change buf size even for Jumbo frames.
+		 * Octeon can send jumbo frames in 4 consecutive descriptors,
+		 */
+		.rx_buf_size			= CN6XXX_OQ_BUF_SIZE,
+
+		.base_queue			= BASE_QUEUE_NOT_REQUESTED,
+
+		.gmx_port_id			= 0,
+	},
+
+	.nic_if_cfg[1] = {
+		/* Max Txqs: Half for each of the two ports :max_iq/2 */
+		.max_txqs			= MAX_TXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_txqs */
+		.num_txqs			= DEF_TXQS_PER_INTF,
+
+		/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+		.max_rxqs			= MAX_RXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_rxqs */
+		.num_rxqs			= DEF_RXQS_PER_INTF,
+
+		/* Num of desc for rx rings */
+		.num_rx_descs			= CN6XXX_MAX_OQ_DESCRIPTORS,
+
+		/* Num of desc for tx rings */
+		.num_tx_descs			= CN6XXX_MAX_IQ_DESCRIPTORS,
+
+		/* SKB size, We need not change buf size even for Jumbo frames.
+		 * Octeon can send jumbo frames in 4 consecutive descriptors,
+		 */
+		.rx_buf_size			= CN6XXX_OQ_BUF_SIZE,
+
+		.base_queue			= BASE_QUEUE_NOT_REQUESTED,
+
+		.gmx_port_id			= 1,
+	},
+
+	.nic_if_cfg[2] = {
+		/* Max Txqs: Half for each of the two ports :max_iq/2 */
+		.max_txqs			= MAX_TXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_txqs */
+		.num_txqs			= DEF_TXQS_PER_INTF,
+
+		/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+		.max_rxqs			= MAX_RXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_rxqs */
+		.num_rxqs			= DEF_RXQS_PER_INTF,
+
+		/* Num of desc for rx rings */
+		.num_rx_descs			= CN6XXX_MAX_OQ_DESCRIPTORS,
+
+		/* Num of desc for tx rings */
+		.num_tx_descs			= CN6XXX_MAX_IQ_DESCRIPTORS,
+
+		/* SKB size, We need not change buf size even for Jumbo frames.
+		 * Octeon can send jumbo frames in 4 consecutive descriptors,
+		 */
+		.rx_buf_size			= CN6XXX_OQ_BUF_SIZE,
+
+		.base_queue			= BASE_QUEUE_NOT_REQUESTED,
+
+		.gmx_port_id			= 2,
+	},
+
+	.nic_if_cfg[3] = {
+		/* Max Txqs: Half for each of the two ports :max_iq/2 */
+		.max_txqs			= MAX_TXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_txqs */
+		.num_txqs			= DEF_TXQS_PER_INTF,
+
+		/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+		.max_rxqs			= MAX_RXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_rxqs */
+		.num_rxqs			= DEF_RXQS_PER_INTF,
+
+		/* Num of desc for rx rings */
+		.num_rx_descs			= CN6XXX_MAX_OQ_DESCRIPTORS,
+
+		/* Num of desc for tx rings */
+		.num_tx_descs			= CN6XXX_MAX_IQ_DESCRIPTORS,
+
+		/* SKB size, We need not change buf size even for Jumbo frames.
+		 * Octeon can send jumbo frames in 4 consecutive descriptors,
+		 */
+		.rx_buf_size			= CN6XXX_OQ_BUF_SIZE,
+
+		.base_queue			= BASE_QUEUE_NOT_REQUESTED,
+
+		.gmx_port_id			= 3,
+	},
+
+	/** Miscellaneous attributes */
+	.misc					= {
+		/* Host driver link query interval */
+		.oct_link_query_interval	= 100,
+
+		/* Octeon link query interval */
+		.host_link_query_interval	= 500,
+
+		.enable_sli_oq_bp		= 0,
+
+		/* Control queue group */
+		.ctrlq_grp			= 1,
+	}
+	,
+};
+
+/** Default configuration
+ *  for CN68XX OCTEON Model.
+ */
+static struct octeon_config default_cn68xx_210nv_conf = {
+	.card_type                              = LIO_210NV,
+	.card_name                              = LIO_210NV_NAME,
+
+	/** IQ attributes */
+
+	.iq					= {
+		.max_iqs			= CN6XXX_CFG_IO_QUEUES,
+		.pending_list_size		=
+			(CN6XXX_MAX_IQ_DESCRIPTORS * CN6XXX_CFG_IO_QUEUES),
+		.instr_type			= OCTEON_64BYTE_INSTR,
+		.db_min				= CN6XXX_DB_MIN,
+		.db_timeout			= CN6XXX_DB_TIMEOUT,
+	}
+	,
+
+	/** OQ attributes */
+	.oq					= {
+		.max_oqs			= CN6XXX_CFG_IO_QUEUES,
+		.refill_threshold		= CN6XXX_OQ_REFIL_THRESHOLD,
+		.oq_intr_pkt			= CN6XXX_OQ_INTR_PKT,
+		.oq_intr_time			= CN6XXX_OQ_INTR_TIME,
+		.pkts_per_intr			= CN6XXX_OQ_PKTSPER_INTR,
+	}
+	,
+
+	.num_nic_ports			= DEFAULT_NUM_NIC_PORTS_68XX_210NV,
+	.num_def_rx_descs		= CN6XXX_MAX_OQ_DESCRIPTORS,
+	.num_def_tx_descs		= CN6XXX_MAX_IQ_DESCRIPTORS,
+	.def_rx_buf_size		= CN6XXX_OQ_BUF_SIZE,
+
+	.nic_if_cfg[0] = {
+		/* Max Txqs: Half for each of the two ports :max_iq/2 */
+		.max_txqs			= MAX_TXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_txqs */
+		.num_txqs			= DEF_TXQS_PER_INTF,
+
+		/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+		.max_rxqs			= MAX_RXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_rxqs */
+		.num_rxqs			= DEF_RXQS_PER_INTF,
+
+		/* Num of desc for rx rings */
+		.num_rx_descs			= CN6XXX_MAX_OQ_DESCRIPTORS,
+
+		/* Num of desc for tx rings */
+		.num_tx_descs			= CN6XXX_MAX_IQ_DESCRIPTORS,
+
+		/* SKB size, We need not change buf size even for Jumbo frames.
+		 * Octeon can send jumbo frames in 4 consecutive descriptors,
+		 */
+		.rx_buf_size			= CN6XXX_OQ_BUF_SIZE,
+
+		.base_queue			= BASE_QUEUE_NOT_REQUESTED,
+
+		.gmx_port_id			= 0,
+	},
+
+	.nic_if_cfg[1] = {
+		/* Max Txqs: Half for each of the two ports :max_iq/2 */
+		.max_txqs			= MAX_TXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_txqs */
+		.num_txqs			= DEF_TXQS_PER_INTF,
+
+		/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+		.max_rxqs			= MAX_RXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_rxqs */
+		.num_rxqs			= DEF_RXQS_PER_INTF,
+
+		/* Num of desc for rx rings */
+		.num_rx_descs			= CN6XXX_MAX_OQ_DESCRIPTORS,
+
+		/* Num of desc for tx rings */
+		.num_tx_descs			= CN6XXX_MAX_IQ_DESCRIPTORS,
+
+		/* SKB size, We need not change buf size even for Jumbo frames.
+		 * Octeon can send jumbo frames in 4 consecutive descriptors,
+		 */
+		.rx_buf_size			= CN6XXX_OQ_BUF_SIZE,
+
+		.base_queue			= BASE_QUEUE_NOT_REQUESTED,
+
+		.gmx_port_id			= 1,
+	},
+
+	/** Miscellaneous attributes */
+	.misc					= {
+		/* Host driver link query interval */
+		.oct_link_query_interval	= 100,
+
+		/* Octeon link query interval */
+		.host_link_query_interval	= 500,
+
+		.enable_sli_oq_bp		= 0,
+
+		/* Control queue group */
+		.ctrlq_grp			= 1,
+	}
+	,
+};
+
+static struct octeon_config default_cn23xx_conf = {
+	.card_type                              = LIO_23XX,
+	.card_name                              = LIO_23XX_NAME,
+	/** IQ attributes */
+	.iq = {
+		.max_iqs		= CN23XX_CFG_IO_QUEUES,
+		.pending_list_size	= (CN23XX_DEFAULT_IQ_DESCRIPTORS *
+					   CN23XX_CFG_IO_QUEUES),
+		.instr_type		= OCTEON_64BYTE_INSTR,
+		.db_min			= CN23XX_DB_MIN,
+		.db_timeout		= CN23XX_DB_TIMEOUT,
+		.iq_intr_pkt		= CN23XX_DEF_IQ_INTR_THRESHOLD,
+	},
+
+	/** OQ attributes */
+	.oq = {
+		.max_oqs		= CN23XX_CFG_IO_QUEUES,
+		.pkts_per_intr	= CN23XX_OQ_PKTSPER_INTR,
+		.refill_threshold	= CN23XX_OQ_REFIL_THRESHOLD,
+		.oq_intr_pkt	= CN23XX_OQ_INTR_PKT,
+		.oq_intr_time	= CN23XX_OQ_INTR_TIME,
+	},
+
+	.num_nic_ports				= DEFAULT_NUM_NIC_PORTS_23XX,
+	.num_def_rx_descs			= CN23XX_DEFAULT_OQ_DESCRIPTORS,
+	.num_def_tx_descs			= CN23XX_DEFAULT_IQ_DESCRIPTORS,
+	.def_rx_buf_size			= CN23XX_OQ_BUF_SIZE,
+
+	/* For ethernet interface 0:  Port cfg Attributes */
+	.nic_if_cfg[0] = {
+		/* Max Txqs: Half for each of the two ports :max_iq/2 */
+		.max_txqs			= MAX_TXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_txqs */
+		.num_txqs			= DEF_TXQS_PER_INTF,
+
+		/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+		.max_rxqs			= MAX_RXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_rxqs */
+		.num_rxqs			= DEF_RXQS_PER_INTF,
+
+		/* Num of desc for rx rings */
+		.num_rx_descs			= CN23XX_DEFAULT_OQ_DESCRIPTORS,
+
+		/* Num of desc for tx rings */
+		.num_tx_descs			= CN23XX_DEFAULT_IQ_DESCRIPTORS,
+
+		/* SKB size, We need not change buf size even for Jumbo frames.
+		 * Octeon can send jumbo frames in 4 consecutive descriptors,
+		 */
+		.rx_buf_size			= CN23XX_OQ_BUF_SIZE,
+
+		.base_queue			= BASE_QUEUE_NOT_REQUESTED,
+
+		.gmx_port_id			= 0,
+	},
+
+	.nic_if_cfg[1] = {
+		/* Max Txqs: Half for each of the two ports :max_iq/2 */
+		.max_txqs			= MAX_TXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_txqs */
+		.num_txqs			= DEF_TXQS_PER_INTF,
+
+		/* Max Rxqs: Half for each of the two ports :max_oq/2  */
+		.max_rxqs			= MAX_RXQS_PER_INTF,
+
+		/* Actual configured value. Range could be: 1...max_rxqs */
+		.num_rxqs			= DEF_RXQS_PER_INTF,
+
+		/* Num of desc for rx rings */
+		.num_rx_descs			= CN23XX_DEFAULT_OQ_DESCRIPTORS,
+
+		/* Num of desc for tx rings */
+		.num_tx_descs			= CN23XX_DEFAULT_IQ_DESCRIPTORS,
+
+		/* SKB size, We need not change buf size even for Jumbo frames.
+		 * Octeon can send jumbo frames in 4 consecutive descriptors,
+		 */
+		.rx_buf_size			= CN23XX_OQ_BUF_SIZE,
+
+		.base_queue			= BASE_QUEUE_NOT_REQUESTED,
+
+		.gmx_port_id			= 1,
+	},
+
+	.misc					= {
+		/* Host driver link query interval */
+		.oct_link_query_interval	= 100,
+
+		/* Octeon link query interval */
+		.host_link_query_interval	= 500,
+
+		.enable_sli_oq_bp		= 0,
+
+		/* Control queue group */
+		.ctrlq_grp			= 1,
+	}
+};
+
+static struct octeon_config_ptr {
+	u32 conf_type;
+} oct_conf_info[MAX_OCTEON_DEVICES] = {
+	{
+		OCTEON_CONFIG_TYPE_DEFAULT,
+	}, {
+		OCTEON_CONFIG_TYPE_DEFAULT,
+	}, {
+		OCTEON_CONFIG_TYPE_DEFAULT,
+	}, {
+		OCTEON_CONFIG_TYPE_DEFAULT,
+	},
+};
+
+static char oct_dev_state_str[OCT_DEV_STATES + 1][32] = {
+	"BEGIN", "PCI-ENABLE-DONE", "PCI-MAP-DONE", "DISPATCH-INIT-DONE",
+	"IQ-INIT-DONE", "SCBUFF-POOL-INIT-DONE", "RESPLIST-INIT-DONE",
+	"DROQ-INIT-DONE", "MBOX-SETUP-DONE", "MSIX-ALLOC-VECTOR-DONE",
+	"INTR-SET-DONE", "IO-QUEUES-INIT-DONE", "CONSOLE-INIT-DONE",
+	"HOST-READY", "CORE-READY", "RUNNING", "IN-RESET",
+	"INVALID"
+};
+
+static char oct_dev_app_str[CVM_DRV_APP_COUNT + 1][32] = {
+	"BASE", "NIC", "UNKNOWN"};
+
+static struct octeon_device *octeon_device[MAX_OCTEON_DEVICES];
+static atomic_t adapter_refcounts[MAX_OCTEON_DEVICES];
+static atomic_t adapter_fw_states[MAX_OCTEON_DEVICES];
+
+static u32 octeon_device_count;
+/* locks device array (i.e. octeon_device[]) */
+static DEFINE_SPINLOCK(octeon_devices_lock);
+
+static struct octeon_core_setup core_setup[MAX_OCTEON_DEVICES];
+
+static void oct_set_config_info(int oct_id, int conf_type)
+{
+	if (conf_type < 0 || conf_type > (NUM_OCTEON_CONFS - 1))
+		conf_type = OCTEON_CONFIG_TYPE_DEFAULT;
+	oct_conf_info[oct_id].conf_type = conf_type;
+}
+
+void octeon_init_device_list(int conf_type)
+{
+	int i;
+
+	memset(octeon_device, 0, (sizeof(void *) * MAX_OCTEON_DEVICES));
+	for (i = 0; i <  MAX_OCTEON_DEVICES; i++)
+		oct_set_config_info(i, conf_type);
+}
+
+static void *__retrieve_octeon_config_info(struct octeon_device *oct,
+					   u16 card_type)
+{
+	u32 oct_id = oct->octeon_id;
+	void *ret = NULL;
+
+	switch (oct_conf_info[oct_id].conf_type) {
+	case OCTEON_CONFIG_TYPE_DEFAULT:
+		if (oct->chip_id == OCTEON_CN66XX) {
+			ret = &default_cn66xx_conf;
+		} else if ((oct->chip_id == OCTEON_CN68XX) &&
+			   (card_type == LIO_210NV)) {
+			ret = &default_cn68xx_210nv_conf;
+		} else if ((oct->chip_id == OCTEON_CN68XX) &&
+			   (card_type == LIO_410NV)) {
+			ret = &default_cn68xx_conf;
+		} else if (oct->chip_id == OCTEON_CN23XX_PF_VID) {
+			ret = &default_cn23xx_conf;
+		} else if (oct->chip_id == OCTEON_CN23XX_VF_VID) {
+			ret = &default_cn23xx_conf;
+		}
+		break;
+	default:
+		break;
+	}
+	return ret;
+}
+
+static int __verify_octeon_config_info(struct octeon_device *oct, void *conf)
+{
+	switch (oct->chip_id) {
+	case OCTEON_CN66XX:
+	case OCTEON_CN68XX:
+		return lio_validate_cn6xxx_config_info(oct, conf);
+	case OCTEON_CN23XX_PF_VID:
+	case OCTEON_CN23XX_VF_VID:
+		return 0;
+	default:
+		break;
+	}
+
+	return 1;
+}
+
+void *oct_get_config_info(struct octeon_device *oct, u16 card_type)
+{
+	void *conf = NULL;
+
+	conf = __retrieve_octeon_config_info(oct, card_type);
+	if (!conf)
+		return NULL;
+
+	if (__verify_octeon_config_info(oct, conf)) {
+		dev_err(&oct->pci_dev->dev, "Configuration verification failed\n");
+		return NULL;
+	}
+
+	return conf;
+}
+
+char *lio_get_state_string(atomic_t *state_ptr)
+{
+	s32 istate = (s32)atomic_read(state_ptr);
+
+	if (istate > OCT_DEV_STATES || istate < 0)
+		return oct_dev_state_str[OCT_DEV_STATE_INVALID];
+	return oct_dev_state_str[istate];
+}
+
+static char *get_oct_app_string(u32 app_mode)
+{
+	if (app_mode <= CVM_DRV_APP_END)
+		return oct_dev_app_str[app_mode - CVM_DRV_APP_START];
+	return oct_dev_app_str[CVM_DRV_INVALID_APP - CVM_DRV_APP_START];
+}
+
+void octeon_free_device_mem(struct octeon_device *oct)
+{
+	int i;
+
+	for (i = 0; i < MAX_OCTEON_OUTPUT_QUEUES(oct); i++) {
+		if (oct->io_qmask.oq & BIT_ULL(i))
+			vfree(oct->droq[i]);
+	}
+
+	for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct); i++) {
+		if (oct->io_qmask.iq & BIT_ULL(i))
+			vfree(oct->instr_queue[i]);
+	}
+
+	i = oct->octeon_id;
+	vfree(oct);
+
+	octeon_device[i] = NULL;
+	octeon_device_count--;
+}
+
+static struct octeon_device *octeon_allocate_device_mem(u32 pci_id,
+							u32 priv_size)
+{
+	struct octeon_device *oct;
+	u8 *buf = NULL;
+	u32 octdevsize = 0, configsize = 0, size;
+
+	switch (pci_id) {
+	case OCTEON_CN68XX:
+	case OCTEON_CN66XX:
+		configsize = sizeof(struct octeon_cn6xxx);
+		break;
+
+	case OCTEON_CN23XX_PF_VID:
+		configsize = sizeof(struct octeon_cn23xx_pf);
+		break;
+	case OCTEON_CN23XX_VF_VID:
+		configsize = sizeof(struct octeon_cn23xx_vf);
+		break;
+	default:
+		pr_err("%s: Unknown PCI Device: 0x%x\n",
+		       __func__,
+		       pci_id);
+		return NULL;
+	}
+
+	if (configsize & 0x7)
+		configsize += (8 - (configsize & 0x7));
+
+	octdevsize = sizeof(struct octeon_device);
+	if (octdevsize & 0x7)
+		octdevsize += (8 - (octdevsize & 0x7));
+
+	if (priv_size & 0x7)
+		priv_size += (8 - (priv_size & 0x7));
+
+	size = octdevsize + priv_size + configsize +
+		(sizeof(struct octeon_dispatch) * DISPATCH_LIST_SIZE);
+
+	buf = vzalloc(size);
+	if (!buf)
+		return NULL;
+
+	oct = (struct octeon_device *)buf;
+	oct->priv = (void *)(buf + octdevsize);
+	oct->chip = (void *)(buf + octdevsize + priv_size);
+	oct->dispatch.dlist = (struct octeon_dispatch *)
+		(buf + octdevsize + priv_size + configsize);
+
+	return oct;
+}
+
+struct octeon_device *octeon_allocate_device(u32 pci_id,
+					     u32 priv_size)
+{
+	u32 oct_idx = 0;
+	struct octeon_device *oct = NULL;
+
+	spin_lock(&octeon_devices_lock);
+
+	for (oct_idx = 0; oct_idx < MAX_OCTEON_DEVICES; oct_idx++)
+		if (!octeon_device[oct_idx])
+			break;
+
+	if (oct_idx < MAX_OCTEON_DEVICES) {
+		oct = octeon_allocate_device_mem(pci_id, priv_size);
+		if (oct) {
+			octeon_device_count++;
+			octeon_device[oct_idx] = oct;
+		}
+	}
+
+	spin_unlock(&octeon_devices_lock);
+	if (!oct)
+		return NULL;
+
+	spin_lock_init(&oct->pci_win_lock);
+	spin_lock_init(&oct->mem_access_lock);
+
+	oct->octeon_id = oct_idx;
+	snprintf(oct->device_name, sizeof(oct->device_name),
+		 "LiquidIO%d", (oct->octeon_id));
+
+	return oct;
+}
+
+/** Register a device's bus location at initialization time.
+ *  @param octeon_dev - pointer to the octeon device structure.
+ *  @param bus        - PCIe bus #
+ *  @param dev        - PCIe device #
+ *  @param func       - PCIe function #
+ *  @param is_pf      - TRUE for PF, FALSE for VF
+ *  @return reference count of device's adapter
+ */
+int octeon_register_device(struct octeon_device *oct,
+			   int bus, int dev, int func, int is_pf)
+{
+	int idx, refcount;
+
+	oct->loc.bus = bus;
+	oct->loc.dev = dev;
+	oct->loc.func = func;
+
+	oct->adapter_refcount = &adapter_refcounts[oct->octeon_id];
+	atomic_set(oct->adapter_refcount, 0);
+
+	/* Like the reference count, the f/w state is shared 'per-adapter' */
+	oct->adapter_fw_state = &adapter_fw_states[oct->octeon_id];
+	atomic_set(oct->adapter_fw_state, FW_NEEDS_TO_BE_LOADED);
+
+	spin_lock(&octeon_devices_lock);
+	for (idx = (int)oct->octeon_id - 1; idx >= 0; idx--) {
+		if (!octeon_device[idx]) {
+			dev_err(&oct->pci_dev->dev,
+				"%s: Internal driver error, missing dev",
+				__func__);
+			spin_unlock(&octeon_devices_lock);
+			atomic_inc(oct->adapter_refcount);
+			return 1; /* here, refcount is guaranteed to be 1 */
+		}
+		/* If another device is at same bus/dev, use its refcounter
+		 * (and f/w state variable).
+		 */
+		if ((octeon_device[idx]->loc.bus == bus) &&
+		    (octeon_device[idx]->loc.dev == dev)) {
+			oct->adapter_refcount =
+				octeon_device[idx]->adapter_refcount;
+			oct->adapter_fw_state =
+				octeon_device[idx]->adapter_fw_state;
+			break;
+		}
+	}
+	spin_unlock(&octeon_devices_lock);
+
+	atomic_inc(oct->adapter_refcount);
+	refcount = atomic_read(oct->adapter_refcount);
+
+	dev_dbg(&oct->pci_dev->dev, "%s: %02x:%02x:%d refcount %u", __func__,
+		oct->loc.bus, oct->loc.dev, oct->loc.func, refcount);
+
+	return refcount;
+}
+
+/** Deregister a device at de-initialization time.
+ *  @param octeon_dev - pointer to the octeon device structure.
+ *  @return reference count of device's adapter
+ */
+int octeon_deregister_device(struct octeon_device *oct)
+{
+	int refcount;
+
+	atomic_dec(oct->adapter_refcount);
+	refcount = atomic_read(oct->adapter_refcount);
+
+	dev_dbg(&oct->pci_dev->dev, "%s: %04d:%02d:%d refcount %u", __func__,
+		oct->loc.bus, oct->loc.dev, oct->loc.func, refcount);
+
+	return refcount;
+}
+
+int
+octeon_allocate_ioq_vector(struct octeon_device *oct, u32 num_ioqs)
+{
+	struct octeon_ioq_vector *ioq_vector;
+	int cpu_num;
+	int size;
+	int i;
+
+	size = sizeof(struct octeon_ioq_vector) * num_ioqs;
+
+	oct->ioq_vector = vzalloc(size);
+	if (!oct->ioq_vector)
+		return -1;
+	for (i = 0; i < num_ioqs; i++) {
+		ioq_vector		= &oct->ioq_vector[i];
+		ioq_vector->oct_dev	= oct;
+		ioq_vector->iq_index	= i;
+		ioq_vector->droq_index	= i;
+		ioq_vector->mbox	= oct->mbox[i];
+
+		cpu_num = i % num_online_cpus();
+		cpumask_set_cpu(cpu_num, &ioq_vector->affinity_mask);
+
+		if (oct->chip_id == OCTEON_CN23XX_PF_VID)
+			ioq_vector->ioq_num	= i + oct->sriov_info.pf_srn;
+		else
+			ioq_vector->ioq_num	= i;
+	}
+
+	return 0;
+}
+
+void
+octeon_free_ioq_vector(struct octeon_device *oct)
+{
+	vfree(oct->ioq_vector);
+}
+
+/* this function is only for setting up the first queue */
+int octeon_setup_instr_queues(struct octeon_device *oct)
+{
+	u32 num_descs = 0;
+	u32 iq_no = 0;
+	union oct_txpciq txpciq;
+	int numa_node = dev_to_node(&oct->pci_dev->dev);
+
+	if (OCTEON_CN6XXX(oct))
+		num_descs =
+			CFG_GET_NUM_DEF_TX_DESCS(CHIP_CONF(oct, cn6xxx));
+	else if (OCTEON_CN23XX_PF(oct))
+		num_descs = CFG_GET_NUM_DEF_TX_DESCS(CHIP_CONF(oct, cn23xx_pf));
+	else if (OCTEON_CN23XX_VF(oct))
+		num_descs = CFG_GET_NUM_DEF_TX_DESCS(CHIP_CONF(oct, cn23xx_vf));
+
+	oct->num_iqs = 0;
+
+	oct->instr_queue[0] = vzalloc_node(sizeof(*oct->instr_queue[0]),
+				numa_node);
+	if (!oct->instr_queue[0])
+		oct->instr_queue[0] =
+			vzalloc(sizeof(struct octeon_instr_queue));
+	if (!oct->instr_queue[0])
+		return 1;
+	memset(oct->instr_queue[0], 0, sizeof(struct octeon_instr_queue));
+	oct->instr_queue[0]->q_index = 0;
+	oct->instr_queue[0]->app_ctx = (void *)(size_t)0;
+	oct->instr_queue[0]->ifidx = 0;
+	txpciq.u64 = 0;
+	txpciq.s.q_no = iq_no;
+	txpciq.s.pkind = oct->pfvf_hsword.pkind;
+	txpciq.s.use_qpg = 0;
+	txpciq.s.qpg = 0;
+	if (octeon_init_instr_queue(oct, txpciq, num_descs)) {
+		/* prevent memory leak */
+		vfree(oct->instr_queue[0]);
+		oct->instr_queue[0] = NULL;
+		return 1;
+	}
+
+	oct->num_iqs++;
+	return 0;
+}
+
+int octeon_setup_output_queues(struct octeon_device *oct)
+{
+	u32 num_descs = 0;
+	u32 desc_size = 0;
+	u32 oq_no = 0;
+	int numa_node = dev_to_node(&oct->pci_dev->dev);
+
+	if (OCTEON_CN6XXX(oct)) {
+		num_descs =
+			CFG_GET_NUM_DEF_RX_DESCS(CHIP_CONF(oct, cn6xxx));
+		desc_size =
+			CFG_GET_DEF_RX_BUF_SIZE(CHIP_CONF(oct, cn6xxx));
+	} else if (OCTEON_CN23XX_PF(oct)) {
+		num_descs = CFG_GET_NUM_DEF_RX_DESCS(CHIP_CONF(oct, cn23xx_pf));
+		desc_size = CFG_GET_DEF_RX_BUF_SIZE(CHIP_CONF(oct, cn23xx_pf));
+	} else if (OCTEON_CN23XX_VF(oct)) {
+		num_descs = CFG_GET_NUM_DEF_RX_DESCS(CHIP_CONF(oct, cn23xx_vf));
+		desc_size = CFG_GET_DEF_RX_BUF_SIZE(CHIP_CONF(oct, cn23xx_vf));
+	}
+	oct->num_oqs = 0;
+	oct->droq[0] = vzalloc_node(sizeof(*oct->droq[0]), numa_node);
+	if (!oct->droq[0])
+		oct->droq[0] = vzalloc(sizeof(*oct->droq[0]));
+	if (!oct->droq[0])
+		return 1;
+
+	if (octeon_init_droq(oct, oq_no, num_descs, desc_size, NULL)) {
+		vfree(oct->droq[oq_no]);
+		oct->droq[oq_no] = NULL;
+		return 1;
+	}
+	oct->num_oqs++;
+
+	return 0;
+}
+
+int octeon_set_io_queues_off(struct octeon_device *oct)
+{
+	int loop = BUSY_READING_REG_VF_LOOP_COUNT;
+
+	if (OCTEON_CN6XXX(oct)) {
+		octeon_write_csr(oct, CN6XXX_SLI_PKT_INSTR_ENB, 0);
+		octeon_write_csr(oct, CN6XXX_SLI_PKT_OUT_ENB, 0);
+	} else if (oct->chip_id == OCTEON_CN23XX_VF_VID) {
+		u32 q_no;
+
+		/* IOQs will already be in reset.
+		 * If RST bit is set, wait for quiet bit to be set.
+		 * Once quiet bit is set, clear the RST bit.
+		 */
+		for (q_no = 0; q_no < oct->sriov_info.rings_per_vf; q_no++) {
+			u64 reg_val = octeon_read_csr64(
+				oct, CN23XX_VF_SLI_IQ_PKT_CONTROL64(q_no));
+
+			while ((reg_val & CN23XX_PKT_INPUT_CTL_RST) &&
+			       !(reg_val &  CN23XX_PKT_INPUT_CTL_QUIET) &&
+			       loop) {
+				reg_val = octeon_read_csr64(
+					oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no));
+				loop--;
+			}
+			if (!loop) {
+				dev_err(&oct->pci_dev->dev,
+					"clearing the reset reg failed or setting the quiet reg failed for qno: %u\n",
+					q_no);
+				return -1;
+			}
+
+			reg_val = reg_val & ~CN23XX_PKT_INPUT_CTL_RST;
+			octeon_write_csr64(oct,
+					   CN23XX_SLI_IQ_PKT_CONTROL64(q_no),
+					   reg_val);
+
+			reg_val = octeon_read_csr64(
+					oct, CN23XX_SLI_IQ_PKT_CONTROL64(q_no));
+			if (reg_val & CN23XX_PKT_INPUT_CTL_RST) {
+				dev_err(&oct->pci_dev->dev,
+					"unable to reset qno %u\n", q_no);
+				return -1;
+			}
+		}
+	}
+	return 0;
+}
+
+void octeon_set_droq_pkt_op(struct octeon_device *oct,
+			    u32 q_no,
+			    u32 enable)
+{
+	u32 reg_val = 0;
+
+	/* Disable the i/p and o/p queues for this Octeon. */
+	if (OCTEON_CN6XXX(oct)) {
+		reg_val = octeon_read_csr(oct, CN6XXX_SLI_PKT_OUT_ENB);
+
+		if (enable)
+			reg_val = reg_val | (1 << q_no);
+		else
+			reg_val = reg_val & (~(1 << q_no));
+
+		octeon_write_csr(oct, CN6XXX_SLI_PKT_OUT_ENB, reg_val);
+	}
+}
+
+int octeon_init_dispatch_list(struct octeon_device *oct)
+{
+	u32 i;
+
+	oct->dispatch.count = 0;
+
+	for (i = 0; i < DISPATCH_LIST_SIZE; i++) {
+		oct->dispatch.dlist[i].opcode = 0;
+		INIT_LIST_HEAD(&oct->dispatch.dlist[i].list);
+	}
+
+	for (i = 0; i <= REQTYPE_LAST; i++)
+		octeon_register_reqtype_free_fn(oct, i, NULL);
+
+	spin_lock_init(&oct->dispatch.lock);
+
+	return 0;
+}
+
+void octeon_delete_dispatch_list(struct octeon_device *oct)
+{
+	u32 i;
+	struct list_head freelist, *temp, *tmp2;
+
+	INIT_LIST_HEAD(&freelist);
+
+	spin_lock_bh(&oct->dispatch.lock);
+
+	for (i = 0; i < DISPATCH_LIST_SIZE; i++) {
+		struct list_head *dispatch;
+
+		dispatch = &oct->dispatch.dlist[i].list;
+		while (dispatch->next != dispatch) {
+			temp = dispatch->next;
+			list_move_tail(temp, &freelist);
+		}
+
+		oct->dispatch.dlist[i].opcode = 0;
+	}
+
+	oct->dispatch.count = 0;
+
+	spin_unlock_bh(&oct->dispatch.lock);
+
+	list_for_each_safe(temp, tmp2, &freelist) {
+		list_del(temp);
+		kfree(temp);
+	}
+}
+
+octeon_dispatch_fn_t
+octeon_get_dispatch(struct octeon_device *octeon_dev, u16 opcode,
+		    u16 subcode)
+{
+	u32 idx;
+	struct list_head *dispatch;
+	octeon_dispatch_fn_t fn = NULL;
+	u16 combined_opcode = OPCODE_SUBCODE(opcode, subcode);
+
+	idx = combined_opcode & OCTEON_OPCODE_MASK;
+
+	spin_lock_bh(&octeon_dev->dispatch.lock);
+
+	if (octeon_dev->dispatch.count == 0) {
+		spin_unlock_bh(&octeon_dev->dispatch.lock);
+		return NULL;
+	}
+
+	if (!(octeon_dev->dispatch.dlist[idx].opcode)) {
+		spin_unlock_bh(&octeon_dev->dispatch.lock);
+		return NULL;
+	}
+
+	if (octeon_dev->dispatch.dlist[idx].opcode == combined_opcode) {
+		fn = octeon_dev->dispatch.dlist[idx].dispatch_fn;
+	} else {
+		list_for_each(dispatch,
+			      &octeon_dev->dispatch.dlist[idx].list) {
+			if (((struct octeon_dispatch *)dispatch)->opcode ==
+			    combined_opcode) {
+				fn = ((struct octeon_dispatch *)
+				      dispatch)->dispatch_fn;
+				break;
+			}
+		}
+	}
+
+	spin_unlock_bh(&octeon_dev->dispatch.lock);
+	return fn;
+}
+
+/* octeon_register_dispatch_fn
+ * Parameters:
+ *   octeon_id - id of the octeon device.
+ *   opcode    - opcode for which driver should call the registered function
+ *   subcode   - subcode for which driver should call the registered function
+ *   fn        - The function to call when a packet with "opcode" arrives in
+ *		  octeon output queues.
+ *   fn_arg    - The argument to be passed when calling function "fn".
+ * Description:
+ *   Registers a function and its argument to be called when a packet
+ *   arrives in Octeon output queues with "opcode".
+ * Returns:
+ *   Success: 0
+ *   Failure: 1
+ * Locks:
+ *   No locks are held.
+ */
+int
+octeon_register_dispatch_fn(struct octeon_device *oct,
+			    u16 opcode,
+			    u16 subcode,
+			    octeon_dispatch_fn_t fn, void *fn_arg)
+{
+	u32 idx;
+	octeon_dispatch_fn_t pfn;
+	u16 combined_opcode = OPCODE_SUBCODE(opcode, subcode);
+
+	idx = combined_opcode & OCTEON_OPCODE_MASK;
+
+	spin_lock_bh(&oct->dispatch.lock);
+	/* Add dispatch function to first level of lookup table */
+	if (oct->dispatch.dlist[idx].opcode == 0) {
+		oct->dispatch.dlist[idx].opcode = combined_opcode;
+		oct->dispatch.dlist[idx].dispatch_fn = fn;
+		oct->dispatch.dlist[idx].arg = fn_arg;
+		oct->dispatch.count++;
+		spin_unlock_bh(&oct->dispatch.lock);
+		return 0;
+	}
+
+	spin_unlock_bh(&oct->dispatch.lock);
+
+	/* Check if there was a function already registered for this
+	 * opcode/subcode.
+	 */
+	pfn = octeon_get_dispatch(oct, opcode, subcode);
+	if (!pfn) {
+		struct octeon_dispatch *dispatch;
+
+		dev_dbg(&oct->pci_dev->dev,
+			"Adding opcode to dispatch list linked list\n");
+		dispatch = kmalloc(sizeof(*dispatch), GFP_KERNEL);
+		if (!dispatch)
+			return 1;
+
+		dispatch->opcode = combined_opcode;
+		dispatch->dispatch_fn = fn;
+		dispatch->arg = fn_arg;
+
+		/* Add dispatch function to linked list of fn ptrs
+		 * at the hashed index.
+		 */
+		spin_lock_bh(&oct->dispatch.lock);
+		list_add(&dispatch->list, &oct->dispatch.dlist[idx].list);
+		oct->dispatch.count++;
+		spin_unlock_bh(&oct->dispatch.lock);
+
+	} else {
+		if (pfn == fn &&
+		    octeon_get_dispatch_arg(oct, opcode, subcode) == fn_arg)
+			return 0;
+
+		dev_err(&oct->pci_dev->dev,
+			"Found previously registered dispatch fn for opcode/subcode: %x/%x\n",
+			opcode, subcode);
+		return 1;
+	}
+
+	return 0;
+}
+
+int octeon_core_drv_init(struct octeon_recv_info *recv_info, void *buf)
+{
+	u32 i;
+	char app_name[16];
+	struct octeon_device *oct = (struct octeon_device *)buf;
+	struct octeon_recv_pkt *recv_pkt = recv_info->recv_pkt;
+	struct octeon_core_setup *cs = NULL;
+	u32 num_nic_ports = 0;
+
+	if (OCTEON_CN6XXX(oct))
+		num_nic_ports =
+			CFG_GET_NUM_NIC_PORTS(CHIP_CONF(oct, cn6xxx));
+	else if (OCTEON_CN23XX_PF(oct))
+		num_nic_ports =
+			CFG_GET_NUM_NIC_PORTS(CHIP_CONF(oct, cn23xx_pf));
+
+	if (atomic_read(&oct->status) >= OCT_DEV_RUNNING) {
+		dev_err(&oct->pci_dev->dev, "Received CORE OK when device state is 0x%x\n",
+			atomic_read(&oct->status));
+		goto core_drv_init_err;
+	}
+
+	strncpy(app_name,
+		get_oct_app_string(
+		(u32)recv_pkt->rh.r_core_drv_init.app_mode),
+		sizeof(app_name) - 1);
+	oct->app_mode = (u32)recv_pkt->rh.r_core_drv_init.app_mode;
+	if (recv_pkt->rh.r_core_drv_init.app_mode == CVM_DRV_NIC_APP) {
+		oct->fw_info.max_nic_ports =
+			(u32)recv_pkt->rh.r_core_drv_init.max_nic_ports;
+		oct->fw_info.num_gmx_ports =
+			(u32)recv_pkt->rh.r_core_drv_init.num_gmx_ports;
+	}
+
+	if (oct->fw_info.max_nic_ports < num_nic_ports) {
+		dev_err(&oct->pci_dev->dev,
+			"Config has more ports than firmware allows (%d > %d).\n",
+			num_nic_ports, oct->fw_info.max_nic_ports);
+		goto core_drv_init_err;
+	}
+	oct->fw_info.app_cap_flags = recv_pkt->rh.r_core_drv_init.app_cap_flags;
+	oct->fw_info.app_mode = (u32)recv_pkt->rh.r_core_drv_init.app_mode;
+	oct->pfvf_hsword.app_mode = (u32)recv_pkt->rh.r_core_drv_init.app_mode;
+
+	oct->pfvf_hsword.pkind = recv_pkt->rh.r_core_drv_init.pkind;
+
+	for (i = 0; i < oct->num_iqs; i++)
+		oct->instr_queue[i]->txpciq.s.pkind = oct->pfvf_hsword.pkind;
+
+	atomic_set(&oct->status, OCT_DEV_CORE_OK);
+
+	cs = &core_setup[oct->octeon_id];
+
+	if (recv_pkt->buffer_size[0] != (sizeof(*cs) + OCT_DROQ_INFO_SIZE)) {
+		dev_dbg(&oct->pci_dev->dev, "Core setup bytes expected %u found %d\n",
+			(u32)sizeof(*cs),
+			recv_pkt->buffer_size[0]);
+	}
+
+	memcpy(cs, get_rbd(
+	       recv_pkt->buffer_ptr[0]) + OCT_DROQ_INFO_SIZE, sizeof(*cs));
+
+	strncpy(oct->boardinfo.name, cs->boardname, OCT_BOARD_NAME);
+	strncpy(oct->boardinfo.serial_number, cs->board_serial_number,
+		OCT_SERIAL_LEN);
+
+	octeon_swap_8B_data((u64 *)cs, (sizeof(*cs) >> 3));
+
+	oct->boardinfo.major = cs->board_rev_major;
+	oct->boardinfo.minor = cs->board_rev_minor;
+
+	dev_info(&oct->pci_dev->dev,
+		 "Running %s (%llu Hz)\n",
+		 app_name, CVM_CAST64(cs->corefreq));
+
+core_drv_init_err:
+	for (i = 0; i < recv_pkt->buffer_count; i++)
+		recv_buffer_free(recv_pkt->buffer_ptr[i]);
+	octeon_free_recv_info(recv_info);
+	return 0;
+}
+
+int octeon_get_tx_qsize(struct octeon_device *oct, u32 q_no)
+
+{
+	if (oct && (q_no < MAX_OCTEON_INSTR_QUEUES(oct)) &&
+	    (oct->io_qmask.iq & BIT_ULL(q_no)))
+		return oct->instr_queue[q_no]->max_count;
+
+	return -1;
+}
+
+int octeon_get_rx_qsize(struct octeon_device *oct, u32 q_no)
+{
+	if (oct && (q_no < MAX_OCTEON_OUTPUT_QUEUES(oct)) &&
+	    (oct->io_qmask.oq & BIT_ULL(q_no)))
+		return oct->droq[q_no]->max_count;
+	return -1;
+}
+
+/* Retruns the host firmware handshake OCTEON specific configuration */
+struct octeon_config *octeon_get_conf(struct octeon_device *oct)
+{
+	struct octeon_config *default_oct_conf = NULL;
+
+	/* check the OCTEON Device model & return the corresponding octeon
+	 * configuration
+	 */
+
+	if (OCTEON_CN6XXX(oct)) {
+		default_oct_conf =
+			(struct octeon_config *)(CHIP_CONF(oct, cn6xxx));
+	} else if (OCTEON_CN23XX_PF(oct)) {
+		default_oct_conf = (struct octeon_config *)
+			(CHIP_CONF(oct, cn23xx_pf));
+	} else if (OCTEON_CN23XX_VF(oct)) {
+		default_oct_conf = (struct octeon_config *)
+			(CHIP_CONF(oct, cn23xx_vf));
+	}
+	return default_oct_conf;
+}
+
+/* scratch register address is same in all the OCT-II and CN70XX models */
+#define CNXX_SLI_SCRATCH1   0x3C0
+
+/* Get the octeon device pointer.
+ *  @param octeon_id  - The id for which the octeon device pointer is required.
+ *  @return Success: Octeon device pointer.
+ *  @return Failure: NULL.
+ */
+struct octeon_device *lio_get_device(u32 octeon_id)
+{
+	if (octeon_id >= MAX_OCTEON_DEVICES)
+		return NULL;
+	else
+		return octeon_device[octeon_id];
+}
+
+u64 lio_pci_readq(struct octeon_device *oct, u64 addr)
+{
+	u64 val64;
+	unsigned long flags;
+	u32 addrhi;
+
+	spin_lock_irqsave(&oct->pci_win_lock, flags);
+
+	/* The windowed read happens when the LSB of the addr is written.
+	 * So write MSB first
+	 */
+	addrhi = (addr >> 32);
+	if ((oct->chip_id == OCTEON_CN66XX) ||
+	    (oct->chip_id == OCTEON_CN68XX) ||
+	    (oct->chip_id == OCTEON_CN23XX_PF_VID))
+		addrhi |= 0x00060000;
+	writel(addrhi, oct->reg_list.pci_win_rd_addr_hi);
+
+	/* Read back to preserve ordering of writes */
+	readl(oct->reg_list.pci_win_rd_addr_hi);
+
+	writel(addr & 0xffffffff, oct->reg_list.pci_win_rd_addr_lo);
+	readl(oct->reg_list.pci_win_rd_addr_lo);
+
+	val64 = readq(oct->reg_list.pci_win_rd_data);
+
+	spin_unlock_irqrestore(&oct->pci_win_lock, flags);
+
+	return val64;
+}
+
+void lio_pci_writeq(struct octeon_device *oct,
+		    u64 val,
+		    u64 addr)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&oct->pci_win_lock, flags);
+
+	writeq(addr, oct->reg_list.pci_win_wr_addr);
+
+	/* The write happens when the LSB is written. So write MSB first. */
+	writel(val >> 32, oct->reg_list.pci_win_wr_data_hi);
+	/* Read the MSB to ensure ordering of writes. */
+	readl(oct->reg_list.pci_win_wr_data_hi);
+
+	writel(val & 0xffffffff, oct->reg_list.pci_win_wr_data_lo);
+
+	spin_unlock_irqrestore(&oct->pci_win_lock, flags);
+}
+
+int octeon_mem_access_ok(struct octeon_device *oct)
+{
+	u64 access_okay = 0;
+	u64 lmc0_reset_ctl;
+
+	/* Check to make sure a DDR interface is enabled */
+	if (OCTEON_CN23XX_PF(oct)) {
+		lmc0_reset_ctl = lio_pci_readq(oct, CN23XX_LMC0_RESET_CTL);
+		access_okay =
+			(lmc0_reset_ctl & CN23XX_LMC0_RESET_CTL_DDR3RST_MASK);
+	} else {
+		lmc0_reset_ctl = lio_pci_readq(oct, CN6XXX_LMC0_RESET_CTL);
+		access_okay =
+			(lmc0_reset_ctl & CN6XXX_LMC0_RESET_CTL_DDR3RST_MASK);
+	}
+
+	return access_okay ? 0 : 1;
+}
+
+int octeon_wait_for_ddr_init(struct octeon_device *oct, u32 *timeout)
+{
+	int ret = 1;
+	u32 ms;
+
+	if (!timeout)
+		return ret;
+
+	for (ms = 0; (ret != 0) && ((*timeout == 0) || (ms <= *timeout));
+	     ms += HZ / 10) {
+		ret = octeon_mem_access_ok(oct);
+
+		/* wait 100 ms */
+		if (ret)
+			schedule_timeout_uninterruptible(HZ / 10);
+	}
+
+	return ret;
+}
+
+/* Get the octeon id assigned to the octeon device passed as argument.
+ *  This function is exported to other modules.
+ *  @param dev - octeon device pointer passed as a void *.
+ *  @return octeon device id
+ */
+int lio_get_device_id(void *dev)
+{
+	struct octeon_device *octeon_dev = (struct octeon_device *)dev;
+	u32 i;
+
+	for (i = 0; i < MAX_OCTEON_DEVICES; i++)
+		if (octeon_device[i] == octeon_dev)
+			return octeon_dev->octeon_id;
+	return -1;
+}
+
+void lio_enable_irq(struct octeon_droq *droq, struct octeon_instr_queue *iq)
+{
+	u64 instr_cnt;
+	u32 pkts_pend;
+	struct octeon_device *oct = NULL;
+
+	/* the whole thing needs to be atomic, ideally */
+	if (droq) {
+		pkts_pend = (u32)atomic_read(&droq->pkts_pending);
+		writel(droq->pkt_count - pkts_pend, droq->pkts_sent_reg);
+		droq->pkt_count = pkts_pend;
+		oct = droq->oct_dev;
+	}
+	if (iq) {
+		spin_lock_bh(&iq->lock);
+		writel(iq->pkts_processed, iq->inst_cnt_reg);
+		iq->pkt_in_done -= iq->pkts_processed;
+		iq->pkts_processed = 0;
+		/* this write needs to be flushed before we release the lock */
+		spin_unlock_bh(&iq->lock);
+		oct = iq->oct_dev;
+	}
+	/*write resend. Writing RESEND in SLI_PKTX_CNTS should be enough
+	 *to trigger tx interrupts as well, if they are pending.
+	 */
+	if (oct && (OCTEON_CN23XX_PF(oct) || OCTEON_CN23XX_VF(oct))) {
+		if (droq)
+			writeq(CN23XX_INTR_RESEND, droq->pkts_sent_reg);
+		/*we race with firmrware here. read and write the IN_DONE_CNTS*/
+		else if (iq) {
+			instr_cnt =  readq(iq->inst_cnt_reg);
+			writeq(((instr_cnt & 0xFFFFFFFF00000000ULL) |
+				CN23XX_INTR_RESEND),
+			       iq->inst_cnt_reg);
+		}
+	}
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_device.h b/drivers/net/ethernet/cavium/liquidio/octeon_device.h
new file mode 100644
index 000000000..fb380b4f3
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_device.h
@@ -0,0 +1,911 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*! \file octeon_device.h
+ *  \brief Host Driver: This file defines the octeon device structure.
+ */
+
+#ifndef _OCTEON_DEVICE_H_
+#define  _OCTEON_DEVICE_H_
+
+#include <linux/interrupt.h>
+#include <net/devlink.h>
+
+/** PCI VendorId Device Id */
+#define  OCTEON_CN68XX_PCIID          0x91177d
+#define  OCTEON_CN66XX_PCIID          0x92177d
+#define  OCTEON_CN23XX_PCIID_PF       0x9702177d
+/** Driver identifies chips by these Ids, created by clubbing together
+ *  DeviceId+RevisionId; Where Revision Id is not used to distinguish
+ *  between chips, a value of 0 is used for revision id.
+ */
+#define  OCTEON_CN68XX                0x0091
+#define  OCTEON_CN66XX                0x0092
+#define  OCTEON_CN23XX_PF_VID         0x9702
+#define  OCTEON_CN23XX_VF_VID         0x9712
+
+/**RevisionId for the chips */
+#define  OCTEON_CN23XX_REV_1_0        0x00
+#define  OCTEON_CN23XX_REV_1_1        0x01
+#define  OCTEON_CN23XX_REV_2_0        0x80
+
+/**SubsystemId for the chips */
+#define	 OCTEON_CN2350_10GB_SUBSYS_ID_1	0X3177d
+#define	 OCTEON_CN2350_10GB_SUBSYS_ID_2	0X4177d
+#define	 OCTEON_CN2360_10GB_SUBSYS_ID	0X5177d
+#define	 OCTEON_CN2350_25GB_SUBSYS_ID	0X7177d
+#define	 OCTEON_CN2360_25GB_SUBSYS_ID	0X6177d
+
+/** Endian-swap modes supported by Octeon. */
+enum octeon_pci_swap_mode {
+	OCTEON_PCI_PASSTHROUGH = 0,
+	OCTEON_PCI_64BIT_SWAP = 1,
+	OCTEON_PCI_32BIT_BYTE_SWAP = 2,
+	OCTEON_PCI_32BIT_LW_SWAP = 3
+};
+
+enum lio_fw_state {
+	FW_IS_PRELOADED = 0,
+	FW_NEEDS_TO_BE_LOADED = 1,
+	FW_IS_BEING_LOADED = 2,
+	FW_HAS_BEEN_LOADED = 3,
+};
+
+enum {
+	OCTEON_CONFIG_TYPE_DEFAULT = 0,
+	NUM_OCTEON_CONFS,
+};
+
+#define  OCTEON_INPUT_INTR    (1)
+#define  OCTEON_OUTPUT_INTR   (2)
+#define  OCTEON_MBOX_INTR     (4)
+#define  OCTEON_ALL_INTR      0xff
+
+/*---------------   PCI BAR1 index registers -------------*/
+
+/* BAR1 Mask */
+#define    PCI_BAR1_ENABLE_CA            1
+#define    PCI_BAR1_ENDIAN_MODE          OCTEON_PCI_64BIT_SWAP
+#define    PCI_BAR1_ENTRY_VALID          1
+#define    PCI_BAR1_MASK                 ((PCI_BAR1_ENABLE_CA << 3)   \
+					    | (PCI_BAR1_ENDIAN_MODE << 1) \
+					    | PCI_BAR1_ENTRY_VALID)
+
+/** Octeon Device state.
+ *  Each octeon device goes through each of these states
+ *  as it is initialized.
+ */
+#define    OCT_DEV_BEGIN_STATE            0x0
+#define    OCT_DEV_PCI_ENABLE_DONE        0x1
+#define    OCT_DEV_PCI_MAP_DONE           0x2
+#define    OCT_DEV_DISPATCH_INIT_DONE     0x3
+#define    OCT_DEV_INSTR_QUEUE_INIT_DONE  0x4
+#define    OCT_DEV_SC_BUFF_POOL_INIT_DONE 0x5
+#define    OCT_DEV_RESP_LIST_INIT_DONE    0x6
+#define    OCT_DEV_DROQ_INIT_DONE         0x7
+#define    OCT_DEV_MBOX_SETUP_DONE        0x8
+#define    OCT_DEV_MSIX_ALLOC_VECTOR_DONE 0x9
+#define    OCT_DEV_INTR_SET_DONE          0xa
+#define    OCT_DEV_IO_QUEUES_DONE         0xb
+#define    OCT_DEV_CONSOLE_INIT_DONE      0xc
+#define    OCT_DEV_HOST_OK                0xd
+#define    OCT_DEV_CORE_OK                0xe
+#define    OCT_DEV_RUNNING                0xf
+#define    OCT_DEV_IN_RESET               0x10
+#define    OCT_DEV_STATE_INVALID          0x11
+
+#define    OCT_DEV_STATES                 OCT_DEV_STATE_INVALID
+
+/** Octeon Device interrupts
+ * These interrupt bits are set in int_status filed of
+ * octeon_device structure
+ */
+#define	   OCT_DEV_INTR_DMA0_FORCE	  0x01
+#define	   OCT_DEV_INTR_DMA1_FORCE	  0x02
+#define	   OCT_DEV_INTR_PKT_DATA	  0x04
+
+#define LIO_RESET_SECS (3)
+
+/*---------------------------DISPATCH LIST-------------------------------*/
+
+/** The dispatch list entry.
+ *  The driver keeps a record of functions registered for each
+ *  response header opcode in this structure. Since the opcode is
+ *  hashed to index into the driver's list, more than one opcode
+ *  can hash to the same entry, in which case the list field points
+ *  to a linked list with the other entries.
+ */
+struct octeon_dispatch {
+	/** List head for this entry */
+	struct list_head list;
+
+	/** The opcode for which the dispatch function & arg should be used */
+	u16 opcode;
+
+	/** The function to be called for a packet received by the driver */
+	octeon_dispatch_fn_t dispatch_fn;
+
+	/* The application specified argument to be passed to the above
+	 * function along with the received packet
+	 */
+	void *arg;
+};
+
+/** The dispatch list structure. */
+struct octeon_dispatch_list {
+	/** access to dispatch list must be atomic */
+	spinlock_t lock;
+
+	/** Count of dispatch functions currently registered */
+	u32 count;
+
+	/** The list of dispatch functions */
+	struct octeon_dispatch *dlist;
+};
+
+/*-----------------------  THE OCTEON DEVICE  ---------------------------*/
+
+#define OCT_MEM_REGIONS     3
+/** PCI address space mapping information.
+ *  Each of the 3 address spaces given by BAR0, BAR2 and BAR4 of
+ *  Octeon gets mapped to different physical address spaces in
+ *  the kernel.
+ */
+struct octeon_mmio {
+	/** PCI address to which the BAR is mapped. */
+	u64 start;
+
+	/** Length of this PCI address space. */
+	u32 len;
+
+	/** Length that has been mapped to phys. address space. */
+	u32 mapped_len;
+
+	/** The physical address to which the PCI address space is mapped. */
+	u8 __iomem *hw_addr;
+
+	/** Flag indicating the mapping was successful. */
+	u32 done;
+};
+
+#define   MAX_OCTEON_MAPS    32
+
+struct octeon_io_enable {
+	u64 iq;
+	u64 oq;
+	u64 iq64B;
+};
+
+struct octeon_reg_list {
+	u32 __iomem *pci_win_wr_addr_hi;
+	u32 __iomem *pci_win_wr_addr_lo;
+	u64 __iomem *pci_win_wr_addr;
+
+	u32 __iomem *pci_win_rd_addr_hi;
+	u32 __iomem *pci_win_rd_addr_lo;
+	u64 __iomem *pci_win_rd_addr;
+
+	u32 __iomem *pci_win_wr_data_hi;
+	u32 __iomem *pci_win_wr_data_lo;
+	u64 __iomem *pci_win_wr_data;
+
+	u32 __iomem *pci_win_rd_data_hi;
+	u32 __iomem *pci_win_rd_data_lo;
+	u64 __iomem *pci_win_rd_data;
+};
+
+#define OCTEON_CONSOLE_MAX_READ_BYTES 512
+typedef int (*octeon_console_print_fn)(struct octeon_device *oct,
+				       u32 num, char *pre, char *suf);
+struct octeon_console {
+	u32 active;
+	u32 waiting;
+	u64 addr;
+	u32 buffer_size;
+	u64 input_base_addr;
+	u64 output_base_addr;
+	octeon_console_print_fn print;
+	char leftover[OCTEON_CONSOLE_MAX_READ_BYTES];
+};
+
+struct octeon_board_info {
+	char name[OCT_BOARD_NAME];
+	char serial_number[OCT_SERIAL_LEN];
+	u64 major;
+	u64 minor;
+};
+
+struct octeon_fn_list {
+	void (*setup_iq_regs)(struct octeon_device *, u32);
+	void (*setup_oq_regs)(struct octeon_device *, u32);
+
+	irqreturn_t (*process_interrupt_regs)(void *);
+	u64 (*msix_interrupt_handler)(void *);
+
+	int (*setup_mbox)(struct octeon_device *);
+	int (*free_mbox)(struct octeon_device *);
+
+	int (*soft_reset)(struct octeon_device *);
+	int (*setup_device_regs)(struct octeon_device *);
+	void (*bar1_idx_setup)(struct octeon_device *, u64, u32, int);
+	void (*bar1_idx_write)(struct octeon_device *, u32, u32);
+	u32 (*bar1_idx_read)(struct octeon_device *, u32);
+	u32 (*update_iq_read_idx)(struct octeon_instr_queue *);
+
+	void (*enable_oq_pkt_time_intr)(struct octeon_device *, u32);
+	void (*disable_oq_pkt_time_intr)(struct octeon_device *, u32);
+
+	void (*enable_interrupt)(struct octeon_device *, u8);
+	void (*disable_interrupt)(struct octeon_device *, u8);
+
+	int (*enable_io_queues)(struct octeon_device *);
+	void (*disable_io_queues)(struct octeon_device *);
+};
+
+/* Must be multiple of 8, changing breaks ABI */
+#define CVMX_BOOTMEM_NAME_LEN 128
+
+/* Structure for named memory blocks
+ * Number of descriptors
+ * available can be changed without affecting compatibility,
+ * but name length changes require a bump in the bootmem
+ * descriptor version
+ * Note: This structure must be naturally 64 bit aligned, as a single
+ * memory image will be used by both 32 and 64 bit programs.
+ */
+struct cvmx_bootmem_named_block_desc {
+	/** Base address of named block */
+	u64 base_addr;
+
+	/** Size actually allocated for named block */
+	u64 size;
+
+	/** name of named block */
+	char name[CVMX_BOOTMEM_NAME_LEN];
+};
+
+struct oct_fw_info {
+	u32 max_nic_ports;      /** max nic ports for the device */
+	u32 num_gmx_ports;      /** num gmx ports */
+	u64 app_cap_flags;      /** firmware cap flags */
+
+	/** The core application is running in this mode.
+	 * See octeon-drv-opcodes.h for values.
+	 */
+	u32 app_mode;
+	char   liquidio_firmware_version[32];
+	/* Fields extracted from legacy string 'liquidio_firmware_version' */
+	struct {
+		u8  maj;
+		u8  min;
+		u8  rev;
+	} ver;
+};
+
+#define OCT_FW_VER(maj, min, rev) \
+	(((u32)(maj) << 16) | ((u32)(min) << 8) | ((u32)(rev)))
+
+/* wrappers around work structs */
+struct cavium_wk {
+	struct delayed_work work;
+	void *ctxptr;
+	u64 ctxul;
+};
+
+struct cavium_wq {
+	struct workqueue_struct *wq;
+	struct cavium_wk wk;
+};
+
+struct octdev_props {
+	/* Each interface in the Octeon device has a network
+	 * device pointer (used for OS specific calls).
+	 */
+	int    rx_on;
+	int    fec;
+	int    fec_boot;
+	int    napi_enabled;
+	int    gmxport;
+	struct net_device *netdev;
+};
+
+#define LIO_FLAG_MSIX_ENABLED	0x1
+#define MSIX_PO_INT		0x1
+#define MSIX_PI_INT		0x2
+#define MSIX_MBOX_INT		0x4
+
+struct octeon_pf_vf_hs_word {
+#ifdef __LITTLE_ENDIAN_BITFIELD
+	/** PKIND value assigned for the DPI interface */
+	u64        pkind : 8;
+
+	/** OCTEON core clock multiplier   */
+	u64        core_tics_per_us : 16;
+
+	/** OCTEON coprocessor clock multiplier  */
+	u64        coproc_tics_per_us : 16;
+
+	/** app that currently running on OCTEON  */
+	u64        app_mode : 8;
+
+	/** RESERVED */
+	u64 reserved : 16;
+
+#else
+
+	/** RESERVED */
+	u64 reserved : 16;
+
+	/** app that currently running on OCTEON  */
+	u64        app_mode : 8;
+
+	/** OCTEON coprocessor clock multiplier  */
+	u64        coproc_tics_per_us : 16;
+
+	/** OCTEON core clock multiplier   */
+	u64        core_tics_per_us : 16;
+
+	/** PKIND value assigned for the DPI interface */
+	u64        pkind : 8;
+#endif
+};
+
+struct octeon_sriov_info {
+	/* Number of rings assigned to VF */
+	u32	rings_per_vf;
+
+	/** Max Number of VF devices that can be enabled. This variable can
+	 *  specified during load time or it will be derived after allocating
+	 *  PF queues. When max_vfs is derived then each VF will get one queue
+	 **/
+	u32	max_vfs;
+
+	/** Number of VF devices enabled using sysfs. */
+	u32	num_vfs_alloced;
+
+	/* Actual rings left for PF device */
+	u32	num_pf_rings;
+
+	/* SRN of PF usable IO queues */
+	u32	pf_srn;
+
+	/* total pf rings */
+	u32	trs;
+
+	u32	sriov_enabled;
+
+	struct lio_trusted_vf	trusted_vf;
+
+	/*lookup table that maps DPI ring number to VF pci_dev struct pointer*/
+	struct pci_dev *dpiring_to_vfpcidev_lut[MAX_POSSIBLE_VFS];
+
+	u64	vf_macaddr[MAX_POSSIBLE_VFS];
+
+	u16	vf_vlantci[MAX_POSSIBLE_VFS];
+
+	int	vf_linkstate[MAX_POSSIBLE_VFS];
+
+	bool    vf_spoofchk[MAX_POSSIBLE_VFS];
+
+	u64	vf_drv_loaded_mask;
+};
+
+struct octeon_ioq_vector {
+	struct octeon_device   *oct_dev;
+	int		        iq_index;
+	int		        droq_index;
+	int			vector;
+	struct octeon_mbox     *mbox;
+	struct cpumask		affinity_mask;
+	u32			ioq_num;
+};
+
+struct lio_vf_rep_list {
+	int num_vfs;
+	struct net_device *ndev[CN23XX_MAX_VFS_PER_PF];
+};
+
+struct lio_devlink_priv {
+	struct octeon_device *oct;
+};
+
+/** The Octeon device.
+ *  Each Octeon device has this structure to represent all its
+ *  components.
+ */
+struct octeon_device {
+	/** Lock for PCI window configuration accesses */
+	spinlock_t pci_win_lock;
+
+	/** Lock for memory accesses */
+	spinlock_t mem_access_lock;
+
+	/** PCI device pointer */
+	struct pci_dev *pci_dev;
+
+	/** Chip specific information. */
+	void *chip;
+
+	/** Number of interfaces detected in this octeon device. */
+	u32 ifcount;
+
+	struct octdev_props props[MAX_OCTEON_LINKS];
+
+	/** Octeon Chip type. */
+	u16 chip_id;
+
+	u16 rev_id;
+
+	u32 subsystem_id;
+
+	u16 pf_num;
+
+	u16 vf_num;
+
+	/** This device's id - set by the driver. */
+	u32 octeon_id;
+
+	/** This device's PCIe port used for traffic. */
+	u16 pcie_port;
+
+	u16 flags;
+#define LIO_FLAG_MSI_ENABLED                  (u32)(1 << 1)
+
+	/** The state of this device */
+	atomic_t status;
+
+	/** memory mapped io range */
+	struct octeon_mmio mmio[OCT_MEM_REGIONS];
+
+	struct octeon_reg_list reg_list;
+
+	struct octeon_fn_list fn_list;
+
+	struct octeon_board_info boardinfo;
+
+	u32 num_iqs;
+
+	/* The pool containing pre allocated buffers used for soft commands */
+	struct octeon_sc_buffer_pool	sc_buf_pool;
+
+	/** The input instruction queues */
+	struct octeon_instr_queue *instr_queue
+		[MAX_POSSIBLE_OCTEON_INSTR_QUEUES];
+
+	/** The doubly-linked list of instruction response */
+	struct octeon_response_list response_list[MAX_RESPONSE_LISTS];
+
+	u32 num_oqs;
+
+	/** The DROQ output queues  */
+	struct octeon_droq *droq[MAX_POSSIBLE_OCTEON_OUTPUT_QUEUES];
+
+	struct octeon_io_enable io_qmask;
+
+	/** List of dispatch functions */
+	struct octeon_dispatch_list dispatch;
+
+	u32 int_status;
+
+	u64 droq_intr;
+
+	/** Physical location of the cvmx_bootmem_desc_t in octeon memory */
+	u64 bootmem_desc_addr;
+
+	/** Placeholder memory for named blocks.
+	 * Assumes single-threaded access
+	 */
+	struct cvmx_bootmem_named_block_desc bootmem_named_block_desc;
+
+	/** Address of consoles descriptor */
+	u64 console_desc_addr;
+
+	/** Number of consoles available. 0 means they are inaccessible */
+	u32 num_consoles;
+
+	/* Console caches */
+	struct octeon_console console[MAX_OCTEON_MAPS];
+
+	/* Console named block info */
+	struct {
+		u64 dram_region_base;
+		int bar1_index;
+	} console_nb_info;
+
+	/* Coprocessor clock rate. */
+	u64 coproc_clock_rate;
+
+	/** The core application is running in this mode. See liquidio_common.h
+	 * for values.
+	 */
+	u32 app_mode;
+
+	struct oct_fw_info fw_info;
+
+	/** The name given to this device. */
+	char device_name[32];
+
+	/** Application Context */
+	void *app_ctx;
+
+	struct cavium_wq dma_comp_wq;
+
+	/** Lock for dma response list */
+	spinlock_t cmd_resp_wqlock;
+	u32 cmd_resp_state;
+
+	struct cavium_wq check_db_wq[MAX_POSSIBLE_OCTEON_INSTR_QUEUES];
+
+	struct cavium_wk nic_poll_work;
+
+	struct cavium_wk console_poll_work[MAX_OCTEON_MAPS];
+
+	void *priv;
+
+	int num_msix_irqs;
+
+	void *msix_entries;
+
+	/* when requesting IRQs, the names are stored here */
+	void *irq_name_storage;
+
+	struct octeon_sriov_info sriov_info;
+
+	struct octeon_pf_vf_hs_word pfvf_hsword;
+
+	int msix_on;
+
+	/** Mail Box details of each octeon queue. */
+	struct octeon_mbox  *mbox[MAX_POSSIBLE_VFS];
+
+	/** IOq information of it's corresponding MSI-X interrupt. */
+	struct octeon_ioq_vector    *ioq_vector;
+
+	int rx_pause;
+	int tx_pause;
+
+	struct oct_link_stats link_stats; /*stastics from firmware*/
+
+	/* private flags to control driver-specific features through ethtool */
+	u32 priv_flags;
+
+	void *watchdog_task;
+
+	u32 rx_coalesce_usecs;
+	u32 rx_max_coalesced_frames;
+	u32 tx_max_coalesced_frames;
+
+	bool cores_crashed;
+
+	struct {
+		int bus;
+		int dev;
+		int func;
+	} loc;
+
+	atomic_t *adapter_refcount; /* reference count of adapter */
+
+	atomic_t *adapter_fw_state; /* per-adapter, lio_fw_state */
+
+	bool ptp_enable;
+
+	struct lio_vf_rep_list vf_rep_list;
+	struct devlink *devlink;
+	enum devlink_eswitch_mode eswitch_mode;
+
+	/* for 25G NIC speed change */
+	u8  speed_boot;
+	u8  speed_setting;
+	u8  no_speed_setting;
+
+	u32    vfstats_poll;
+#define LIO_VFSTATS_POLL 10
+};
+
+#define  OCT_DRV_ONLINE 1
+#define  OCT_DRV_OFFLINE 2
+#define  OCTEON_CN6XXX(oct)	({					\
+				 typeof(oct) _oct = (oct);		\
+				 ((_oct->chip_id == OCTEON_CN66XX) ||	\
+				  (_oct->chip_id == OCTEON_CN68XX));	})
+#define  OCTEON_CN23XX_PF(oct)        ((oct)->chip_id == OCTEON_CN23XX_PF_VID)
+#define  OCTEON_CN23XX_VF(oct)        ((oct)->chip_id == OCTEON_CN23XX_VF_VID)
+#define CHIP_CONF(oct, TYPE)             \
+	(((struct octeon_ ## TYPE  *)((oct)->chip))->conf)
+
+#define MAX_IO_PENDING_PKT_COUNT 100
+
+/*------------------ Function Prototypes ----------------------*/
+
+/** Initialize device list memory */
+void octeon_init_device_list(int conf_type);
+
+/** Free memory for Input and Output queue structures for a octeon device */
+void octeon_free_device_mem(struct octeon_device *oct);
+
+/* Look up a free entry in the octeon_device table and allocate resources
+ * for the octeon_device structure for an octeon device. Called at init
+ * time.
+ */
+struct octeon_device *octeon_allocate_device(u32 pci_id,
+					     u32 priv_size);
+
+/** Register a device's bus location at initialization time.
+ *  @param octeon_dev - pointer to the octeon device structure.
+ *  @param bus        - PCIe bus #
+ *  @param dev        - PCIe device #
+ *  @param func       - PCIe function #
+ *  @param is_pf      - TRUE for PF, FALSE for VF
+ *  @return reference count of device's adapter
+ */
+int octeon_register_device(struct octeon_device *oct,
+			   int bus, int dev, int func, int is_pf);
+
+/** Deregister a device at de-initialization time.
+ *  @param octeon_dev - pointer to the octeon device structure.
+ *  @return reference count of device's adapter
+ */
+int octeon_deregister_device(struct octeon_device *oct);
+
+/**  Initialize the driver's dispatch list which is a mix of a hash table
+ *  and a linked list. This is done at driver load time.
+ *  @param octeon_dev - pointer to the octeon device structure.
+ *  @return 0 on success, else -ve error value
+ */
+int octeon_init_dispatch_list(struct octeon_device *octeon_dev);
+
+/**  Delete the driver's dispatch list and all registered entries.
+ * This is done at driver unload time.
+ *  @param octeon_dev - pointer to the octeon device structure.
+ */
+void octeon_delete_dispatch_list(struct octeon_device *octeon_dev);
+
+/** Initialize the core device fields with the info returned by the FW.
+ * @param recv_info - Receive info structure
+ * @param buf       - Receive buffer
+ */
+int octeon_core_drv_init(struct octeon_recv_info *recv_info, void *buf);
+
+/** Gets the dispatch function registered to receive packets with a
+ *  given opcode/subcode.
+ *  @param  octeon_dev  - the octeon device pointer.
+ *  @param  opcode      - the opcode for which the dispatch function
+ *                        is to checked.
+ *  @param  subcode     - the subcode for which the dispatch function
+ *                        is to checked.
+ *
+ *  @return Success: octeon_dispatch_fn_t (dispatch function pointer)
+ *  @return Failure: NULL
+ *
+ *  Looks up the dispatch list to get the dispatch function for a
+ *  given opcode.
+ */
+octeon_dispatch_fn_t
+octeon_get_dispatch(struct octeon_device *octeon_dev, u16 opcode,
+		    u16 subcode);
+
+/** Get the octeon device pointer.
+ *  @param octeon_id  - The id for which the octeon device pointer is required.
+ *  @return Success: Octeon device pointer.
+ *  @return Failure: NULL.
+ */
+struct octeon_device *lio_get_device(u32 octeon_id);
+
+/** Get the octeon id assigned to the octeon device passed as argument.
+ *  This function is exported to other modules.
+ *  @param dev - octeon device pointer passed as a void *.
+ *  @return octeon device id
+ */
+int lio_get_device_id(void *dev);
+
+/** Read windowed register.
+ *  @param  oct   -  pointer to the Octeon device.
+ *  @param  addr  -  Address of the register to read.
+ *
+ *  This routine is called to read from the indirectly accessed
+ *  Octeon registers that are visible through a PCI BAR0 mapped window
+ *  register.
+ *  @return  - 64 bit value read from the register.
+ */
+
+u64 lio_pci_readq(struct octeon_device *oct, u64 addr);
+
+/** Write windowed register.
+ *  @param  oct  -  pointer to the Octeon device.
+ *  @param  val  -  Value to write
+ *  @param  addr -  Address of the register to write
+ *
+ *  This routine is called to write to the indirectly accessed
+ *  Octeon registers that are visible through a PCI BAR0 mapped window
+ *  register.
+ *  @return   Nothing.
+ */
+void lio_pci_writeq(struct octeon_device *oct, u64 val, u64 addr);
+
+/* Routines for reading and writing CSRs */
+#define   octeon_write_csr(oct_dev, reg_off, value) \
+		writel(value, (oct_dev)->mmio[0].hw_addr + (reg_off))
+
+#define   octeon_write_csr64(oct_dev, reg_off, val64) \
+		writeq(val64, (oct_dev)->mmio[0].hw_addr + (reg_off))
+
+#define   octeon_read_csr(oct_dev, reg_off)         \
+		readl((oct_dev)->mmio[0].hw_addr + (reg_off))
+
+#define   octeon_read_csr64(oct_dev, reg_off)         \
+		readq((oct_dev)->mmio[0].hw_addr + (reg_off))
+
+/**
+ * Checks if memory access is okay
+ *
+ * @param oct which octeon to send to
+ * @return Zero on success, negative on failure.
+ */
+int octeon_mem_access_ok(struct octeon_device *oct);
+
+/**
+ * Waits for DDR initialization.
+ *
+ * @param oct which octeon to send to
+ * @param timeout_in_ms pointer to how long to wait until DDR is initialized
+ * in ms.
+ *                      If contents are 0, it waits until contents are non-zero
+ *                      before starting to check.
+ * @return Zero on success, negative on failure.
+ */
+int octeon_wait_for_ddr_init(struct octeon_device *oct,
+			     u32 *timeout_in_ms);
+
+/**
+ * Wait for u-boot to boot and be waiting for a command.
+ *
+ * @param wait_time_hundredths
+ *               Maximum time to wait
+ *
+ * @return Zero on success, negative on failure.
+ */
+int octeon_wait_for_bootloader(struct octeon_device *oct,
+			       u32 wait_time_hundredths);
+
+/**
+ * Initialize console access
+ *
+ * @param oct which octeon initialize
+ * @return Zero on success, negative on failure.
+ */
+int octeon_init_consoles(struct octeon_device *oct);
+
+/**
+ * Adds access to a console to the device.
+ *
+ * @param oct:          which octeon to add to
+ * @param console_num:  which console
+ * @param dbg_enb:      ptr to debug enablement string, one of:
+ *                    * NULL for no debug output (i.e. disabled)
+ *                    * empty string enables debug output (via default method)
+ *                    * specific string to enable debug console output
+ *
+ * @return Zero on success, negative on failure.
+ */
+int octeon_add_console(struct octeon_device *oct, u32 console_num,
+		       char *dbg_enb);
+
+/** write or read from a console */
+int octeon_console_write(struct octeon_device *oct, u32 console_num,
+			 char *buffer, u32 write_request_size, u32 flags);
+int octeon_console_write_avail(struct octeon_device *oct, u32 console_num);
+
+int octeon_console_read_avail(struct octeon_device *oct, u32 console_num);
+
+/** Removes all attached consoles. */
+void octeon_remove_consoles(struct octeon_device *oct);
+
+/**
+ * Send a string to u-boot on console 0 as a command.
+ *
+ * @param oct which octeon to send to
+ * @param cmd_str String to send
+ * @param wait_hundredths Time to wait for u-boot to accept the command.
+ *
+ * @return Zero on success, negative on failure.
+ */
+int octeon_console_send_cmd(struct octeon_device *oct, char *cmd_str,
+			    u32 wait_hundredths);
+
+/** Parses, validates, and downloads firmware, then boots associated cores.
+ *  @param oct which octeon to download firmware to
+ *  @param data  - The complete firmware file image
+ *  @param size  - The size of the data
+ *
+ *  @return 0 if success.
+ *         -EINVAL if file is incompatible or badly formatted.
+ *         -ENODEV if no handler was found for the application type or an
+ *         invalid octeon id was passed.
+ */
+int octeon_download_firmware(struct octeon_device *oct, const u8 *data,
+			     size_t size);
+
+char *lio_get_state_string(atomic_t *state_ptr);
+
+/** Sets up instruction queues for the device
+ *  @param oct which octeon to setup
+ *
+ *  @return 0 if success. 1 if fails
+ */
+int octeon_setup_instr_queues(struct octeon_device *oct);
+
+/** Sets up output queues for the device
+ *  @param oct which octeon to setup
+ *
+ *  @return 0 if success. 1 if fails
+ */
+int octeon_setup_output_queues(struct octeon_device *oct);
+
+int octeon_get_tx_qsize(struct octeon_device *oct, u32 q_no);
+
+int octeon_get_rx_qsize(struct octeon_device *oct, u32 q_no);
+
+/** Turns off the input and output queues for the device
+ *  @param oct which octeon to disable
+ */
+int octeon_set_io_queues_off(struct octeon_device *oct);
+
+/** Turns on or off the given output queue for the device
+ *  @param oct which octeon to change
+ *  @param q_no which queue
+ *  @param enable 1 to enable, 0 to disable
+ */
+void octeon_set_droq_pkt_op(struct octeon_device *oct, u32 q_no, u32 enable);
+
+/** Retrieve the config for the device
+ *  @param oct which octeon
+ *  @param card_type type of card
+ *
+ *  @returns pointer to configuration
+ */
+void *oct_get_config_info(struct octeon_device *oct, u16 card_type);
+
+/** Gets the octeon device configuration
+ *  @return - pointer to the octeon configuration struture
+ */
+struct octeon_config *octeon_get_conf(struct octeon_device *oct);
+
+void octeon_free_ioq_vector(struct octeon_device *oct);
+int octeon_allocate_ioq_vector(struct octeon_device  *oct, u32 num_ioqs);
+void lio_enable_irq(struct octeon_droq *droq, struct octeon_instr_queue *iq);
+
+/* LiquidIO driver pivate flags */
+enum {
+	OCT_PRIV_FLAG_TX_BYTES = 0, /* Tx interrupts by pending byte count */
+};
+
+#define OCT_PRIV_FLAG_DEFAULT 0x0
+
+static inline u32 lio_get_priv_flag(struct octeon_device *octdev, u32 flag)
+{
+	return !!(octdev->priv_flags & (0x1 << flag));
+}
+
+static inline void lio_set_priv_flag(struct octeon_device *octdev,
+				     u32 flag, u32 val)
+{
+	if (val)
+		octdev->priv_flags |= (0x1 << flag);
+	else
+		octdev->priv_flags &= ~(0x1 << flag);
+}
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_droq.c b/drivers/net/ethernet/cavium/liquidio/octeon_droq.c
new file mode 100644
index 000000000..d4080bddc
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_droq.c
@@ -0,0 +1,969 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include <linux/vmalloc.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_main.h"
+#include "octeon_network.h"
+#include "cn66xx_regs.h"
+#include "cn66xx_device.h"
+#include "cn23xx_pf_device.h"
+#include "cn23xx_vf_device.h"
+
+struct niclist {
+	struct list_head list;
+	void *ptr;
+};
+
+struct __dispatch {
+	struct list_head list;
+	struct octeon_recv_info *rinfo;
+	octeon_dispatch_fn_t disp_fn;
+};
+
+/** Get the argument that the user set when registering dispatch
+ *  function for a given opcode/subcode.
+ *  @param  octeon_dev - the octeon device pointer.
+ *  @param  opcode     - the opcode for which the dispatch argument
+ *                       is to be checked.
+ *  @param  subcode    - the subcode for which the dispatch argument
+ *                       is to be checked.
+ *  @return  Success: void * (argument to the dispatch function)
+ *  @return  Failure: NULL
+ *
+ */
+void *octeon_get_dispatch_arg(struct octeon_device *octeon_dev,
+			      u16 opcode, u16 subcode)
+{
+	int idx;
+	struct list_head *dispatch;
+	void *fn_arg = NULL;
+	u16 combined_opcode = OPCODE_SUBCODE(opcode, subcode);
+
+	idx = combined_opcode & OCTEON_OPCODE_MASK;
+
+	spin_lock_bh(&octeon_dev->dispatch.lock);
+
+	if (octeon_dev->dispatch.count == 0) {
+		spin_unlock_bh(&octeon_dev->dispatch.lock);
+		return NULL;
+	}
+
+	if (octeon_dev->dispatch.dlist[idx].opcode == combined_opcode) {
+		fn_arg = octeon_dev->dispatch.dlist[idx].arg;
+	} else {
+		list_for_each(dispatch,
+			      &octeon_dev->dispatch.dlist[idx].list) {
+			if (((struct octeon_dispatch *)dispatch)->opcode ==
+			    combined_opcode) {
+				fn_arg = ((struct octeon_dispatch *)
+					  dispatch)->arg;
+				break;
+			}
+		}
+	}
+
+	spin_unlock_bh(&octeon_dev->dispatch.lock);
+	return fn_arg;
+}
+
+/** Check for packets on Droq. This function should be called with lock held.
+ *  @param  droq - Droq on which count is checked.
+ *  @return Returns packet count.
+ */
+u32 octeon_droq_check_hw_for_pkts(struct octeon_droq *droq)
+{
+	u32 pkt_count = 0;
+	u32 last_count;
+
+	pkt_count = readl(droq->pkts_sent_reg);
+
+	last_count = pkt_count - droq->pkt_count;
+	droq->pkt_count = pkt_count;
+
+	/* we shall write to cnts  at napi irq enable or end of droq tasklet */
+	if (last_count)
+		atomic_add(last_count, &droq->pkts_pending);
+
+	return last_count;
+}
+
+static void octeon_droq_compute_max_packet_bufs(struct octeon_droq *droq)
+{
+	u32 count = 0;
+
+	/* max_empty_descs is the max. no. of descs that can have no buffers.
+	 * If the empty desc count goes beyond this value, we cannot safely
+	 * read in a 64K packet sent by Octeon
+	 * (64K is max pkt size from Octeon)
+	 */
+	droq->max_empty_descs = 0;
+
+	do {
+		droq->max_empty_descs++;
+		count += droq->buffer_size;
+	} while (count < (64 * 1024));
+
+	droq->max_empty_descs = droq->max_count - droq->max_empty_descs;
+}
+
+static void octeon_droq_reset_indices(struct octeon_droq *droq)
+{
+	droq->read_idx = 0;
+	droq->write_idx = 0;
+	droq->refill_idx = 0;
+	droq->refill_count = 0;
+	atomic_set(&droq->pkts_pending, 0);
+}
+
+static void
+octeon_droq_destroy_ring_buffers(struct octeon_device *oct,
+				 struct octeon_droq *droq)
+{
+	u32 i;
+	struct octeon_skb_page_info *pg_info;
+
+	for (i = 0; i < droq->max_count; i++) {
+		pg_info = &droq->recv_buf_list[i].pg_info;
+		if (!pg_info)
+			continue;
+
+		if (pg_info->dma)
+			lio_unmap_ring(oct->pci_dev,
+				       (u64)pg_info->dma);
+		pg_info->dma = 0;
+
+		if (pg_info->page)
+			recv_buffer_destroy(droq->recv_buf_list[i].buffer,
+					    pg_info);
+
+		droq->recv_buf_list[i].buffer = NULL;
+	}
+
+	octeon_droq_reset_indices(droq);
+}
+
+static int
+octeon_droq_setup_ring_buffers(struct octeon_device *oct,
+			       struct octeon_droq *droq)
+{
+	u32 i;
+	void *buf;
+	struct octeon_droq_desc *desc_ring = droq->desc_ring;
+
+	for (i = 0; i < droq->max_count; i++) {
+		buf = recv_buffer_alloc(oct, &droq->recv_buf_list[i].pg_info);
+
+		if (!buf) {
+			dev_err(&oct->pci_dev->dev, "%s buffer alloc failed\n",
+				__func__);
+			droq->stats.rx_alloc_failure++;
+			return -ENOMEM;
+		}
+
+		droq->recv_buf_list[i].buffer = buf;
+		droq->recv_buf_list[i].data = get_rbd(buf);
+		desc_ring[i].info_ptr = 0;
+		desc_ring[i].buffer_ptr =
+			lio_map_ring(droq->recv_buf_list[i].buffer);
+	}
+
+	octeon_droq_reset_indices(droq);
+
+	octeon_droq_compute_max_packet_bufs(droq);
+
+	return 0;
+}
+
+int octeon_delete_droq(struct octeon_device *oct, u32 q_no)
+{
+	struct octeon_droq *droq = oct->droq[q_no];
+
+	dev_dbg(&oct->pci_dev->dev, "%s[%d]\n", __func__, q_no);
+
+	octeon_droq_destroy_ring_buffers(oct, droq);
+	vfree(droq->recv_buf_list);
+
+	if (droq->desc_ring)
+		lio_dma_free(oct, (droq->max_count * OCT_DROQ_DESC_SIZE),
+			     droq->desc_ring, droq->desc_ring_dma);
+
+	memset(droq, 0, OCT_DROQ_SIZE);
+	oct->io_qmask.oq &= ~(1ULL << q_no);
+	vfree(oct->droq[q_no]);
+	oct->droq[q_no] = NULL;
+	oct->num_oqs--;
+
+	return 0;
+}
+
+int octeon_init_droq(struct octeon_device *oct,
+		     u32 q_no,
+		     u32 num_descs,
+		     u32 desc_size,
+		     void *app_ctx)
+{
+	struct octeon_droq *droq;
+	u32 desc_ring_size = 0, c_num_descs = 0, c_buf_size = 0;
+	u32 c_pkts_per_intr = 0, c_refill_threshold = 0;
+	int numa_node = dev_to_node(&oct->pci_dev->dev);
+
+	dev_dbg(&oct->pci_dev->dev, "%s[%d]\n", __func__, q_no);
+
+	droq = oct->droq[q_no];
+	memset(droq, 0, OCT_DROQ_SIZE);
+
+	droq->oct_dev = oct;
+	droq->q_no = q_no;
+	if (app_ctx)
+		droq->app_ctx = app_ctx;
+	else
+		droq->app_ctx = (void *)(size_t)q_no;
+
+	c_num_descs = num_descs;
+	c_buf_size = desc_size;
+	if (OCTEON_CN6XXX(oct)) {
+		struct octeon_config *conf6x = CHIP_CONF(oct, cn6xxx);
+
+		c_pkts_per_intr = (u32)CFG_GET_OQ_PKTS_PER_INTR(conf6x);
+		c_refill_threshold =
+			(u32)CFG_GET_OQ_REFILL_THRESHOLD(conf6x);
+	} else if (OCTEON_CN23XX_PF(oct)) {
+		struct octeon_config *conf23 = CHIP_CONF(oct, cn23xx_pf);
+
+		c_pkts_per_intr = (u32)CFG_GET_OQ_PKTS_PER_INTR(conf23);
+		c_refill_threshold = (u32)CFG_GET_OQ_REFILL_THRESHOLD(conf23);
+	} else if (OCTEON_CN23XX_VF(oct)) {
+		struct octeon_config *conf23 = CHIP_CONF(oct, cn23xx_vf);
+
+		c_pkts_per_intr = (u32)CFG_GET_OQ_PKTS_PER_INTR(conf23);
+		c_refill_threshold = (u32)CFG_GET_OQ_REFILL_THRESHOLD(conf23);
+	} else {
+		return 1;
+	}
+
+	droq->max_count = c_num_descs;
+	droq->buffer_size = c_buf_size;
+
+	desc_ring_size = droq->max_count * OCT_DROQ_DESC_SIZE;
+	droq->desc_ring = lio_dma_alloc(oct, desc_ring_size,
+					(dma_addr_t *)&droq->desc_ring_dma);
+
+	if (!droq->desc_ring) {
+		dev_err(&oct->pci_dev->dev,
+			"Output queue %d ring alloc failed\n", q_no);
+		return 1;
+	}
+
+	dev_dbg(&oct->pci_dev->dev, "droq[%d]: desc_ring: virt: 0x%p, dma: %lx\n",
+		q_no, droq->desc_ring, droq->desc_ring_dma);
+	dev_dbg(&oct->pci_dev->dev, "droq[%d]: num_desc: %d\n", q_no,
+		droq->max_count);
+
+	droq->recv_buf_list = vzalloc_node(array_size(droq->max_count, OCT_DROQ_RECVBUF_SIZE),
+					   numa_node);
+	if (!droq->recv_buf_list)
+		droq->recv_buf_list = vzalloc(array_size(droq->max_count, OCT_DROQ_RECVBUF_SIZE));
+	if (!droq->recv_buf_list) {
+		dev_err(&oct->pci_dev->dev, "Output queue recv buf list alloc failed\n");
+		goto init_droq_fail;
+	}
+
+	if (octeon_droq_setup_ring_buffers(oct, droq))
+		goto init_droq_fail;
+
+	droq->pkts_per_intr = c_pkts_per_intr;
+	droq->refill_threshold = c_refill_threshold;
+
+	dev_dbg(&oct->pci_dev->dev, "DROQ INIT: max_empty_descs: %d\n",
+		droq->max_empty_descs);
+
+	INIT_LIST_HEAD(&droq->dispatch_list);
+
+	/* For 56xx Pass1, this function won't be called, so no checks. */
+	oct->fn_list.setup_oq_regs(oct, q_no);
+
+	oct->io_qmask.oq |= BIT_ULL(q_no);
+
+	return 0;
+
+init_droq_fail:
+	octeon_delete_droq(oct, q_no);
+	return 1;
+}
+
+/* octeon_create_recv_info
+ * Parameters:
+ *  octeon_dev - pointer to the octeon device structure
+ *  droq       - droq in which the packet arrived.
+ *  buf_cnt    - no. of buffers used by the packet.
+ *  idx        - index in the descriptor for the first buffer in the packet.
+ * Description:
+ *  Allocates a recv_info_t and copies the buffer addresses for packet data
+ *  into the recv_pkt space which starts at an 8B offset from recv_info_t.
+ *  Flags the descriptors for refill later. If available descriptors go
+ *  below the threshold to receive a 64K pkt, new buffers are first allocated
+ *  before the recv_pkt_t is created.
+ *  This routine will be called in interrupt context.
+ * Returns:
+ *  Success: Pointer to recv_info_t
+ *  Failure: NULL.
+ */
+static inline struct octeon_recv_info *octeon_create_recv_info(
+		struct octeon_device *octeon_dev,
+		struct octeon_droq *droq,
+		u32 buf_cnt,
+		u32 idx)
+{
+	struct octeon_droq_info *info;
+	struct octeon_recv_pkt *recv_pkt;
+	struct octeon_recv_info *recv_info;
+	u32 i, bytes_left;
+	struct octeon_skb_page_info *pg_info;
+
+	info = (struct octeon_droq_info *)droq->recv_buf_list[idx].data;
+
+	recv_info = octeon_alloc_recv_info(sizeof(struct __dispatch));
+	if (!recv_info)
+		return NULL;
+
+	recv_pkt = recv_info->recv_pkt;
+	recv_pkt->rh = info->rh;
+	recv_pkt->length = (u32)info->length;
+	recv_pkt->buffer_count = (u16)buf_cnt;
+	recv_pkt->octeon_id = (u16)octeon_dev->octeon_id;
+
+	i = 0;
+	bytes_left = (u32)info->length;
+
+	while (buf_cnt) {
+		{
+			pg_info = &droq->recv_buf_list[idx].pg_info;
+
+			lio_unmap_ring(octeon_dev->pci_dev,
+				       (u64)pg_info->dma);
+			pg_info->page = NULL;
+			pg_info->dma = 0;
+		}
+
+		recv_pkt->buffer_size[i] =
+			(bytes_left >=
+			 droq->buffer_size) ? droq->buffer_size : bytes_left;
+
+		recv_pkt->buffer_ptr[i] = droq->recv_buf_list[idx].buffer;
+		droq->recv_buf_list[idx].buffer = NULL;
+
+		idx = incr_index(idx, 1, droq->max_count);
+		bytes_left -= droq->buffer_size;
+		i++;
+		buf_cnt--;
+	}
+
+	return recv_info;
+}
+
+/* If we were not able to refill all buffers, try to move around
+ * the buffers that were not dispatched.
+ */
+static inline u32
+octeon_droq_refill_pullup_descs(struct octeon_droq *droq,
+				struct octeon_droq_desc *desc_ring)
+{
+	u32 desc_refilled = 0;
+
+	u32 refill_index = droq->refill_idx;
+
+	while (refill_index != droq->read_idx) {
+		if (droq->recv_buf_list[refill_index].buffer) {
+			droq->recv_buf_list[droq->refill_idx].buffer =
+				droq->recv_buf_list[refill_index].buffer;
+			droq->recv_buf_list[droq->refill_idx].data =
+				droq->recv_buf_list[refill_index].data;
+			desc_ring[droq->refill_idx].buffer_ptr =
+				desc_ring[refill_index].buffer_ptr;
+			droq->recv_buf_list[refill_index].buffer = NULL;
+			desc_ring[refill_index].buffer_ptr = 0;
+			do {
+				droq->refill_idx = incr_index(droq->refill_idx,
+							      1,
+							      droq->max_count);
+				desc_refilled++;
+				droq->refill_count--;
+			} while (droq->recv_buf_list[droq->refill_idx].buffer);
+		}
+		refill_index = incr_index(refill_index, 1, droq->max_count);
+	}                       /* while */
+	return desc_refilled;
+}
+
+/* octeon_droq_refill
+ * Parameters:
+ *  droq       - droq in which descriptors require new buffers.
+ * Description:
+ *  Called during normal DROQ processing in interrupt mode or by the poll
+ *  thread to refill the descriptors from which buffers were dispatched
+ *  to upper layers. Attempts to allocate new buffers. If that fails, moves
+ *  up buffers (that were not dispatched) to form a contiguous ring.
+ * Returns:
+ *  No of descriptors refilled.
+ */
+static u32
+octeon_droq_refill(struct octeon_device *octeon_dev, struct octeon_droq *droq)
+{
+	struct octeon_droq_desc *desc_ring;
+	void *buf = NULL;
+	u8 *data;
+	u32 desc_refilled = 0;
+	struct octeon_skb_page_info *pg_info;
+
+	desc_ring = droq->desc_ring;
+
+	while (droq->refill_count && (desc_refilled < droq->max_count)) {
+		/* If a valid buffer exists (happens if there is no dispatch),
+		 * reuse the buffer, else allocate.
+		 */
+		if (!droq->recv_buf_list[droq->refill_idx].buffer) {
+			pg_info =
+				&droq->recv_buf_list[droq->refill_idx].pg_info;
+			/* Either recycle the existing pages or go for
+			 * new page alloc
+			 */
+			if (pg_info->page)
+				buf = recv_buffer_reuse(octeon_dev, pg_info);
+			else
+				buf = recv_buffer_alloc(octeon_dev, pg_info);
+			/* If a buffer could not be allocated, no point in
+			 * continuing
+			 */
+			if (!buf) {
+				droq->stats.rx_alloc_failure++;
+				break;
+			}
+			droq->recv_buf_list[droq->refill_idx].buffer =
+				buf;
+			data = get_rbd(buf);
+		} else {
+			data = get_rbd(droq->recv_buf_list
+				       [droq->refill_idx].buffer);
+		}
+
+		droq->recv_buf_list[droq->refill_idx].data = data;
+
+		desc_ring[droq->refill_idx].buffer_ptr =
+			lio_map_ring(droq->recv_buf_list[
+				     droq->refill_idx].buffer);
+
+		droq->refill_idx = incr_index(droq->refill_idx, 1,
+					      droq->max_count);
+		desc_refilled++;
+		droq->refill_count--;
+	}
+
+	if (droq->refill_count)
+		desc_refilled +=
+			octeon_droq_refill_pullup_descs(droq, desc_ring);
+
+	/* if droq->refill_count
+	 * The refill count would not change in pass two. We only moved buffers
+	 * to close the gap in the ring, but we would still have the same no. of
+	 * buffers to refill.
+	 */
+	return desc_refilled;
+}
+
+/** check if we can allocate packets to get out of oom.
+ *  @param  droq - Droq being checked.
+ *  @return 1 if fails to refill minimum
+ */
+int octeon_retry_droq_refill(struct octeon_droq *droq)
+{
+	struct octeon_device *oct = droq->oct_dev;
+	int desc_refilled, reschedule = 1;
+	u32 pkts_credit;
+
+	pkts_credit = readl(droq->pkts_credit_reg);
+	desc_refilled = octeon_droq_refill(oct, droq);
+	if (desc_refilled) {
+		/* Flush the droq descriptor data to memory to be sure
+		 * that when we update the credits the data in memory
+		 * is accurate.
+		 */
+		wmb();
+		writel(desc_refilled, droq->pkts_credit_reg);
+
+		if (pkts_credit + desc_refilled >= CN23XX_SLI_DEF_BP)
+			reschedule = 0;
+	}
+
+	return reschedule;
+}
+
+static inline u32
+octeon_droq_get_bufcount(u32 buf_size, u32 total_len)
+{
+	return DIV_ROUND_UP(total_len, buf_size);
+}
+
+static int
+octeon_droq_dispatch_pkt(struct octeon_device *oct,
+			 struct octeon_droq *droq,
+			 union octeon_rh *rh,
+			 struct octeon_droq_info *info)
+{
+	u32 cnt;
+	octeon_dispatch_fn_t disp_fn;
+	struct octeon_recv_info *rinfo;
+
+	cnt = octeon_droq_get_bufcount(droq->buffer_size, (u32)info->length);
+
+	disp_fn = octeon_get_dispatch(oct, (u16)rh->r.opcode,
+				      (u16)rh->r.subcode);
+	if (disp_fn) {
+		rinfo = octeon_create_recv_info(oct, droq, cnt, droq->read_idx);
+		if (rinfo) {
+			struct __dispatch *rdisp = rinfo->rsvd;
+
+			rdisp->rinfo = rinfo;
+			rdisp->disp_fn = disp_fn;
+			rinfo->recv_pkt->rh = *rh;
+			list_add_tail(&rdisp->list,
+				      &droq->dispatch_list);
+		} else {
+			droq->stats.dropped_nomem++;
+		}
+	} else {
+		dev_err(&oct->pci_dev->dev, "DROQ: No dispatch function (opcode %u/%u)\n",
+			(unsigned int)rh->r.opcode,
+			(unsigned int)rh->r.subcode);
+		droq->stats.dropped_nodispatch++;
+	}
+
+	return cnt;
+}
+
+static inline void octeon_droq_drop_packets(struct octeon_device *oct,
+					    struct octeon_droq *droq,
+					    u32 cnt)
+{
+	u32 i = 0, buf_cnt;
+	struct octeon_droq_info *info;
+
+	for (i = 0; i < cnt; i++) {
+		info = (struct octeon_droq_info *)
+			droq->recv_buf_list[droq->read_idx].data;
+		octeon_swap_8B_data((u64 *)info, 2);
+
+		if (info->length) {
+			info->length += OCTNET_FRM_LENGTH_SIZE;
+			droq->stats.bytes_received += info->length;
+			buf_cnt = octeon_droq_get_bufcount(droq->buffer_size,
+							   (u32)info->length);
+		} else {
+			dev_err(&oct->pci_dev->dev, "DROQ: In drop: pkt with len 0\n");
+			buf_cnt = 1;
+		}
+
+		droq->read_idx = incr_index(droq->read_idx, buf_cnt,
+					    droq->max_count);
+		droq->refill_count += buf_cnt;
+	}
+}
+
+static u32
+octeon_droq_fast_process_packets(struct octeon_device *oct,
+				 struct octeon_droq *droq,
+				 u32 pkts_to_process)
+{
+	u32 pkt, total_len = 0, pkt_count, retval;
+	struct octeon_droq_info *info;
+	union octeon_rh *rh;
+
+	pkt_count = pkts_to_process;
+
+	for (pkt = 0; pkt < pkt_count; pkt++) {
+		u32 pkt_len = 0;
+		struct sk_buff *nicbuf = NULL;
+		struct octeon_skb_page_info *pg_info;
+		void *buf;
+
+		info = (struct octeon_droq_info *)
+			droq->recv_buf_list[droq->read_idx].data;
+		octeon_swap_8B_data((u64 *)info, 2);
+
+		if (!info->length) {
+			dev_err(&oct->pci_dev->dev,
+				"DROQ[%d] idx: %d len:0, pkt_cnt: %d\n",
+				droq->q_no, droq->read_idx, pkt_count);
+			print_hex_dump_bytes("", DUMP_PREFIX_ADDRESS,
+					     (u8 *)info,
+					     OCT_DROQ_INFO_SIZE);
+			break;
+		}
+
+		/* Len of resp hdr in included in the received data len. */
+		rh = &info->rh;
+
+		info->length += OCTNET_FRM_LENGTH_SIZE;
+		rh->r_dh.len += (ROUNDUP8(OCT_DROQ_INFO_SIZE) / sizeof(u64));
+		total_len += (u32)info->length;
+		if (opcode_slow_path(rh)) {
+			u32 buf_cnt;
+
+			buf_cnt = octeon_droq_dispatch_pkt(oct, droq, rh, info);
+			droq->read_idx = incr_index(droq->read_idx,
+						    buf_cnt, droq->max_count);
+			droq->refill_count += buf_cnt;
+		} else {
+			if (info->length <= droq->buffer_size) {
+				pkt_len = (u32)info->length;
+				nicbuf = droq->recv_buf_list[
+					droq->read_idx].buffer;
+				pg_info = &droq->recv_buf_list[
+					droq->read_idx].pg_info;
+				if (recv_buffer_recycle(oct, pg_info))
+					pg_info->page = NULL;
+				droq->recv_buf_list[droq->read_idx].buffer =
+					NULL;
+
+				droq->read_idx = incr_index(droq->read_idx, 1,
+							    droq->max_count);
+				droq->refill_count++;
+			} else {
+				nicbuf = octeon_fast_packet_alloc((u32)
+								  info->length);
+				pkt_len = 0;
+				/* nicbuf allocation can fail. We'll handle it
+				 * inside the loop.
+				 */
+				while (pkt_len < info->length) {
+					int cpy_len, idx = droq->read_idx;
+
+					cpy_len = ((pkt_len + droq->buffer_size)
+						   > info->length) ?
+						((u32)info->length - pkt_len) :
+						droq->buffer_size;
+
+					if (nicbuf) {
+						octeon_fast_packet_next(droq,
+									nicbuf,
+									cpy_len,
+									idx);
+						buf = droq->recv_buf_list[
+							idx].buffer;
+						recv_buffer_fast_free(buf);
+						droq->recv_buf_list[idx].buffer
+							= NULL;
+					} else {
+						droq->stats.rx_alloc_failure++;
+					}
+
+					pkt_len += cpy_len;
+					droq->read_idx =
+						incr_index(droq->read_idx, 1,
+							   droq->max_count);
+					droq->refill_count++;
+				}
+			}
+
+			if (nicbuf) {
+				if (droq->ops.fptr) {
+					droq->ops.fptr(oct->octeon_id,
+						       nicbuf, pkt_len,
+						       rh, &droq->napi,
+						       droq->ops.farg);
+				} else {
+					recv_buffer_free(nicbuf);
+				}
+			}
+		}
+
+		if (droq->refill_count >= droq->refill_threshold) {
+			int desc_refilled = octeon_droq_refill(oct, droq);
+
+			if (desc_refilled) {
+				/* Flush the droq descriptor data to memory to
+				 * be sure that when we update the credits the
+				 * data in memory is accurate.
+				 */
+				wmb();
+				writel(desc_refilled, droq->pkts_credit_reg);
+			}
+		}
+	}                       /* for (each packet)... */
+
+	/* Increment refill_count by the number of buffers processed. */
+	droq->stats.pkts_received += pkt;
+	droq->stats.bytes_received += total_len;
+
+	retval = pkt;
+	if ((droq->ops.drop_on_max) && (pkts_to_process - pkt)) {
+		octeon_droq_drop_packets(oct, droq, (pkts_to_process - pkt));
+
+		droq->stats.dropped_toomany += (pkts_to_process - pkt);
+		retval = pkts_to_process;
+	}
+
+	atomic_sub(retval, &droq->pkts_pending);
+
+	if (droq->refill_count >= droq->refill_threshold &&
+	    readl(droq->pkts_credit_reg) < CN23XX_SLI_DEF_BP) {
+		octeon_droq_check_hw_for_pkts(droq);
+
+		/* Make sure there are no pkts_pending */
+		if (!atomic_read(&droq->pkts_pending))
+			octeon_schedule_rxq_oom_work(oct, droq);
+	}
+
+	return retval;
+}
+
+int
+octeon_droq_process_packets(struct octeon_device *oct,
+			    struct octeon_droq *droq,
+			    u32 budget)
+{
+	u32 pkt_count = 0;
+	struct list_head *tmp, *tmp2;
+
+	octeon_droq_check_hw_for_pkts(droq);
+	pkt_count = atomic_read(&droq->pkts_pending);
+
+	if (!pkt_count)
+		return 0;
+
+	if (pkt_count > budget)
+		pkt_count = budget;
+
+	octeon_droq_fast_process_packets(oct, droq, pkt_count);
+
+	list_for_each_safe(tmp, tmp2, &droq->dispatch_list) {
+		struct __dispatch *rdisp = (struct __dispatch *)tmp;
+
+		list_del(tmp);
+		rdisp->disp_fn(rdisp->rinfo,
+			       octeon_get_dispatch_arg
+			       (oct,
+				(u16)rdisp->rinfo->recv_pkt->rh.r.opcode,
+				(u16)rdisp->rinfo->recv_pkt->rh.r.subcode));
+	}
+
+	/* If there are packets pending. schedule tasklet again */
+	if (atomic_read(&droq->pkts_pending))
+		return 1;
+
+	return 0;
+}
+
+/*
+ * Utility function to poll for packets. check_hw_for_packets must be
+ * called before calling this routine.
+ */
+
+int
+octeon_droq_process_poll_pkts(struct octeon_device *oct,
+			      struct octeon_droq *droq, u32 budget)
+{
+	struct list_head *tmp, *tmp2;
+	u32 pkts_available = 0, pkts_processed = 0;
+	u32 total_pkts_processed = 0;
+
+	if (budget > droq->max_count)
+		budget = droq->max_count;
+
+	while (total_pkts_processed < budget) {
+		octeon_droq_check_hw_for_pkts(droq);
+
+		pkts_available = min((budget - total_pkts_processed),
+				     (u32)(atomic_read(&droq->pkts_pending)));
+
+		if (pkts_available == 0)
+			break;
+
+		pkts_processed =
+			octeon_droq_fast_process_packets(oct, droq,
+							 pkts_available);
+
+		total_pkts_processed += pkts_processed;
+	}
+
+	list_for_each_safe(tmp, tmp2, &droq->dispatch_list) {
+		struct __dispatch *rdisp = (struct __dispatch *)tmp;
+
+		list_del(tmp);
+		rdisp->disp_fn(rdisp->rinfo,
+			       octeon_get_dispatch_arg
+			       (oct,
+				(u16)rdisp->rinfo->recv_pkt->rh.r.opcode,
+				(u16)rdisp->rinfo->recv_pkt->rh.r.subcode));
+	}
+
+	return total_pkts_processed;
+}
+
+/* Enable Pkt Interrupt */
+int
+octeon_enable_irq(struct octeon_device *oct, u32 q_no)
+{
+	switch (oct->chip_id) {
+	case OCTEON_CN66XX:
+	case OCTEON_CN68XX: {
+		struct octeon_cn6xxx *cn6xxx =
+			(struct octeon_cn6xxx *)oct->chip;
+		unsigned long flags;
+		u32 value;
+
+		spin_lock_irqsave
+			(&cn6xxx->lock_for_droq_int_enb_reg, flags);
+		value = octeon_read_csr(oct, CN6XXX_SLI_PKT_TIME_INT_ENB);
+		value |= (1 << q_no);
+		octeon_write_csr(oct, CN6XXX_SLI_PKT_TIME_INT_ENB, value);
+		value = octeon_read_csr(oct, CN6XXX_SLI_PKT_CNT_INT_ENB);
+		value |= (1 << q_no);
+		octeon_write_csr(oct, CN6XXX_SLI_PKT_CNT_INT_ENB, value);
+
+		/* don't bother flushing the enables */
+
+		spin_unlock_irqrestore
+			(&cn6xxx->lock_for_droq_int_enb_reg, flags);
+	}
+		break;
+	case OCTEON_CN23XX_PF_VID:
+		lio_enable_irq(oct->droq[q_no], oct->instr_queue[q_no]);
+		break;
+
+	case OCTEON_CN23XX_VF_VID:
+		lio_enable_irq(oct->droq[q_no], oct->instr_queue[q_no]);
+		break;
+	default:
+		dev_err(&oct->pci_dev->dev, "%s Unknown Chip\n", __func__);
+		return 1;
+	}
+
+	return 0;
+}
+
+int octeon_register_droq_ops(struct octeon_device *oct, u32 q_no,
+			     struct octeon_droq_ops *ops)
+{
+	struct octeon_config *oct_cfg = NULL;
+	struct octeon_droq *droq;
+
+	oct_cfg = octeon_get_conf(oct);
+
+	if (!oct_cfg)
+		return -EINVAL;
+
+	if (!(ops)) {
+		dev_err(&oct->pci_dev->dev, "%s: droq_ops pointer is NULL\n",
+			__func__);
+		return -EINVAL;
+	}
+
+	if (q_no >= CFG_GET_OQ_MAX_Q(oct_cfg)) {
+		dev_err(&oct->pci_dev->dev, "%s: droq id (%d) exceeds MAX (%d)\n",
+			__func__, q_no, (oct->num_oqs - 1));
+		return -EINVAL;
+	}
+
+	droq = oct->droq[q_no];
+	memcpy(&droq->ops, ops, sizeof(struct octeon_droq_ops));
+
+	return 0;
+}
+
+int octeon_unregister_droq_ops(struct octeon_device *oct, u32 q_no)
+{
+	struct octeon_config *oct_cfg = NULL;
+	struct octeon_droq *droq;
+
+	oct_cfg = octeon_get_conf(oct);
+
+	if (!oct_cfg)
+		return -EINVAL;
+
+	if (q_no >= CFG_GET_OQ_MAX_Q(oct_cfg)) {
+		dev_err(&oct->pci_dev->dev, "%s: droq id (%d) exceeds MAX (%d)\n",
+			__func__, q_no, oct->num_oqs - 1);
+		return -EINVAL;
+	}
+
+	droq = oct->droq[q_no];
+
+	if (!droq) {
+		dev_info(&oct->pci_dev->dev,
+			 "Droq id (%d) not available.\n", q_no);
+		return 0;
+	}
+
+	droq->ops.fptr = NULL;
+	droq->ops.farg = NULL;
+	droq->ops.drop_on_max = 0;
+
+	return 0;
+}
+
+int octeon_create_droq(struct octeon_device *oct,
+		       u32 q_no, u32 num_descs,
+		       u32 desc_size, void *app_ctx)
+{
+	struct octeon_droq *droq;
+	int numa_node = dev_to_node(&oct->pci_dev->dev);
+
+	if (oct->droq[q_no]) {
+		dev_dbg(&oct->pci_dev->dev, "Droq already in use. Cannot create droq %d again\n",
+			q_no);
+		return 1;
+	}
+
+	/* Allocate the DS for the new droq. */
+	droq = vmalloc_node(sizeof(*droq), numa_node);
+	if (!droq)
+		droq = vmalloc(sizeof(*droq));
+	if (!droq)
+		return -1;
+
+	memset(droq, 0, sizeof(struct octeon_droq));
+
+	/*Disable the pkt o/p for this Q  */
+	octeon_set_droq_pkt_op(oct, q_no, 0);
+	oct->droq[q_no] = droq;
+
+	/* Initialize the Droq */
+	if (octeon_init_droq(oct, q_no, num_descs, desc_size, app_ctx)) {
+		vfree(oct->droq[q_no]);
+		oct->droq[q_no] = NULL;
+		return -1;
+	}
+
+	oct->num_oqs++;
+
+	dev_dbg(&oct->pci_dev->dev, "%s: Total number of OQ: %d\n", __func__,
+		oct->num_oqs);
+
+	/* Global Droq register settings */
+
+	/* As of now not required, as setting are done for all 32 Droqs at
+	 * the same time.
+	 */
+	return 0;
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_droq.h b/drivers/net/ethernet/cavium/liquidio/octeon_droq.h
new file mode 100644
index 000000000..c9b19e624
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_droq.h
@@ -0,0 +1,416 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*!  \file  octeon_droq.h
+ *   \brief Implementation of Octeon Output queues. "Output" is with
+ *   respect to the Octeon device on the NIC. From this driver's point of
+ *   view they are ingress queues.
+ */
+
+#ifndef __OCTEON_DROQ_H__
+#define __OCTEON_DROQ_H__
+
+/* Default number of packets that will be processed in one iteration. */
+#define MAX_PACKET_BUDGET 0xFFFFFFFF
+
+/** Octeon descriptor format.
+ *  The descriptor ring is made of descriptors which have 2 64-bit values:
+ *  -# Physical (bus) address of the data buffer.
+ *  -# Physical (bus) address of a octeon_droq_info structure.
+ *  The Octeon device DMA's incoming packets and its information at the address
+ *  given by these descriptor fields.
+ */
+struct octeon_droq_desc {
+	/** The buffer pointer */
+	u64 buffer_ptr;
+
+	/** The Info pointer */
+	u64 info_ptr;
+};
+
+#define OCT_DROQ_DESC_SIZE    (sizeof(struct octeon_droq_desc))
+
+/** Information about packet DMA'ed by Octeon.
+ *  The format of the information available at Info Pointer after Octeon
+ *  has posted a packet. Not all descriptors have valid information. Only
+ *  the Info field of the first descriptor for a packet has information
+ *  about the packet.
+ */
+struct octeon_droq_info {
+	/** The Length of the packet. */
+	u64 length;
+
+	/** The Output Receive Header. */
+	union octeon_rh rh;
+};
+
+#define OCT_DROQ_INFO_SIZE   (sizeof(struct octeon_droq_info))
+
+struct octeon_skb_page_info {
+	/* DMA address for the page */
+	dma_addr_t dma;
+
+	/* Page for the rx dma  **/
+	struct page *page;
+
+	/** which offset into page */
+	unsigned int page_offset;
+};
+
+/** Pointer to data buffer.
+ *  Driver keeps a pointer to the data buffer that it made available to
+ *  the Octeon device. Since the descriptor ring keeps physical (bus)
+ *  addresses, this field is required for the driver to keep track of
+ *  the virtual address pointers.
+ */
+struct octeon_recv_buffer {
+	/** Packet buffer, including metadata. */
+	void *buffer;
+
+	/** Data in the packet buffer.  */
+	u8 *data;
+
+	/** pg_info **/
+	struct octeon_skb_page_info pg_info;
+};
+
+#define OCT_DROQ_RECVBUF_SIZE    (sizeof(struct octeon_recv_buffer))
+
+/** Output Queue statistics. Each output queue has four stats fields. */
+struct oct_droq_stats {
+	/** Number of packets received in this queue. */
+	u64 pkts_received;
+
+	/** Bytes received by this queue. */
+	u64 bytes_received;
+
+	/** Packets dropped due to no dispatch function. */
+	u64 dropped_nodispatch;
+
+	/** Packets dropped due to no memory available. */
+	u64 dropped_nomem;
+
+	/** Packets dropped due to large number of pkts to process. */
+	u64 dropped_toomany;
+
+	/** Number of packets  sent to stack from this queue. */
+	u64 rx_pkts_received;
+
+	/** Number of Bytes sent to stack from this queue. */
+	u64 rx_bytes_received;
+
+	/** Num of Packets dropped due to receive path failures. */
+	u64 rx_dropped;
+
+	u64 rx_vxlan;
+
+	/** Num of failures of recv_buffer_alloc() */
+	u64 rx_alloc_failure;
+
+};
+
+/* The maximum number of buffers that can be dispatched from the
+ * output/dma queue. Set to 64 assuming 1K buffers in DROQ and the fact that
+ * max packet size from DROQ is 64K.
+ */
+#define    MAX_RECV_BUFS    64
+
+/** Receive Packet format used when dispatching output queue packets
+ *  with non-raw opcodes.
+ *  The received packet will be sent to the upper layers using this
+ *  structure which is passed as a parameter to the dispatch function
+ */
+struct octeon_recv_pkt {
+	/**  Number of buffers in this received packet */
+	u16 buffer_count;
+
+	/** Id of the device that is sending the packet up */
+	u16 octeon_id;
+
+	/** Length of data in the packet buffer */
+	u32 length;
+
+	/** The receive header */
+	union octeon_rh rh;
+
+	/** Pointer to the OS-specific packet buffer */
+	void *buffer_ptr[MAX_RECV_BUFS];
+
+	/** Size of the buffers pointed to by ptr's in buffer_ptr */
+	u32 buffer_size[MAX_RECV_BUFS];
+};
+
+#define OCT_RECV_PKT_SIZE    (sizeof(struct octeon_recv_pkt))
+
+/** The first parameter of a dispatch function.
+ *  For a raw mode opcode, the driver dispatches with the device
+ *  pointer in this structure.
+ *  For non-raw mode opcode, the driver dispatches the recv_pkt
+ *  created to contain the buffers with data received from Octeon.
+ *  ---------------------
+ *  |     *recv_pkt ----|---
+ *  |-------------------|   |
+ *  | 0 or more bytes   |   |
+ *  | reserved by driver|   |
+ *  |-------------------|<-/
+ *  | octeon_recv_pkt   |
+ *  |                   |
+ *  |___________________|
+ */
+struct octeon_recv_info {
+	void *rsvd;
+	struct octeon_recv_pkt *recv_pkt;
+};
+
+#define  OCT_RECV_INFO_SIZE    (sizeof(struct octeon_recv_info))
+
+/** Allocate a recv_info structure. The recv_pkt pointer in the recv_info
+ *  structure is filled in before this call returns.
+ *  @param extra_bytes - extra bytes to be allocated at the end of the recv info
+ *                       structure.
+ *  @return - pointer to a newly allocated recv_info structure.
+ */
+static inline struct octeon_recv_info *octeon_alloc_recv_info(int extra_bytes)
+{
+	struct octeon_recv_info *recv_info;
+	u8 *buf;
+
+	buf = kmalloc(OCT_RECV_PKT_SIZE + OCT_RECV_INFO_SIZE +
+		      extra_bytes, GFP_ATOMIC);
+	if (!buf)
+		return NULL;
+
+	recv_info = (struct octeon_recv_info *)buf;
+	recv_info->recv_pkt =
+		(struct octeon_recv_pkt *)(buf + OCT_RECV_INFO_SIZE);
+	recv_info->rsvd = NULL;
+	if (extra_bytes)
+		recv_info->rsvd = buf + OCT_RECV_INFO_SIZE + OCT_RECV_PKT_SIZE;
+
+	return recv_info;
+}
+
+/** Free a recv_info structure.
+ *  @param recv_info - Pointer to receive_info to be freed
+ */
+static inline void octeon_free_recv_info(struct octeon_recv_info *recv_info)
+{
+	kfree(recv_info);
+}
+
+typedef int (*octeon_dispatch_fn_t)(struct octeon_recv_info *, void *);
+
+/** Used by NIC module to register packet handler and to get device
+ * information for each octeon device.
+ */
+struct octeon_droq_ops {
+	/** This registered function will be called by the driver with
+	 *  the octeon id, pointer to buffer from droq and length of
+	 *  data in the buffer. The receive header gives the port
+	 *  number to the caller.  Function pointer is set by caller.
+	 */
+	void (*fptr)(u32, void *, u32, union octeon_rh *, void *, void *);
+	void *farg;
+
+	/* This function will be called by the driver for all NAPI related
+	 * events. The first param is the octeon id. The second param is the
+	 * output queue number. The third is the NAPI event that occurred.
+	 */
+	void (*napi_fn)(void *);
+
+	u32 poll_mode;
+
+	/** Flag indicating if the DROQ handler should drop packets that
+	 *  it cannot handle in one iteration. Set by caller.
+	 */
+	u32 drop_on_max;
+};
+
+/** The Descriptor Ring Output Queue structure.
+ *  This structure has all the information required to implement a
+ *  Octeon DROQ.
+ */
+struct octeon_droq {
+	u32 q_no;
+
+	u32 pkt_count;
+
+	struct octeon_droq_ops ops;
+
+	struct octeon_device *oct_dev;
+
+	/** The 8B aligned descriptor ring starts at this address. */
+	struct octeon_droq_desc *desc_ring;
+
+	/** Index in the ring where the driver should read the next packet */
+	u32 read_idx;
+
+	/** Index in the ring where Octeon will write the next packet */
+	u32 write_idx;
+
+	/** Index in the ring where the driver will refill the descriptor's
+	 * buffer
+	 */
+	u32 refill_idx;
+
+	/** Packets pending to be processed */
+	atomic_t pkts_pending;
+
+	/** Number of  descriptors in this ring. */
+	u32 max_count;
+
+	/** The number of descriptors pending refill. */
+	u32 refill_count;
+
+	u32 pkts_per_intr;
+	u32 refill_threshold;
+
+	/** The max number of descriptors in DROQ without a buffer.
+	 * This field is used to keep track of empty space threshold. If the
+	 * refill_count reaches this value, the DROQ cannot accept a max-sized
+	 * (64K) packet.
+	 */
+	u32 max_empty_descs;
+
+	/** The receive buffer list. This list has the virtual addresses of the
+	 * buffers.
+	 */
+	struct octeon_recv_buffer *recv_buf_list;
+
+	/** The size of each buffer pointed by the buffer pointer. */
+	u32 buffer_size;
+
+	/** Pointer to the mapped packet credit register.
+	 * Host writes number of info/buffer ptrs available to this register
+	 */
+	void  __iomem *pkts_credit_reg;
+
+	/** Pointer to the mapped packet sent register.
+	 * Octeon writes the number of packets DMA'ed to host memory
+	 * in this register.
+	 */
+	void __iomem *pkts_sent_reg;
+
+	struct list_head dispatch_list;
+
+	/** Statistics for this DROQ. */
+	struct oct_droq_stats stats;
+
+	/** DMA mapped address of the DROQ descriptor ring. */
+	size_t desc_ring_dma;
+
+	/** application context */
+	void *app_ctx;
+
+	struct napi_struct napi;
+
+	u32 cpu_id;
+
+	call_single_data_t csd;
+};
+
+#define OCT_DROQ_SIZE   (sizeof(struct octeon_droq))
+
+/**
+ *  Allocates space for the descriptor ring for the droq and sets the
+ *   base addr, num desc etc in Octeon registers.
+ *
+ * @param  oct_dev    - pointer to the octeon device structure
+ * @param  q_no       - droq no. ranges from 0 - 3.
+ * @param app_ctx     - pointer to application context
+ * @return Success: 0    Failure: 1
+ */
+int octeon_init_droq(struct octeon_device *oct_dev,
+		     u32 q_no,
+		     u32 num_descs,
+		     u32 desc_size,
+		     void *app_ctx);
+
+/**
+ *  Frees the space for descriptor ring for the droq.
+ *
+ *  @param oct_dev - pointer to the octeon device structure
+ *  @param q_no    - droq no. ranges from 0 - 3.
+ *  @return:    Success: 0    Failure: 1
+ */
+int octeon_delete_droq(struct octeon_device *oct_dev, u32 q_no);
+
+/** Register a change in droq operations. The ops field has a pointer to a
+ * function which will called by the DROQ handler for all packets arriving
+ * on output queues given by q_no irrespective of the type of packet.
+ * The ops field also has a flag which if set tells the DROQ handler to
+ * drop packets if it receives more than what it can process in one
+ * invocation of the handler.
+ * @param oct       - octeon device
+ * @param q_no      - octeon output queue number (0 <= q_no <= MAX_OCTEON_DROQ-1
+ * @param ops       - the droq_ops settings for this queue
+ * @return          - 0 on success, -ENODEV or -EINVAL on error.
+ */
+int
+octeon_register_droq_ops(struct octeon_device *oct,
+			 u32 q_no,
+			 struct octeon_droq_ops *ops);
+
+/** Resets the function pointer and flag settings made by
+ * octeon_register_droq_ops(). After this routine is called, the DROQ handler
+ * will lookup dispatch function for each arriving packet on the output queue
+ * given by q_no.
+ * @param oct       - octeon device
+ * @param q_no      - octeon output queue number (0 <= q_no <= MAX_OCTEON_DROQ-1
+ * @return          - 0 on success, -ENODEV or -EINVAL on error.
+ */
+int octeon_unregister_droq_ops(struct octeon_device *oct, u32 q_no);
+
+/**   Register a dispatch function for a opcode/subcode. The driver will call
+ *    this dispatch function when it receives a packet with the given
+ *    opcode/subcode in its output queues along with the user specified
+ *    argument.
+ *    @param  oct        - the octeon device to register with.
+ *    @param  opcode     - the opcode for which the dispatch will be registered.
+ *    @param  subcode    - the subcode for which the dispatch will be registered
+ *    @param  fn         - the dispatch function.
+ *    @param  fn_arg     - user specified that will be passed along with the
+ *                         dispatch function by the driver.
+ *    @return Success: 0; Failure: 1
+ */
+int octeon_register_dispatch_fn(struct octeon_device *oct,
+				u16 opcode,
+				u16 subcode,
+				octeon_dispatch_fn_t fn, void *fn_arg);
+
+void *octeon_get_dispatch_arg(struct octeon_device *oct,
+			      u16 opcode, u16 subcode);
+
+void octeon_droq_print_stats(void);
+
+u32 octeon_droq_check_hw_for_pkts(struct octeon_droq *droq);
+
+int octeon_create_droq(struct octeon_device *oct, u32 q_no,
+		       u32 num_descs, u32 desc_size, void *app_ctx);
+
+int octeon_droq_process_packets(struct octeon_device *oct,
+				struct octeon_droq *droq,
+				u32 budget);
+
+int octeon_droq_process_poll_pkts(struct octeon_device *oct,
+				  struct octeon_droq *droq, u32 budget);
+
+int octeon_enable_irq(struct octeon_device *oct, u32 q_no);
+
+int octeon_retry_droq_refill(struct octeon_droq *droq);
+
+#endif	/*__OCTEON_DROQ_H__ */
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_iq.h b/drivers/net/ethernet/cavium/liquidio/octeon_iq.h
new file mode 100644
index 000000000..bebf3bd34
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_iq.h
@@ -0,0 +1,399 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*!  \file  octeon_iq.h
+ *   \brief Host Driver: Implementation of Octeon input queues. "Input" is
+ *   with respect to the Octeon device on the NIC. From this driver's
+ *   point of view they are egress queues.
+ */
+
+#ifndef __OCTEON_IQ_H__
+#define  __OCTEON_IQ_H__
+
+#define IQ_STATUS_RUNNING   1
+
+#define IQ_SEND_OK          0
+#define IQ_SEND_STOP        1
+#define IQ_SEND_FAILED     -1
+
+/*-------------------------  INSTRUCTION QUEUE --------------------------*/
+
+/* \cond */
+
+#define REQTYPE_NONE                 0
+#define REQTYPE_NORESP_NET           1
+#define REQTYPE_NORESP_NET_SG        2
+#define REQTYPE_RESP_NET             3
+#define REQTYPE_RESP_NET_SG          4
+#define REQTYPE_SOFT_COMMAND         5
+#define REQTYPE_LAST                 5
+
+struct octeon_request_list {
+	u32 reqtype;
+	void *buf;
+};
+
+/* \endcond */
+
+/** Input Queue statistics. Each input queue has four stats fields. */
+struct oct_iq_stats {
+	u64 instr_posted; /**< Instructions posted to this queue. */
+	u64 instr_processed; /**< Instructions processed in this queue. */
+	u64 instr_dropped; /**< Instructions that could not be processed */
+	u64 bytes_sent;  /**< Bytes sent through this queue. */
+	u64 sgentry_sent;/**< Gather entries sent through this queue. */
+	u64 tx_done;/**< Num of packets sent to network. */
+	u64 tx_iq_busy;/**< Numof times this iq was found to be full. */
+	u64 tx_dropped;/**< Numof pkts dropped dueto xmitpath errors. */
+	u64 tx_tot_bytes;/**< Total count of bytes sento to network. */
+	u64 tx_gso;  /* count of tso */
+	u64 tx_vxlan; /* tunnel */
+	u64 tx_dmamap_fail; /* Number of times dma mapping failed */
+	u64 tx_restart; /* Number of times this queue restarted */
+};
+
+#define OCT_IQ_STATS_SIZE   (sizeof(struct oct_iq_stats))
+
+/** The instruction (input) queue.
+ *  The input queue is used to post raw (instruction) mode data or packet
+ *  data to Octeon device from the host. Each input queue (upto 4) for
+ *  a Octeon device has one such structure to represent it.
+ */
+struct octeon_instr_queue {
+	struct octeon_device *oct_dev;
+
+	/** A spinlock to protect access to the input ring.  */
+	spinlock_t lock;
+
+	/** A spinlock to protect while posting on the ring.  */
+	spinlock_t post_lock;
+
+	/** This flag indicates if the queue can be used for soft commands.
+	 *  If this flag is set, post_lock must be acquired before posting
+	 *  a command to the queue.
+	 *  If this flag is clear, post_lock is invalid for the queue.
+	 *  All control commands (soft commands) will go through only Queue 0
+	 *  (control and data queue). So only queue-0 needs post_lock,
+	 *  other queues are only data queues and does not need post_lock
+	 */
+	bool allow_soft_cmds;
+
+	u32 pkt_in_done;
+
+	u32 pkts_processed;
+
+	/** A spinlock to protect access to the input ring.*/
+	spinlock_t iq_flush_running_lock;
+
+	/** Flag that indicates if the queue uses 64 byte commands. */
+	u32 iqcmd_64B:1;
+
+	/** Queue info. */
+	union oct_txpciq txpciq;
+
+	u32 rsvd:17;
+
+	/* Controls whether extra flushing of IQ is done on Tx */
+	u32 do_auto_flush:1;
+
+	u32 status:8;
+
+	/** Maximum no. of instructions in this queue. */
+	u32 max_count;
+
+	/** Index in input ring where the driver should write the next packet */
+	u32 host_write_index;
+
+	/** Index in input ring where Octeon is expected to read the next
+	 * packet.
+	 */
+	u32 octeon_read_index;
+
+	/** This index aids in finding the window in the queue where Octeon
+	 *  has read the commands.
+	 */
+	u32 flush_index;
+
+	/** This field keeps track of the instructions pending in this queue. */
+	atomic_t instr_pending;
+
+	u32 reset_instr_cnt;
+
+	/** Pointer to the Virtual Base addr of the input ring. */
+	u8 *base_addr;
+
+	struct octeon_request_list *request_list;
+
+	/** Octeon doorbell register for the ring. */
+	void __iomem *doorbell_reg;
+
+	/** Octeon instruction count register for this ring. */
+	void __iomem *inst_cnt_reg;
+
+	/** Number of instructions pending to be posted to Octeon. */
+	u32 fill_cnt;
+
+	/** The max. number of instructions that can be held pending by the
+	 * driver.
+	 */
+	u32 fill_threshold;
+
+	/** The last time that the doorbell was rung. */
+	u64 last_db_time;
+
+	/** The doorbell timeout. If the doorbell was not rung for this time and
+	 * fill_cnt is non-zero, ring the doorbell again.
+	 */
+	u32 db_timeout;
+
+	/** Statistics for this input queue. */
+	struct oct_iq_stats stats;
+
+	/** DMA mapped base address of the input descriptor ring. */
+	dma_addr_t base_addr_dma;
+
+	/** Application context */
+	void *app_ctx;
+
+	/* network stack queue index */
+	int q_index;
+
+	/*os ifidx associated with this queue */
+	int ifidx;
+
+};
+
+/*----------------------  INSTRUCTION FORMAT ----------------------------*/
+
+/** 32-byte instruction format.
+ *  Format of instruction for a 32-byte mode input queue.
+ */
+struct octeon_instr_32B {
+	/** Pointer where the input data is available. */
+	u64 dptr;
+
+	/** Instruction Header.  */
+	u64 ih;
+
+	/** Pointer where the response for a RAW mode packet will be written
+	 * by Octeon.
+	 */
+	u64 rptr;
+
+	/** Input Request Header. Additional info about the input. */
+	u64 irh;
+
+};
+
+#define OCT_32B_INSTR_SIZE     (sizeof(struct octeon_instr_32B))
+
+/** 64-byte instruction format.
+ *  Format of instruction for a 64-byte mode input queue.
+ */
+struct octeon_instr2_64B {
+	/** Pointer where the input data is available. */
+	u64 dptr;
+
+	/** Instruction Header. */
+	u64 ih2;
+
+	/** Input Request Header. */
+	u64 irh;
+
+	/** opcode/subcode specific parameters */
+	u64 ossp[2];
+
+	/** Return Data Parameters */
+	u64 rdp;
+
+	/** Pointer where the response for a RAW mode packet will be written
+	 * by Octeon.
+	 */
+	u64 rptr;
+
+	u64 reserved;
+};
+
+struct octeon_instr3_64B {
+	/** Pointer where the input data is available. */
+	u64 dptr;
+
+	/** Instruction Header. */
+	u64 ih3;
+
+	/** Instruction Header. */
+	u64 pki_ih3;
+
+	/** Input Request Header. */
+	u64 irh;
+
+	/** opcode/subcode specific parameters */
+	u64 ossp[2];
+
+	/** Return Data Parameters */
+	u64 rdp;
+
+	/** Pointer where the response for a RAW mode packet will be written
+	 * by Octeon.
+	 */
+	u64 rptr;
+
+};
+
+union octeon_instr_64B {
+	struct octeon_instr2_64B cmd2;
+	struct octeon_instr3_64B cmd3;
+};
+
+#define OCT_64B_INSTR_SIZE     (sizeof(union octeon_instr_64B))
+
+/** The size of each buffer in soft command buffer pool
+ */
+#define  SOFT_COMMAND_BUFFER_SIZE	2048
+
+struct octeon_soft_command {
+	/** Soft command buffer info. */
+	struct list_head node;
+	u64 dma_addr;
+	u32 size;
+
+	/** Command and return status */
+	union octeon_instr_64B cmd;
+
+#define COMPLETION_WORD_INIT    0xffffffffffffffffULL
+	u64 *status_word;
+
+	/** Data buffer info */
+	void *virtdptr;
+	u64 dmadptr;
+	u32 datasize;
+
+	/** Return buffer info */
+	void *virtrptr;
+	u64 dmarptr;
+	u32 rdatasize;
+
+	/** Context buffer info */
+	void *ctxptr;
+	u32  ctxsize;
+
+	/** Time out and callback */
+	size_t expiry_time;
+	u32 iq_no;
+	void (*callback)(struct octeon_device *, u32, void *);
+	void *callback_arg;
+
+	int caller_is_done;
+	u32 sc_status;
+	struct completion complete;
+};
+
+/* max timeout (in milli sec) for soft request */
+#define LIO_SC_MAX_TMO_MS       60000
+
+/** Maximum number of buffers to allocate into soft command buffer pool
+ */
+#define  MAX_SOFT_COMMAND_BUFFERS	256
+
+/** Head of a soft command buffer pool.
+ */
+struct octeon_sc_buffer_pool {
+	/** List structure to add delete pending entries to */
+	struct list_head head;
+
+	/** A lock for this response list */
+	spinlock_t lock;
+
+	atomic_t alloc_buf_count;
+};
+
+#define INCR_INSTRQUEUE_PKT_COUNT(octeon_dev_ptr, iq_no, field, count)  \
+		(((octeon_dev_ptr)->instr_queue[iq_no]->stats.field) += count)
+
+int octeon_setup_sc_buffer_pool(struct octeon_device *oct);
+int octeon_free_sc_done_list(struct octeon_device *oct);
+int octeon_free_sc_zombie_list(struct octeon_device *oct);
+int octeon_free_sc_buffer_pool(struct octeon_device *oct);
+struct octeon_soft_command *
+	octeon_alloc_soft_command(struct octeon_device *oct,
+				  u32 datasize, u32 rdatasize,
+				  u32 ctxsize);
+void octeon_free_soft_command(struct octeon_device *oct,
+			      struct octeon_soft_command *sc);
+
+/**
+ *  octeon_init_instr_queue()
+ *  @param octeon_dev      - pointer to the octeon device structure.
+ *  @param txpciq          - queue to be initialized (0 <= q_no <= 3).
+ *
+ *  Called at driver init time for each input queue. iq_conf has the
+ *  configuration parameters for the queue.
+ *
+ *  @return  Success: 0   Failure: 1
+ */
+int octeon_init_instr_queue(struct octeon_device *octeon_dev,
+			    union oct_txpciq txpciq,
+			    u32 num_descs);
+
+/**
+ *  octeon_delete_instr_queue()
+ *  @param octeon_dev      - pointer to the octeon device structure.
+ *  @param iq_no           - queue to be deleted (0 <= q_no <= 3).
+ *
+ *  Called at driver unload time for each input queue. Deletes all
+ *  allocated resources for the input queue.
+ *
+ *  @return  Success: 0   Failure: 1
+ */
+int octeon_delete_instr_queue(struct octeon_device *octeon_dev, u32 iq_no);
+
+int lio_wait_for_instr_fetch(struct octeon_device *oct);
+
+void
+octeon_ring_doorbell_locked(struct octeon_device *oct, u32 iq_no);
+
+int
+octeon_register_reqtype_free_fn(struct octeon_device *oct, int reqtype,
+				void (*fn)(void *));
+
+int
+lio_process_iq_request_list(struct octeon_device *oct,
+			    struct octeon_instr_queue *iq, u32 napi_budget);
+
+int octeon_send_command(struct octeon_device *oct, u32 iq_no,
+			u32 force_db, void *cmd, void *buf,
+			u32 datasize, u32 reqtype);
+
+void octeon_dump_soft_command(struct octeon_device *oct,
+			      struct octeon_soft_command *sc);
+
+void octeon_prepare_soft_command(struct octeon_device *oct,
+				 struct octeon_soft_command *sc,
+				 u8 opcode, u8 subcode,
+				 u32 irh_ossp, u64 ossp0,
+				 u64 ossp1);
+
+int octeon_send_soft_command(struct octeon_device *oct,
+			     struct octeon_soft_command *sc);
+
+int octeon_setup_iq(struct octeon_device *oct, int ifidx,
+		    int q_index, union oct_txpciq iq_no, u32 num_descs,
+		    void *app_ctx);
+int
+octeon_flush_iq(struct octeon_device *oct, struct octeon_instr_queue *iq,
+		u32 napi_budget);
+#endif				/* __OCTEON_IQ_H__ */
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_mailbox.c b/drivers/net/ethernet/cavium/liquidio/octeon_mailbox.c
new file mode 100644
index 000000000..ad685f5d0
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_mailbox.c
@@ -0,0 +1,375 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_main.h"
+#include "octeon_mailbox.h"
+#include "cn23xx_pf_device.h"
+
+/**
+ * octeon_mbox_read:
+ * @mbox: Pointer mailbox
+ *
+ * Reads the 8-bytes of data from the mbox register
+ * Writes back the acknowldgement inidcating completion of read
+ */
+int octeon_mbox_read(struct octeon_mbox *mbox)
+{
+	union octeon_mbox_message msg;
+	int ret = 0;
+
+	spin_lock(&mbox->lock);
+
+	msg.u64 = readq(mbox->mbox_read_reg);
+
+	if ((msg.u64 == OCTEON_PFVFACK) || (msg.u64 == OCTEON_PFVFSIG)) {
+		spin_unlock(&mbox->lock);
+		return 0;
+	}
+
+	if (mbox->state & OCTEON_MBOX_STATE_REQUEST_RECEIVING) {
+		mbox->mbox_req.data[mbox->mbox_req.recv_len - 1] = msg.u64;
+		mbox->mbox_req.recv_len++;
+	} else {
+		if (mbox->state & OCTEON_MBOX_STATE_RESPONSE_RECEIVING) {
+			mbox->mbox_resp.data[mbox->mbox_resp.recv_len - 1] =
+				msg.u64;
+			mbox->mbox_resp.recv_len++;
+		} else {
+			if ((mbox->state & OCTEON_MBOX_STATE_IDLE) &&
+			    (msg.s.type == OCTEON_MBOX_REQUEST)) {
+				mbox->state &= ~OCTEON_MBOX_STATE_IDLE;
+				mbox->state |=
+				    OCTEON_MBOX_STATE_REQUEST_RECEIVING;
+				mbox->mbox_req.msg.u64 = msg.u64;
+				mbox->mbox_req.q_no = mbox->q_no;
+				mbox->mbox_req.recv_len = 1;
+			} else {
+				if ((mbox->state &
+				     OCTEON_MBOX_STATE_RESPONSE_PENDING) &&
+				    (msg.s.type == OCTEON_MBOX_RESPONSE)) {
+					mbox->state &=
+					    ~OCTEON_MBOX_STATE_RESPONSE_PENDING;
+					mbox->state |=
+					    OCTEON_MBOX_STATE_RESPONSE_RECEIVING
+					    ;
+					mbox->mbox_resp.msg.u64 = msg.u64;
+					mbox->mbox_resp.q_no = mbox->q_no;
+					mbox->mbox_resp.recv_len = 1;
+				} else {
+					writeq(OCTEON_PFVFERR,
+					       mbox->mbox_read_reg);
+					mbox->state |= OCTEON_MBOX_STATE_ERROR;
+					spin_unlock(&mbox->lock);
+					return 1;
+				}
+			}
+		}
+	}
+
+	if (mbox->state & OCTEON_MBOX_STATE_REQUEST_RECEIVING) {
+		if (mbox->mbox_req.recv_len < mbox->mbox_req.msg.s.len) {
+			ret = 0;
+		} else {
+			mbox->state &= ~OCTEON_MBOX_STATE_REQUEST_RECEIVING;
+			mbox->state |= OCTEON_MBOX_STATE_REQUEST_RECEIVED;
+			ret = 1;
+		}
+	} else {
+		if (mbox->state & OCTEON_MBOX_STATE_RESPONSE_RECEIVING) {
+			if (mbox->mbox_resp.recv_len <
+			    mbox->mbox_resp.msg.s.len) {
+				ret = 0;
+			} else {
+				mbox->state &=
+				    ~OCTEON_MBOX_STATE_RESPONSE_RECEIVING;
+				mbox->state |=
+				    OCTEON_MBOX_STATE_RESPONSE_RECEIVED;
+				ret = 1;
+			}
+		} else {
+			WARN_ON(1);
+		}
+	}
+
+	writeq(OCTEON_PFVFACK, mbox->mbox_read_reg);
+
+	spin_unlock(&mbox->lock);
+
+	return ret;
+}
+
+/**
+ * octeon_mbox_write:
+ * @oct: Pointer Octeon Device
+ * @mbox_cmd: Cmd to send to mailbox.
+ *
+ * Populates the queue specific mbox structure
+ * with cmd information.
+ * Write the cmd to mbox register
+ */
+int octeon_mbox_write(struct octeon_device *oct,
+		      struct octeon_mbox_cmd *mbox_cmd)
+{
+	struct octeon_mbox *mbox = oct->mbox[mbox_cmd->q_no];
+	u32 count, i, ret = OCTEON_MBOX_STATUS_SUCCESS;
+	long timeout = LIO_MBOX_WRITE_WAIT_TIME;
+	unsigned long flags;
+
+	spin_lock_irqsave(&mbox->lock, flags);
+
+	if ((mbox_cmd->msg.s.type == OCTEON_MBOX_RESPONSE) &&
+	    !(mbox->state & OCTEON_MBOX_STATE_REQUEST_RECEIVED)) {
+		spin_unlock_irqrestore(&mbox->lock, flags);
+		return OCTEON_MBOX_STATUS_FAILED;
+	}
+
+	if ((mbox_cmd->msg.s.type == OCTEON_MBOX_REQUEST) &&
+	    !(mbox->state & OCTEON_MBOX_STATE_IDLE)) {
+		spin_unlock_irqrestore(&mbox->lock, flags);
+		return OCTEON_MBOX_STATUS_BUSY;
+	}
+
+	if (mbox_cmd->msg.s.type == OCTEON_MBOX_REQUEST) {
+		memcpy(&mbox->mbox_resp, mbox_cmd,
+		       sizeof(struct octeon_mbox_cmd));
+		mbox->state = OCTEON_MBOX_STATE_RESPONSE_PENDING;
+	}
+
+	spin_unlock_irqrestore(&mbox->lock, flags);
+
+	count = 0;
+
+	while (readq(mbox->mbox_write_reg) != OCTEON_PFVFSIG) {
+		schedule_timeout_uninterruptible(timeout);
+		if (count++ == LIO_MBOX_WRITE_WAIT_CNT) {
+			ret = OCTEON_MBOX_STATUS_FAILED;
+			break;
+		}
+	}
+
+	if (ret == OCTEON_MBOX_STATUS_SUCCESS) {
+		writeq(mbox_cmd->msg.u64, mbox->mbox_write_reg);
+		for (i = 0; i < (u32)(mbox_cmd->msg.s.len - 1); i++) {
+			count = 0;
+			while (readq(mbox->mbox_write_reg) !=
+			       OCTEON_PFVFACK) {
+				schedule_timeout_uninterruptible(timeout);
+				if (count++ == LIO_MBOX_WRITE_WAIT_CNT) {
+					ret = OCTEON_MBOX_STATUS_FAILED;
+					break;
+				}
+			}
+			if (ret == OCTEON_MBOX_STATUS_SUCCESS)
+				writeq(mbox_cmd->data[i], mbox->mbox_write_reg);
+			else
+				break;
+		}
+	}
+
+	spin_lock_irqsave(&mbox->lock, flags);
+	if (mbox_cmd->msg.s.type == OCTEON_MBOX_RESPONSE) {
+		mbox->state = OCTEON_MBOX_STATE_IDLE;
+		writeq(OCTEON_PFVFSIG, mbox->mbox_read_reg);
+	} else {
+		if ((!mbox_cmd->msg.s.resp_needed) ||
+		    (ret == OCTEON_MBOX_STATUS_FAILED)) {
+			mbox->state &= ~OCTEON_MBOX_STATE_RESPONSE_PENDING;
+			if (!(mbox->state &
+			      (OCTEON_MBOX_STATE_REQUEST_RECEIVING |
+			       OCTEON_MBOX_STATE_REQUEST_RECEIVED)))
+				mbox->state = OCTEON_MBOX_STATE_IDLE;
+		}
+	}
+	spin_unlock_irqrestore(&mbox->lock, flags);
+
+	return ret;
+}
+
+static void get_vf_stats(struct octeon_device *oct,
+			 struct oct_vf_stats *stats)
+{
+	int i;
+
+	for (i = 0; i < oct->num_iqs; i++) {
+		if (!oct->instr_queue[i])
+			continue;
+		stats->tx_packets += oct->instr_queue[i]->stats.tx_done;
+		stats->tx_bytes += oct->instr_queue[i]->stats.tx_tot_bytes;
+	}
+
+	for (i = 0; i < oct->num_oqs; i++) {
+		if (!oct->droq[i])
+			continue;
+		stats->rx_packets += oct->droq[i]->stats.rx_pkts_received;
+		stats->rx_bytes += oct->droq[i]->stats.rx_bytes_received;
+	}
+}
+
+/**
+ * octeon_mbox_process_cmd:
+ * @mbox: Pointer mailbox
+ * @mbox_cmd: Pointer to command received
+ *
+ * Process the cmd received in mbox
+ */
+static int octeon_mbox_process_cmd(struct octeon_mbox *mbox,
+				   struct octeon_mbox_cmd *mbox_cmd)
+{
+	struct octeon_device *oct = mbox->oct_dev;
+
+	switch (mbox_cmd->msg.s.cmd) {
+	case OCTEON_VF_ACTIVE:
+		dev_dbg(&oct->pci_dev->dev, "got vfactive sending data back\n");
+		mbox_cmd->msg.s.type = OCTEON_MBOX_RESPONSE;
+		mbox_cmd->msg.s.resp_needed = 1;
+		mbox_cmd->msg.s.len = 2;
+		mbox_cmd->data[0] = 0; /* VF version is in mbox_cmd->data[0] */
+		((struct lio_version *)&mbox_cmd->data[0])->major =
+			LIQUIDIO_BASE_MAJOR_VERSION;
+		((struct lio_version *)&mbox_cmd->data[0])->minor =
+			LIQUIDIO_BASE_MINOR_VERSION;
+		((struct lio_version *)&mbox_cmd->data[0])->micro =
+			LIQUIDIO_BASE_MICRO_VERSION;
+		memcpy(mbox_cmd->msg.s.params, (uint8_t *)&oct->pfvf_hsword, 6);
+		/* Sending core cofig info to the corresponding active VF.*/
+		octeon_mbox_write(oct, mbox_cmd);
+		break;
+
+	case OCTEON_VF_FLR_REQUEST:
+		dev_info(&oct->pci_dev->dev,
+			 "got a request for FLR from VF that owns DPI ring %u\n",
+			 mbox->q_no);
+		pcie_flr(oct->sriov_info.dpiring_to_vfpcidev_lut[mbox->q_no]);
+		break;
+
+	case OCTEON_PF_CHANGED_VF_MACADDR:
+		if (OCTEON_CN23XX_VF(oct))
+			octeon_pf_changed_vf_macaddr(oct,
+						     mbox_cmd->msg.s.params);
+		break;
+
+	case OCTEON_GET_VF_STATS:
+		dev_dbg(&oct->pci_dev->dev, "Got VF stats request. Sending data back\n");
+		mbox_cmd->msg.s.type = OCTEON_MBOX_RESPONSE;
+		mbox_cmd->msg.s.resp_needed = 1;
+		mbox_cmd->msg.s.len = 1 +
+			sizeof(struct oct_vf_stats) / sizeof(u64);
+		get_vf_stats(oct, (struct oct_vf_stats *)mbox_cmd->data);
+		octeon_mbox_write(oct, mbox_cmd);
+		break;
+	default:
+		break;
+	}
+	return 0;
+}
+
+/**
+ * octeon_mbox_process_message
+ * @mbox: mailbox
+ *
+ * Process the received mbox message.
+ */
+int octeon_mbox_process_message(struct octeon_mbox *mbox)
+{
+	struct octeon_mbox_cmd mbox_cmd;
+	unsigned long flags;
+
+	spin_lock_irqsave(&mbox->lock, flags);
+
+	if (mbox->state & OCTEON_MBOX_STATE_ERROR) {
+		if (mbox->state & (OCTEON_MBOX_STATE_RESPONSE_PENDING |
+				   OCTEON_MBOX_STATE_RESPONSE_RECEIVING)) {
+			memcpy(&mbox_cmd, &mbox->mbox_resp,
+			       sizeof(struct octeon_mbox_cmd));
+			mbox->state = OCTEON_MBOX_STATE_IDLE;
+			writeq(OCTEON_PFVFSIG, mbox->mbox_read_reg);
+			spin_unlock_irqrestore(&mbox->lock, flags);
+			mbox_cmd.recv_status = 1;
+			if (mbox_cmd.fn)
+				mbox_cmd.fn(mbox->oct_dev, &mbox_cmd,
+					    mbox_cmd.fn_arg);
+			return 0;
+		}
+
+		mbox->state = OCTEON_MBOX_STATE_IDLE;
+		writeq(OCTEON_PFVFSIG, mbox->mbox_read_reg);
+		spin_unlock_irqrestore(&mbox->lock, flags);
+		return 0;
+	}
+
+	if (mbox->state & OCTEON_MBOX_STATE_RESPONSE_RECEIVED) {
+		memcpy(&mbox_cmd, &mbox->mbox_resp,
+		       sizeof(struct octeon_mbox_cmd));
+		mbox->state = OCTEON_MBOX_STATE_IDLE;
+		writeq(OCTEON_PFVFSIG, mbox->mbox_read_reg);
+		spin_unlock_irqrestore(&mbox->lock, flags);
+		mbox_cmd.recv_status = 0;
+		if (mbox_cmd.fn)
+			mbox_cmd.fn(mbox->oct_dev, &mbox_cmd, mbox_cmd.fn_arg);
+		return 0;
+	}
+
+	if (mbox->state & OCTEON_MBOX_STATE_REQUEST_RECEIVED) {
+		memcpy(&mbox_cmd, &mbox->mbox_req,
+		       sizeof(struct octeon_mbox_cmd));
+		if (!mbox_cmd.msg.s.resp_needed) {
+			mbox->state &= ~OCTEON_MBOX_STATE_REQUEST_RECEIVED;
+			if (!(mbox->state &
+			      OCTEON_MBOX_STATE_RESPONSE_PENDING))
+				mbox->state = OCTEON_MBOX_STATE_IDLE;
+			writeq(OCTEON_PFVFSIG, mbox->mbox_read_reg);
+		}
+
+		spin_unlock_irqrestore(&mbox->lock, flags);
+		octeon_mbox_process_cmd(mbox, &mbox_cmd);
+		return 0;
+	}
+
+	spin_unlock_irqrestore(&mbox->lock, flags);
+	WARN_ON(1);
+
+	return 0;
+}
+
+int octeon_mbox_cancel(struct octeon_device *oct, int q_no)
+{
+	struct octeon_mbox *mbox = oct->mbox[q_no];
+	struct octeon_mbox_cmd *mbox_cmd;
+	unsigned long flags = 0;
+
+	spin_lock_irqsave(&mbox->lock, flags);
+	mbox_cmd = &mbox->mbox_resp;
+
+	if (!(mbox->state & OCTEON_MBOX_STATE_RESPONSE_PENDING)) {
+		spin_unlock_irqrestore(&mbox->lock, flags);
+		return 1;
+	}
+
+	mbox->state = OCTEON_MBOX_STATE_IDLE;
+	memset(mbox_cmd, 0, sizeof(*mbox_cmd));
+	writeq(OCTEON_PFVFSIG, mbox->mbox_read_reg);
+	spin_unlock_irqrestore(&mbox->lock, flags);
+
+	return 0;
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_mailbox.h b/drivers/net/ethernet/cavium/liquidio/octeon_mailbox.h
new file mode 100644
index 000000000..d92bd7e16
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_mailbox.h
@@ -0,0 +1,122 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+#ifndef __MAILBOX_H__
+#define __MAILBOX_H__
+
+/* Macros for Mail Box Communication */
+
+#define OCTEON_MBOX_DATA_MAX		32
+
+#define OCTEON_VF_ACTIVE		0x1
+#define OCTEON_VF_FLR_REQUEST		0x2
+#define OCTEON_PF_CHANGED_VF_MACADDR	0x4
+#define OCTEON_GET_VF_STATS		0x8
+
+/*Macro for Read acknowldgement*/
+#define OCTEON_PFVFACK			0xffffffffffffffffULL
+#define OCTEON_PFVFSIG			0x1122334455667788ULL
+#define OCTEON_PFVFERR			0xDEADDEADDEADDEADULL
+
+#define LIO_MBOX_WRITE_WAIT_CNT         1000
+#define LIO_MBOX_WRITE_WAIT_TIME        msecs_to_jiffies(1)
+
+enum octeon_mbox_cmd_status {
+	OCTEON_MBOX_STATUS_SUCCESS = 0,
+	OCTEON_MBOX_STATUS_FAILED = 1,
+	OCTEON_MBOX_STATUS_BUSY = 2
+};
+
+enum octeon_mbox_message_type {
+	OCTEON_MBOX_REQUEST = 0,
+	OCTEON_MBOX_RESPONSE = 1
+};
+
+union octeon_mbox_message {
+	u64 u64;
+	struct {
+		u16 type : 1;
+		u16 resp_needed : 1;
+		u16 cmd : 6;
+		u16 len : 8;
+		u8 params[6];
+	} s;
+};
+
+typedef void (*octeon_mbox_callback_t)(void *, void *, void *);
+
+struct octeon_mbox_cmd {
+	union octeon_mbox_message msg;
+	u64 data[OCTEON_MBOX_DATA_MAX];
+	u32 q_no;
+	u32 recv_len;
+	u32 recv_status;
+	octeon_mbox_callback_t fn;
+	void *fn_arg;
+};
+
+enum octeon_mbox_state {
+	OCTEON_MBOX_STATE_IDLE = 1,
+	OCTEON_MBOX_STATE_REQUEST_RECEIVING = 2,
+	OCTEON_MBOX_STATE_REQUEST_RECEIVED = 4,
+	OCTEON_MBOX_STATE_RESPONSE_PENDING = 8,
+	OCTEON_MBOX_STATE_RESPONSE_RECEIVING = 16,
+	OCTEON_MBOX_STATE_RESPONSE_RECEIVED = 32,
+	OCTEON_MBOX_STATE_ERROR = 64
+};
+
+struct octeon_mbox {
+	/** A spinlock to protect access to this q_mbox. */
+	spinlock_t lock;
+
+	struct octeon_device *oct_dev;
+
+	u32 q_no;
+
+	enum octeon_mbox_state state;
+
+	struct cavium_wk mbox_poll_wk;
+
+	/** SLI_MAC_PF_MBOX_INT for PF, SLI_PKT_MBOX_INT for VF. */
+	void *mbox_int_reg;
+
+	/** SLI_PKT_PF_VF_MBOX_SIG(0) for PF, SLI_PKT_PF_VF_MBOX_SIG(1) for VF.
+	 */
+	void *mbox_write_reg;
+
+	/** SLI_PKT_PF_VF_MBOX_SIG(1) for PF, SLI_PKT_PF_VF_MBOX_SIG(0) for VF.
+	 */
+	void *mbox_read_reg;
+
+	struct octeon_mbox_cmd mbox_req;
+
+	struct octeon_mbox_cmd mbox_resp;
+
+};
+
+struct oct_vf_stats_ctx {
+	atomic_t status;
+	struct oct_vf_stats *stats;
+};
+
+int octeon_mbox_read(struct octeon_mbox *mbox);
+int octeon_mbox_write(struct octeon_device *oct,
+		      struct octeon_mbox_cmd *mbox_cmd);
+int octeon_mbox_process_message(struct octeon_mbox *mbox);
+int octeon_mbox_cancel(struct octeon_device *oct, int q_no);
+
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_main.h b/drivers/net/ethernet/cavium/liquidio/octeon_main.h
new file mode 100644
index 000000000..5b4cb725f
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_main.h
@@ -0,0 +1,236 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more details.
+ ***********************************************************************/
+/*! \file octeon_main.h
+ *  \brief Host Driver: This file is included by all host driver source files
+ *  to include common definitions.
+ */
+
+#ifndef _OCTEON_MAIN_H_
+#define  _OCTEON_MAIN_H_
+
+#include <linux/sched/signal.h>
+
+#if BITS_PER_LONG == 32
+#define CVM_CAST64(v) ((long long)(v))
+#elif BITS_PER_LONG == 64
+#define CVM_CAST64(v) ((long long)(long)(v))
+#else
+#error "Unknown system architecture"
+#endif
+
+#define DRV_NAME "LiquidIO"
+
+struct octeon_device_priv {
+	/** Tasklet structures for this device. */
+	struct tasklet_struct droq_tasklet;
+	unsigned long napi_mask;
+	struct octeon_device *dev;
+};
+
+/** This structure is used by NIC driver to store information required
+ * to free the sk_buff when the packet has been fetched by Octeon.
+ * Bytes offset below assume worst-case of a 64-bit system.
+ */
+struct octnet_buf_free_info {
+	/** Bytes 1-8.  Pointer to network device private structure. */
+	struct lio *lio;
+
+	/** Bytes 9-16.  Pointer to sk_buff. */
+	struct sk_buff *skb;
+
+	/** Bytes 17-24.  Pointer to gather list. */
+	struct octnic_gather *g;
+
+	/** Bytes 25-32. Physical address of skb->data or gather list. */
+	u64 dptr;
+
+	/** Bytes 33-47. Piggybacked soft command, if any */
+	struct octeon_soft_command *sc;
+};
+
+/* BQL-related functions */
+int octeon_report_sent_bytes_to_bql(void *buf, int reqtype);
+void octeon_update_tx_completion_counters(void *buf, int reqtype,
+					  unsigned int *pkts_compl,
+					  unsigned int *bytes_compl);
+void octeon_report_tx_completion_to_bql(void *txq, unsigned int pkts_compl,
+					unsigned int bytes_compl);
+void octeon_pf_changed_vf_macaddr(struct octeon_device *oct, u8 *mac);
+
+void octeon_schedule_rxq_oom_work(struct octeon_device *oct,
+				  struct octeon_droq *droq);
+
+/** Swap 8B blocks */
+static inline void octeon_swap_8B_data(u64 *data, u32 blocks)
+{
+	while (blocks) {
+		cpu_to_be64s(data);
+		blocks--;
+		data++;
+	}
+}
+
+/**
+ * \brief unmaps a PCI BAR
+ * @param oct Pointer to Octeon device
+ * @param baridx bar index
+ */
+static inline void octeon_unmap_pci_barx(struct octeon_device *oct, int baridx)
+{
+	dev_dbg(&oct->pci_dev->dev, "Freeing PCI mapped regions for Bar%d\n",
+		baridx);
+
+	if (oct->mmio[baridx].done)
+		iounmap(oct->mmio[baridx].hw_addr);
+
+	if (oct->mmio[baridx].start)
+		pci_release_region(oct->pci_dev, baridx * 2);
+}
+
+/**
+ * \brief maps a PCI BAR
+ * @param oct Pointer to Octeon device
+ * @param baridx bar index
+ * @param max_map_len maximum length of mapped memory
+ */
+static inline int octeon_map_pci_barx(struct octeon_device *oct,
+				      int baridx, int max_map_len)
+{
+	u32 mapped_len = 0;
+
+	if (pci_request_region(oct->pci_dev, baridx * 2, DRV_NAME)) {
+		dev_err(&oct->pci_dev->dev, "pci_request_region failed for bar %d\n",
+			baridx);
+		return 1;
+	}
+
+	oct->mmio[baridx].start = pci_resource_start(oct->pci_dev, baridx * 2);
+	oct->mmio[baridx].len = pci_resource_len(oct->pci_dev, baridx * 2);
+
+	mapped_len = oct->mmio[baridx].len;
+	if (!mapped_len)
+		goto err_release_region;
+
+	if (max_map_len && (mapped_len > max_map_len))
+		mapped_len = max_map_len;
+
+	oct->mmio[baridx].hw_addr =
+		ioremap(oct->mmio[baridx].start, mapped_len);
+	oct->mmio[baridx].mapped_len = mapped_len;
+
+	dev_dbg(&oct->pci_dev->dev, "BAR%d start: 0x%llx mapped %u of %u bytes\n",
+		baridx, oct->mmio[baridx].start, mapped_len,
+		oct->mmio[baridx].len);
+
+	if (!oct->mmio[baridx].hw_addr) {
+		dev_err(&oct->pci_dev->dev, "error ioremap for bar %d\n",
+			baridx);
+		goto err_release_region;
+	}
+	oct->mmio[baridx].done = 1;
+
+	return 0;
+
+err_release_region:
+	pci_release_region(oct->pci_dev, baridx * 2);
+	return 1;
+}
+
+/* input parameter:
+ * sc: pointer to a soft request
+ * timeout: milli sec which an application wants to wait for the
+	    response of the request.
+ *          0: the request will wait until its response gets back
+ *	       from the firmware within LIO_SC_MAX_TMO_MS milli sec.
+ *	       It the response does not return within
+ *	       LIO_SC_MAX_TMO_MS milli sec, lio_process_ordered_list()
+ *	       will move the request to zombie response list.
+ *
+ * return value:
+ * 0: got the response from firmware for the sc request.
+ * errno -EINTR: user abort the command.
+ * errno -ETIME: user spefified timeout value has been expired.
+ * errno -EBUSY: the response of the request does not return in
+ *               resonable time (LIO_SC_MAX_TMO_MS).
+ *               the sc wll be move to zombie response list by
+ *               lio_process_ordered_list()
+ *
+ * A request with non-zero return value, the sc->caller_is_done
+ *  will be marked 1.
+ * When getting a request with zero return value, the requestor
+ *  should mark sc->caller_is_done with 1 after examing the
+ *  response of sc.
+ * lio_process_ordered_list() will free the soft command on behalf
+ * of the soft command requestor.
+ * This is to fix the possible race condition of both timeout process
+ * and lio_process_ordered_list()/callback function to free a
+ * sc strucutre.
+ */
+static inline int
+wait_for_sc_completion_timeout(struct octeon_device *oct_dev,
+			       struct octeon_soft_command *sc,
+			       unsigned long timeout)
+{
+	int errno = 0;
+	long timeout_jiff;
+
+	if (timeout)
+		timeout_jiff = msecs_to_jiffies(timeout);
+	else
+		timeout_jiff = MAX_SCHEDULE_TIMEOUT;
+
+	timeout_jiff =
+		wait_for_completion_interruptible_timeout(&sc->complete,
+							  timeout_jiff);
+	if (timeout_jiff == 0) {
+		dev_err(&oct_dev->pci_dev->dev, "%s: sc is timeout\n",
+			__func__);
+		WRITE_ONCE(sc->caller_is_done, true);
+		errno = -ETIME;
+	} else if (timeout_jiff == -ERESTARTSYS) {
+		dev_err(&oct_dev->pci_dev->dev, "%s: sc is interrupted\n",
+			__func__);
+		WRITE_ONCE(sc->caller_is_done, true);
+		errno = -EINTR;
+	} else  if (sc->sc_status == OCTEON_REQUEST_TIMEOUT) {
+		dev_err(&oct_dev->pci_dev->dev, "%s: sc has fatal timeout\n",
+			__func__);
+		WRITE_ONCE(sc->caller_is_done, true);
+		errno = -EBUSY;
+	}
+
+	return errno;
+}
+
+#ifndef ROUNDUP4
+#define ROUNDUP4(val) (((val) + 3) & 0xfffffffc)
+#endif
+
+#ifndef ROUNDUP8
+#define ROUNDUP8(val) (((val) + 7) & 0xfffffff8)
+#endif
+
+#ifndef ROUNDUP16
+#define ROUNDUP16(val) (((val) + 15) & 0xfffffff0)
+#endif
+
+#ifndef ROUNDUP128
+#define ROUNDUP128(val) (((val) + 127) & 0xffffff80)
+#endif
+
+#endif /* _OCTEON_MAIN_H_ */
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_mem_ops.c b/drivers/net/ethernet/cavium/liquidio/octeon_mem_ops.c
new file mode 100644
index 000000000..7ccab3614
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_mem_ops.c
@@ -0,0 +1,201 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more
+ * details.
+ **********************************************************************/
+#include <linux/netdevice.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_mem_ops.h"
+
+#define MEMOPS_IDX   BAR1_INDEX_DYNAMIC_MAP
+
+#ifdef __BIG_ENDIAN_BITFIELD
+static inline void
+octeon_toggle_bar1_swapmode(struct octeon_device *oct, u32 idx)
+{
+	u32 mask;
+
+	mask = oct->fn_list.bar1_idx_read(oct, idx);
+	mask = (mask & 0x2) ? (mask & ~2) : (mask | 2);
+	oct->fn_list.bar1_idx_write(oct, idx, mask);
+}
+#else
+#define octeon_toggle_bar1_swapmode(oct, idx)
+#endif
+
+static void
+octeon_pci_fastwrite(struct octeon_device *oct, u8 __iomem *mapped_addr,
+		     u8 *hostbuf, u32 len)
+{
+	while ((len) && ((unsigned long)mapped_addr) & 7) {
+		writeb(*(hostbuf++), mapped_addr++);
+		len--;
+	}
+
+	octeon_toggle_bar1_swapmode(oct, MEMOPS_IDX);
+
+	while (len >= 8) {
+		writeq(*((u64 *)hostbuf), mapped_addr);
+		mapped_addr += 8;
+		hostbuf += 8;
+		len -= 8;
+	}
+
+	octeon_toggle_bar1_swapmode(oct, MEMOPS_IDX);
+
+	while (len--)
+		writeb(*(hostbuf++), mapped_addr++);
+}
+
+static void
+octeon_pci_fastread(struct octeon_device *oct, u8 __iomem *mapped_addr,
+		    u8 *hostbuf, u32 len)
+{
+	while ((len) && ((unsigned long)mapped_addr) & 7) {
+		*(hostbuf++) = readb(mapped_addr++);
+		len--;
+	}
+
+	octeon_toggle_bar1_swapmode(oct, MEMOPS_IDX);
+
+	while (len >= 8) {
+		*((u64 *)hostbuf) = readq(mapped_addr);
+		mapped_addr += 8;
+		hostbuf += 8;
+		len -= 8;
+	}
+
+	octeon_toggle_bar1_swapmode(oct, MEMOPS_IDX);
+
+	while (len--)
+		*(hostbuf++) = readb(mapped_addr++);
+}
+
+/* Core mem read/write with temporary bar1 settings. */
+/* op = 1 to read, op = 0 to write. */
+static void
+__octeon_pci_rw_core_mem(struct octeon_device *oct, u64 addr,
+			 u8 *hostbuf, u32 len, u32 op)
+{
+	u32 copy_len = 0, index_reg_val = 0;
+	unsigned long flags;
+	u8 __iomem *mapped_addr;
+	u64 static_mapping_base;
+
+	static_mapping_base = oct->console_nb_info.dram_region_base;
+
+	if (static_mapping_base &&
+	    static_mapping_base == (addr & ~(OCTEON_BAR1_ENTRY_SIZE - 1ULL))) {
+		int bar1_index = oct->console_nb_info.bar1_index;
+
+		mapped_addr = oct->mmio[1].hw_addr
+			+ (bar1_index << ilog2(OCTEON_BAR1_ENTRY_SIZE))
+			+ (addr & (OCTEON_BAR1_ENTRY_SIZE - 1ULL));
+
+		if (op)
+			octeon_pci_fastread(oct, mapped_addr, hostbuf, len);
+		else
+			octeon_pci_fastwrite(oct, mapped_addr, hostbuf, len);
+
+		return;
+	}
+
+	spin_lock_irqsave(&oct->mem_access_lock, flags);
+
+	/* Save the original index reg value. */
+	index_reg_val = oct->fn_list.bar1_idx_read(oct, MEMOPS_IDX);
+	do {
+		oct->fn_list.bar1_idx_setup(oct, addr, MEMOPS_IDX, 1);
+		mapped_addr = oct->mmio[1].hw_addr
+		    + (MEMOPS_IDX << 22) + (addr & 0x3fffff);
+
+		/* If operation crosses a 4MB boundary, split the transfer
+		 * at the 4MB
+		 * boundary.
+		 */
+		if (((addr + len - 1) & ~(0x3fffff)) != (addr & ~(0x3fffff))) {
+			copy_len = (u32)(((addr & ~(0x3fffff)) +
+				   (MEMOPS_IDX << 22)) - addr);
+		} else {
+			copy_len = len;
+		}
+
+		if (op) {	/* read from core */
+			octeon_pci_fastread(oct, mapped_addr, hostbuf,
+					    copy_len);
+		} else {
+			octeon_pci_fastwrite(oct, mapped_addr, hostbuf,
+					     copy_len);
+		}
+
+		len -= copy_len;
+		addr += copy_len;
+		hostbuf += copy_len;
+
+	} while (len);
+
+	oct->fn_list.bar1_idx_write(oct, MEMOPS_IDX, index_reg_val);
+
+	spin_unlock_irqrestore(&oct->mem_access_lock, flags);
+}
+
+void
+octeon_pci_read_core_mem(struct octeon_device *oct,
+			 u64 coreaddr,
+			 u8 *buf,
+			 u32 len)
+{
+	__octeon_pci_rw_core_mem(oct, coreaddr, buf, len, 1);
+}
+
+void
+octeon_pci_write_core_mem(struct octeon_device *oct,
+			  u64 coreaddr,
+			  const u8 *buf,
+			  u32 len)
+{
+	__octeon_pci_rw_core_mem(oct, coreaddr, (u8 *)buf, len, 0);
+}
+
+u64 octeon_read_device_mem64(struct octeon_device *oct, u64 coreaddr)
+{
+	__be64 ret;
+
+	__octeon_pci_rw_core_mem(oct, coreaddr, (u8 *)&ret, 8, 1);
+
+	return be64_to_cpu(ret);
+}
+
+u32 octeon_read_device_mem32(struct octeon_device *oct, u64 coreaddr)
+{
+	__be32 ret;
+
+	__octeon_pci_rw_core_mem(oct, coreaddr, (u8 *)&ret, 4, 1);
+
+	return be32_to_cpu(ret);
+}
+
+void octeon_write_device_mem32(struct octeon_device *oct, u64 coreaddr,
+			       u32 val)
+{
+	__be32 t = cpu_to_be32(val);
+
+	__octeon_pci_rw_core_mem(oct, coreaddr, (u8 *)&t, 4, 0);
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_mem_ops.h b/drivers/net/ethernet/cavium/liquidio/octeon_mem_ops.h
new file mode 100644
index 000000000..47a3ff5f9
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_mem_ops.h
@@ -0,0 +1,72 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more
+ * details.
+ **********************************************************************/
+
+/*!  \file octeon_mem_ops.h
+ *   \brief Host Driver: Routines used to read/write Octeon memory.
+ */
+
+#ifndef __OCTEON_MEM_OPS_H__
+#define __OCTEON_MEM_OPS_H__
+
+/**  Read a 64-bit value from a BAR1 mapped core memory address.
+ *   @param  oct        -  pointer to the octeon device.
+ *   @param  core_addr  -  the address to read from.
+ *
+ *   The range_idx gives the BAR1 index register for the range of address
+ *   in which core_addr is mapped.
+ *
+ *   @return  64-bit value read from Core memory
+ */
+u64 octeon_read_device_mem64(struct octeon_device *oct, u64 core_addr);
+
+/**  Read a 32-bit value from a BAR1 mapped core memory address.
+ *   @param  oct        -  pointer to the octeon device.
+ *   @param  core_addr  -  the address to read from.
+ *
+ *   @return  32-bit value read from Core memory
+ */
+u32 octeon_read_device_mem32(struct octeon_device *oct, u64 core_addr);
+
+/**  Write a 32-bit value to a BAR1 mapped core memory address.
+ *   @param  oct        -  pointer to the octeon device.
+ *   @param  core_addr  -  the address to write to.
+ *   @param  val        -  32-bit value to write.
+ */
+void
+octeon_write_device_mem32(struct octeon_device *oct,
+			  u64 core_addr,
+			  u32 val);
+
+/** Read multiple bytes from Octeon memory.
+ */
+void
+octeon_pci_read_core_mem(struct octeon_device *oct,
+			 u64 coreaddr,
+			 u8 *buf,
+			 u32 len);
+
+/** Write multiple bytes into Octeon memory.
+ */
+void
+octeon_pci_write_core_mem(struct octeon_device *oct,
+			  u64 coreaddr,
+			  const u8 *buf,
+			  u32 len);
+
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_network.h b/drivers/net/ethernet/cavium/liquidio/octeon_network.h
new file mode 100644
index 000000000..ebe56bd88
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_network.h
@@ -0,0 +1,626 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more
+ * details.
+ **********************************************************************/
+
+/*!  \file  octeon_network.h
+ *   \brief Host NIC Driver: Structure and Macro definitions used by NIC Module.
+ */
+
+#ifndef __OCTEON_NETWORK_H__
+#define __OCTEON_NETWORK_H__
+#include <linux/ptp_clock_kernel.h>
+
+#define LIO_MAX_MTU_SIZE (OCTNET_MAX_FRM_SIZE - OCTNET_FRM_HEADER_SIZE)
+#define LIO_MIN_MTU_SIZE ETH_MIN_MTU
+
+/* Bit mask values for lio->ifstate */
+#define   LIO_IFSTATE_DROQ_OPS             0x01
+#define   LIO_IFSTATE_REGISTERED           0x02
+#define   LIO_IFSTATE_RUNNING              0x04
+#define   LIO_IFSTATE_RX_TIMESTAMP_ENABLED 0x08
+#define   LIO_IFSTATE_RESETTING		   0x10
+
+struct liquidio_if_cfg_resp {
+	u64 rh;
+	struct liquidio_if_cfg_info cfg_info;
+	u64 status;
+};
+
+#define LIO_IFCFG_WAIT_TIME    3000 /* In milli seconds */
+#define LIQUIDIO_NDEV_STATS_POLL_TIME_MS 200
+
+/* Structure of a node in list of gather components maintained by
+ * NIC driver for each network device.
+ */
+struct octnic_gather {
+	/* List manipulation. Next and prev pointers. */
+	struct list_head list;
+
+	/* Size of the gather component at sg in bytes. */
+	int sg_size;
+
+	/* Number of bytes that sg was adjusted to make it 8B-aligned. */
+	int adjust;
+
+	/* Gather component that can accommodate max sized fragment list
+	 * received from the IP layer.
+	 */
+	struct octeon_sg_entry *sg;
+
+	dma_addr_t sg_dma_ptr;
+};
+
+struct oct_nic_stats_resp {
+	u64     rh;
+	struct oct_link_stats stats;
+	u64     status;
+};
+
+struct oct_nic_vf_stats_resp {
+	u64     rh;
+	u64	spoofmac_cnt;
+	u64     status;
+};
+
+struct oct_nic_stats_ctrl {
+	struct completion complete;
+	struct net_device *netdev;
+};
+
+struct oct_nic_seapi_resp {
+	u64 rh;
+	union {
+		u32 fec_setting;
+		u32 speed;
+	};
+	u64 status;
+};
+
+/** LiquidIO per-interface network private data */
+struct lio {
+	/** State of the interface. Rx/Tx happens only in the RUNNING state.  */
+	atomic_t ifstate;
+
+	/** Octeon Interface index number. This device will be represented as
+	 *  oct<ifidx> in the system.
+	 */
+	int ifidx;
+
+	/** Octeon Input queue to use to transmit for this network interface. */
+	int txq;
+
+	/** Octeon Output queue from which pkts arrive
+	 * for this network interface.
+	 */
+	int rxq;
+
+	/** Guards each glist */
+	spinlock_t *glist_lock;
+
+	/** Array of gather component linked lists */
+	struct list_head *glist;
+	void **glists_virt_base;
+	dma_addr_t *glists_dma_base;
+	u32 glist_entry_size;
+
+	/** Pointer to the NIC properties for the Octeon device this network
+	 *  interface is associated with.
+	 */
+	struct octdev_props *octprops;
+
+	/** Pointer to the octeon device structure. */
+	struct octeon_device *oct_dev;
+
+	struct net_device *netdev;
+
+	/** Link information sent by the core application for this interface. */
+	struct oct_link_info linfo;
+
+	/** counter of link changes */
+	u64 link_changes;
+
+	/** Size of Tx queue for this octeon device. */
+	u32 tx_qsize;
+
+	/** Size of Rx queue for this octeon device. */
+	u32 rx_qsize;
+
+	/** Size of MTU this octeon device. */
+	u32 mtu;
+
+	/** msg level flag per interface. */
+	u32 msg_enable;
+
+	/** Copy of Interface capabilities: TSO, TSO6, LRO, Chescksums . */
+	u64 dev_capability;
+
+	/* Copy of transmit encapsulation capabilities:
+	 * TSO, TSO6, Checksums for this device for Kernel
+	 * 3.10.0 onwards
+	 */
+	u64 enc_dev_capability;
+
+	/** Copy of beacaon reg in phy */
+	u32 phy_beacon_val;
+
+	/** Copy of ctrl reg in phy */
+	u32 led_ctrl_val;
+
+	/* PTP clock information */
+	struct ptp_clock_info ptp_info;
+	struct ptp_clock *ptp_clock;
+	s64 ptp_adjust;
+
+	/* for atomic access to Octeon PTP reg and data struct */
+	spinlock_t ptp_lock;
+
+	/* Interface info */
+	u32	intf_open;
+
+	/* work queue for  txq status */
+	struct cavium_wq	txq_status_wq;
+
+	/* work queue for  rxq oom status */
+	struct cavium_wq rxq_status_wq[MAX_POSSIBLE_OCTEON_OUTPUT_QUEUES];
+
+	/* work queue for  link status */
+	struct cavium_wq	link_status_wq;
+
+	/* work queue to regularly send local time to octeon firmware */
+	struct cavium_wq	sync_octeon_time_wq;
+
+	int netdev_uc_count;
+	struct cavium_wk stats_wk;
+};
+
+#define LIO_SIZE         (sizeof(struct lio))
+#define GET_LIO(netdev)  ((struct lio *)netdev_priv(netdev))
+
+#define LIO_MAX_CORES                16
+
+/**
+ * \brief Enable or disable feature
+ * @param netdev    pointer to network device
+ * @param cmd       Command that just requires acknowledgment
+ * @param param1    Parameter to command
+ */
+int liquidio_set_feature(struct net_device *netdev, int cmd, u16 param1);
+
+int setup_rx_oom_poll_fn(struct net_device *netdev);
+
+void cleanup_rx_oom_poll_fn(struct net_device *netdev);
+
+/**
+ * \brief Link control command completion callback
+ * @param nctrl_ptr pointer to control packet structure
+ *
+ * This routine is called by the callback function when a ctrl pkt sent to
+ * core app completes. The nctrl_ptr contains a copy of the command type
+ * and data sent to the core app. This routine is only called if the ctrl
+ * pkt was sent successfully to the core app.
+ */
+void liquidio_link_ctrl_cmd_completion(void *nctrl_ptr);
+
+int liquidio_setup_io_queues(struct octeon_device *octeon_dev, int ifidx,
+			     u32 num_iqs, u32 num_oqs);
+
+irqreturn_t liquidio_msix_intr_handler(int irq __attribute__((unused)),
+				       void *dev);
+
+int octeon_setup_interrupt(struct octeon_device *oct, u32 num_ioqs);
+
+void lio_fetch_stats(struct work_struct *work);
+
+int lio_wait_for_clean_oq(struct octeon_device *oct);
+/**
+ * \brief Register ethtool operations
+ * @param netdev    pointer to network device
+ */
+void liquidio_set_ethtool_ops(struct net_device *netdev);
+
+void lio_delete_glists(struct lio *lio);
+
+int lio_setup_glists(struct octeon_device *oct, struct lio *lio, int num_qs);
+
+int liquidio_get_speed(struct lio *lio);
+int liquidio_set_speed(struct lio *lio, int speed);
+int liquidio_get_fec(struct lio *lio);
+int liquidio_set_fec(struct lio *lio, int on_off);
+
+/**
+ * \brief Net device change_mtu
+ * @param netdev network device
+ */
+int liquidio_change_mtu(struct net_device *netdev, int new_mtu);
+#define LIO_CHANGE_MTU_SUCCESS 1
+#define LIO_CHANGE_MTU_FAIL    2
+
+#define SKB_ADJ_MASK  0x3F
+#define SKB_ADJ       (SKB_ADJ_MASK + 1)
+
+#define MIN_SKB_SIZE       256 /* 8 bytes and more - 8 bytes for PTP */
+#define LIO_RXBUFFER_SZ    2048
+
+static inline void
+*recv_buffer_alloc(struct octeon_device *oct,
+		   struct octeon_skb_page_info *pg_info)
+{
+	struct page *page;
+	struct sk_buff *skb;
+	struct octeon_skb_page_info *skb_pg_info;
+
+	page = alloc_page(GFP_ATOMIC);
+	if (unlikely(!page))
+		return NULL;
+
+	skb = dev_alloc_skb(MIN_SKB_SIZE + SKB_ADJ);
+	if (unlikely(!skb)) {
+		__free_page(page);
+		pg_info->page = NULL;
+		return NULL;
+	}
+
+	if ((unsigned long)skb->data & SKB_ADJ_MASK) {
+		u32 r = SKB_ADJ - ((unsigned long)skb->data & SKB_ADJ_MASK);
+
+		skb_reserve(skb, r);
+	}
+
+	skb_pg_info = ((struct octeon_skb_page_info *)(skb->cb));
+	/* Get DMA info */
+	pg_info->dma = dma_map_page(&oct->pci_dev->dev, page, 0,
+				    PAGE_SIZE, DMA_FROM_DEVICE);
+
+	/* Mapping failed!! */
+	if (dma_mapping_error(&oct->pci_dev->dev, pg_info->dma)) {
+		__free_page(page);
+		dev_kfree_skb_any((struct sk_buff *)skb);
+		pg_info->page = NULL;
+		return NULL;
+	}
+
+	pg_info->page = page;
+	pg_info->page_offset = 0;
+	skb_pg_info->page = page;
+	skb_pg_info->page_offset = 0;
+	skb_pg_info->dma = pg_info->dma;
+
+	return (void *)skb;
+}
+
+static inline void
+*recv_buffer_fast_alloc(u32 size)
+{
+	struct sk_buff *skb;
+	struct octeon_skb_page_info *skb_pg_info;
+
+	skb = dev_alloc_skb(size + SKB_ADJ);
+	if (unlikely(!skb))
+		return NULL;
+
+	if ((unsigned long)skb->data & SKB_ADJ_MASK) {
+		u32 r = SKB_ADJ - ((unsigned long)skb->data & SKB_ADJ_MASK);
+
+		skb_reserve(skb, r);
+	}
+
+	skb_pg_info = ((struct octeon_skb_page_info *)(skb->cb));
+	skb_pg_info->page = NULL;
+	skb_pg_info->page_offset = 0;
+	skb_pg_info->dma = 0;
+
+	return skb;
+}
+
+static inline int
+recv_buffer_recycle(struct octeon_device *oct, void *buf)
+{
+	struct octeon_skb_page_info *pg_info = buf;
+
+	if (!pg_info->page) {
+		dev_err(&oct->pci_dev->dev, "%s: pg_info->page NULL\n",
+			__func__);
+		return -ENOMEM;
+	}
+
+	if (unlikely(page_count(pg_info->page) != 1) ||
+	    unlikely(page_to_nid(pg_info->page)	!= numa_node_id())) {
+		dma_unmap_page(&oct->pci_dev->dev,
+			       pg_info->dma, (PAGE_SIZE << 0),
+			       DMA_FROM_DEVICE);
+		pg_info->dma = 0;
+		pg_info->page = NULL;
+		pg_info->page_offset = 0;
+		return -ENOMEM;
+	}
+
+	/* Flip to other half of the buffer */
+	if (pg_info->page_offset == 0)
+		pg_info->page_offset = LIO_RXBUFFER_SZ;
+	else
+		pg_info->page_offset = 0;
+	page_ref_inc(pg_info->page);
+
+	return 0;
+}
+
+static inline void
+*recv_buffer_reuse(struct octeon_device *oct, void *buf)
+{
+	struct octeon_skb_page_info *pg_info = buf, *skb_pg_info;
+	struct sk_buff *skb;
+
+	skb = dev_alloc_skb(MIN_SKB_SIZE + SKB_ADJ);
+	if (unlikely(!skb)) {
+		dma_unmap_page(&oct->pci_dev->dev,
+			       pg_info->dma, (PAGE_SIZE << 0),
+			       DMA_FROM_DEVICE);
+		return NULL;
+	}
+
+	if ((unsigned long)skb->data & SKB_ADJ_MASK) {
+		u32 r = SKB_ADJ - ((unsigned long)skb->data & SKB_ADJ_MASK);
+
+		skb_reserve(skb, r);
+	}
+
+	skb_pg_info = ((struct octeon_skb_page_info *)(skb->cb));
+	skb_pg_info->page = pg_info->page;
+	skb_pg_info->page_offset = pg_info->page_offset;
+	skb_pg_info->dma = pg_info->dma;
+
+	return skb;
+}
+
+static inline void
+recv_buffer_destroy(void *buffer, struct octeon_skb_page_info *pg_info)
+{
+	struct sk_buff *skb = (struct sk_buff *)buffer;
+
+	put_page(pg_info->page);
+	pg_info->dma = 0;
+	pg_info->page = NULL;
+	pg_info->page_offset = 0;
+
+	if (skb)
+		dev_kfree_skb_any(skb);
+}
+
+static inline void recv_buffer_free(void *buffer)
+{
+	struct sk_buff *skb = (struct sk_buff *)buffer;
+	struct octeon_skb_page_info *pg_info;
+
+	pg_info = ((struct octeon_skb_page_info *)(skb->cb));
+
+	if (pg_info->page) {
+		put_page(pg_info->page);
+		pg_info->dma = 0;
+		pg_info->page = NULL;
+		pg_info->page_offset = 0;
+	}
+
+	dev_kfree_skb_any((struct sk_buff *)buffer);
+}
+
+static inline void
+recv_buffer_fast_free(void *buffer)
+{
+	dev_kfree_skb_any((struct sk_buff *)buffer);
+}
+
+static inline void tx_buffer_free(void *buffer)
+{
+	dev_kfree_skb_any((struct sk_buff *)buffer);
+}
+
+#define lio_dma_alloc(oct, size, dma_addr) \
+	dma_alloc_coherent(&(oct)->pci_dev->dev, size, dma_addr, GFP_KERNEL)
+#define lio_dma_free(oct, size, virt_addr, dma_addr) \
+	dma_free_coherent(&(oct)->pci_dev->dev, size, virt_addr, dma_addr)
+
+static inline
+void *get_rbd(struct sk_buff *skb)
+{
+	struct octeon_skb_page_info *pg_info;
+	unsigned char *va;
+
+	pg_info = ((struct octeon_skb_page_info *)(skb->cb));
+	va = page_address(pg_info->page) + pg_info->page_offset;
+
+	return va;
+}
+
+static inline u64
+lio_map_ring(void *buf)
+{
+	dma_addr_t dma_addr;
+
+	struct sk_buff *skb = (struct sk_buff *)buf;
+	struct octeon_skb_page_info *pg_info;
+
+	pg_info = ((struct octeon_skb_page_info *)(skb->cb));
+	if (!pg_info->page) {
+		pr_err("%s: pg_info->page NULL\n", __func__);
+		WARN_ON(1);
+	}
+
+	/* Get DMA info */
+	dma_addr = pg_info->dma;
+	if (!pg_info->dma) {
+		pr_err("%s: ERROR it should be already available\n",
+		       __func__);
+		WARN_ON(1);
+	}
+	dma_addr += pg_info->page_offset;
+
+	return (u64)dma_addr;
+}
+
+static inline void
+lio_unmap_ring(struct pci_dev *pci_dev,
+	       u64 buf_ptr)
+
+{
+	dma_unmap_page(&pci_dev->dev,
+		       buf_ptr, (PAGE_SIZE << 0),
+		       DMA_FROM_DEVICE);
+}
+
+static inline void *octeon_fast_packet_alloc(u32 size)
+{
+	return recv_buffer_fast_alloc(size);
+}
+
+static inline void octeon_fast_packet_next(struct octeon_droq *droq,
+					   struct sk_buff *nicbuf,
+					   int copy_len,
+					   int idx)
+{
+	skb_put_data(nicbuf, get_rbd(droq->recv_buf_list[idx].buffer),
+		     copy_len);
+}
+
+/**
+ * \brief check interface state
+ * @param lio per-network private data
+ * @param state_flag flag state to check
+ */
+static inline int ifstate_check(struct lio *lio, int state_flag)
+{
+	return atomic_read(&lio->ifstate) & state_flag;
+}
+
+/**
+ * \brief set interface state
+ * @param lio per-network private data
+ * @param state_flag flag state to set
+ */
+static inline void ifstate_set(struct lio *lio, int state_flag)
+{
+	atomic_set(&lio->ifstate, (atomic_read(&lio->ifstate) | state_flag));
+}
+
+/**
+ * \brief clear interface state
+ * @param lio per-network private data
+ * @param state_flag flag state to clear
+ */
+static inline void ifstate_reset(struct lio *lio, int state_flag)
+{
+	atomic_set(&lio->ifstate, (atomic_read(&lio->ifstate) & ~(state_flag)));
+}
+
+/**
+ * \brief wait for all pending requests to complete
+ * @param oct Pointer to Octeon device
+ *
+ * Called during shutdown sequence
+ */
+static inline int wait_for_pending_requests(struct octeon_device *oct)
+{
+	int i, pcount = 0;
+
+	for (i = 0; i < MAX_IO_PENDING_PKT_COUNT; i++) {
+		pcount = atomic_read(
+		    &oct->response_list[OCTEON_ORDERED_SC_LIST]
+			 .pending_req_count);
+		if (pcount)
+			schedule_timeout_uninterruptible(HZ / 10);
+		else
+			break;
+	}
+
+	if (pcount)
+		return 1;
+
+	return 0;
+}
+
+/**
+ * \brief Stop Tx queues
+ * @param netdev network device
+ */
+static inline void stop_txqs(struct net_device *netdev)
+{
+	int i;
+
+	for (i = 0; i < netdev->real_num_tx_queues; i++)
+		netif_stop_subqueue(netdev, i);
+}
+
+/**
+ * \brief Wake Tx queues
+ * @param netdev network device
+ */
+static inline void wake_txqs(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	int i, qno;
+
+	for (i = 0; i < netdev->real_num_tx_queues; i++) {
+		qno = lio->linfo.txpciq[i % lio->oct_dev->num_iqs].s.q_no;
+
+		if (__netif_subqueue_stopped(netdev, i)) {
+			INCR_INSTRQUEUE_PKT_COUNT(lio->oct_dev, qno,
+						  tx_restart, 1);
+			netif_wake_subqueue(netdev, i);
+		}
+	}
+}
+
+/**
+ * \brief Start Tx queues
+ * @param netdev network device
+ */
+static inline void start_txqs(struct net_device *netdev)
+{
+	struct lio *lio = GET_LIO(netdev);
+	int i;
+
+	if (lio->linfo.link.s.link_up) {
+		for (i = 0; i < netdev->real_num_tx_queues; i++)
+			netif_start_subqueue(netdev, i);
+	}
+}
+
+static inline int skb_iq(struct octeon_device *oct, struct sk_buff *skb)
+{
+	return skb->queue_mapping % oct->num_iqs;
+}
+
+/**
+ * Remove the node at the head of the list. The list would be empty at
+ * the end of this call if there are no more nodes in the list.
+ */
+static inline struct list_head *lio_list_delete_head(struct list_head *root)
+{
+	struct list_head *node;
+
+	if (list_empty_careful(root))
+		node = NULL;
+	else
+		node = root->next;
+
+	if (node)
+		list_del(node);
+
+	return node;
+}
+
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_nic.c b/drivers/net/ethernet/cavium/liquidio/octeon_nic.c
new file mode 100644
index 000000000..1a706f81b
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_nic.c
@@ -0,0 +1,198 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more
+ * details.
+ **********************************************************************/
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_nic.h"
+#include "octeon_main.h"
+
+void *
+octeon_alloc_soft_command_resp(struct octeon_device    *oct,
+			       union octeon_instr_64B *cmd,
+			       u32		       rdatasize)
+{
+	struct octeon_soft_command *sc;
+	struct octeon_instr_ih3  *ih3;
+	struct octeon_instr_ih2  *ih2;
+	struct octeon_instr_irh *irh;
+	struct octeon_instr_rdp *rdp;
+
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct, 0, rdatasize, 0);
+
+	if (!sc)
+		return NULL;
+
+	/* Copy existing command structure into the soft command */
+	memcpy(&sc->cmd, cmd, sizeof(union octeon_instr_64B));
+
+	/* Add in the response related fields. Opcode and Param are already
+	 * there.
+	 */
+	if (OCTEON_CN23XX_PF(oct) || OCTEON_CN23XX_VF(oct)) {
+		ih3      = (struct octeon_instr_ih3 *)&sc->cmd.cmd3.ih3;
+		rdp     = (struct octeon_instr_rdp *)&sc->cmd.cmd3.rdp;
+		irh     = (struct octeon_instr_irh *)&sc->cmd.cmd3.irh;
+		/*pkiih3 + irh + ossp[0] + ossp[1] + rdp + rptr = 40 bytes */
+		ih3->fsz = LIO_SOFTCMDRESP_IH3;
+	} else {
+		ih2      = (struct octeon_instr_ih2 *)&sc->cmd.cmd2.ih2;
+		rdp     = (struct octeon_instr_rdp *)&sc->cmd.cmd2.rdp;
+		irh     = (struct octeon_instr_irh *)&sc->cmd.cmd2.irh;
+		/* irh + ossp[0] + ossp[1] + rdp + rptr = 40 bytes */
+		ih2->fsz = LIO_SOFTCMDRESP_IH2;
+	}
+
+	irh->rflag = 1; /* a response is required */
+
+	rdp->pcie_port = oct->pcie_port;
+	rdp->rlen      = rdatasize;
+
+	*sc->status_word = COMPLETION_WORD_INIT;
+
+	if (OCTEON_CN23XX_PF(oct) || OCTEON_CN23XX_VF(oct))
+		sc->cmd.cmd3.rptr =  sc->dmarptr;
+	else
+		sc->cmd.cmd2.rptr =  sc->dmarptr;
+
+	sc->expiry_time = jiffies + msecs_to_jiffies(LIO_SC_MAX_TMO_MS);
+
+	return sc;
+}
+
+int octnet_send_nic_data_pkt(struct octeon_device *oct,
+			     struct octnic_data_pkt *ndata,
+			     int xmit_more)
+{
+	int ring_doorbell = !xmit_more;
+
+	return octeon_send_command(oct, ndata->q_no, ring_doorbell, &ndata->cmd,
+				   ndata->buf, ndata->datasize,
+				   ndata->reqtype);
+}
+
+static inline struct octeon_soft_command
+*octnic_alloc_ctrl_pkt_sc(struct octeon_device *oct,
+			  struct octnic_ctrl_pkt *nctrl)
+{
+	struct octeon_soft_command *sc = NULL;
+	u8 *data;
+	u32 rdatasize;
+	u32 uddsize = 0, datasize = 0;
+
+	uddsize = (u32)(nctrl->ncmd.s.more * 8);
+
+	datasize = OCTNET_CMD_SIZE + uddsize;
+	rdatasize = 16;
+
+	sc = (struct octeon_soft_command *)
+		octeon_alloc_soft_command(oct, datasize, rdatasize, 0);
+
+	if (!sc)
+		return NULL;
+
+	data = (u8 *)sc->virtdptr;
+
+	memcpy(data, &nctrl->ncmd, OCTNET_CMD_SIZE);
+
+	octeon_swap_8B_data((u64 *)data, (OCTNET_CMD_SIZE >> 3));
+
+	if (uddsize) {
+		/* Endian-Swap for UDD should have been done by caller. */
+		memcpy(data + OCTNET_CMD_SIZE, nctrl->udd, uddsize);
+	}
+
+	sc->iq_no = (u32)nctrl->iq_no;
+
+	octeon_prepare_soft_command(oct, sc, OPCODE_NIC, OPCODE_NIC_CMD,
+				    0, 0, 0);
+
+	init_completion(&sc->complete);
+	sc->sc_status = OCTEON_REQUEST_PENDING;
+
+	return sc;
+}
+
+int
+octnet_send_nic_ctrl_pkt(struct octeon_device *oct,
+			 struct octnic_ctrl_pkt *nctrl)
+{
+	int retval;
+	struct octeon_soft_command *sc = NULL;
+
+	spin_lock_bh(&oct->cmd_resp_wqlock);
+	/* Allow only rx ctrl command to stop traffic on the chip
+	 * during offline operations
+	 */
+	if ((oct->cmd_resp_state == OCT_DRV_OFFLINE) &&
+	    (nctrl->ncmd.s.cmd != OCTNET_CMD_RX_CTL)) {
+		spin_unlock_bh(&oct->cmd_resp_wqlock);
+		dev_err(&oct->pci_dev->dev,
+			"%s cmd:%d not processed since driver offline\n",
+			__func__, nctrl->ncmd.s.cmd);
+		return -1;
+	}
+
+	sc = octnic_alloc_ctrl_pkt_sc(oct, nctrl);
+	if (!sc) {
+		dev_err(&oct->pci_dev->dev, "%s soft command alloc failed\n",
+			__func__);
+		spin_unlock_bh(&oct->cmd_resp_wqlock);
+		return -1;
+	}
+
+	retval = octeon_send_soft_command(oct, sc);
+	if (retval == IQ_SEND_FAILED) {
+		octeon_free_soft_command(oct, sc);
+		dev_err(&oct->pci_dev->dev, "%s pf_num:%d soft command:%d send failed status: %x\n",
+			__func__, oct->pf_num, nctrl->ncmd.s.cmd, retval);
+		spin_unlock_bh(&oct->cmd_resp_wqlock);
+		return -1;
+	}
+
+	spin_unlock_bh(&oct->cmd_resp_wqlock);
+
+	if (nctrl->ncmd.s.cmdgroup == 0) {
+		switch (nctrl->ncmd.s.cmd) {
+			/* caller holds lock, can not sleep */
+		case OCTNET_CMD_CHANGE_DEVFLAGS:
+		case OCTNET_CMD_SET_MULTI_LIST:
+		case OCTNET_CMD_SET_UC_LIST:
+			WRITE_ONCE(sc->caller_is_done, true);
+			return retval;
+		}
+	}
+
+	retval = wait_for_sc_completion_timeout(oct, sc, 0);
+	if (retval)
+		return (retval);
+
+	nctrl->sc_status = sc->sc_status;
+	retval = nctrl->sc_status;
+	if (nctrl->cb_fn)
+		nctrl->cb_fn(nctrl);
+
+	WRITE_ONCE(sc->caller_is_done, true);
+
+	return retval;
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/octeon_nic.h b/drivers/net/ethernet/cavium/liquidio/octeon_nic.h
new file mode 100644
index 000000000..87dd6f89c
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/octeon_nic.h
@@ -0,0 +1,288 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more
+ * details.
+ **********************************************************************/
+
+/*!  \file octeon_nic.h
+ *   \brief Host NIC Driver: Routine to send network data &
+ *   control packet to Octeon.
+ */
+
+#ifndef __OCTEON_NIC_H__
+#define  __OCTEON_NIC_H__
+
+/* Maximum number of 8-byte words can be sent in a NIC control message.
+ */
+#define  MAX_NCTRL_UDD  32
+
+typedef void (*octnic_ctrl_pkt_cb_fn_t) (void *);
+
+/* Structure of control information passed by the NIC module to the OSI
+ * layer when sending control commands to Octeon device software.
+ */
+struct octnic_ctrl_pkt {
+	/** Command to be passed to the Octeon device software. */
+	union octnet_cmd ncmd;
+
+	/** Send buffer  */
+	void *data;
+	u64 dmadata;
+
+	/** Response buffer */
+	void *rdata;
+	u64 dmardata;
+
+	/** Additional data that may be needed by some commands. */
+	u64 udd[MAX_NCTRL_UDD];
+
+	/** Input queue to use to send this command. */
+	u64 iq_no;
+
+	/** The network device that issued the control command. */
+	u64 netpndev;
+
+	/** Callback function called when the command has been fetched */
+	octnic_ctrl_pkt_cb_fn_t cb_fn;
+
+	u32 sc_status;
+};
+
+#define MAX_UDD_SIZE(nctrl) (sizeof((nctrl)->udd))
+
+/** Structure of data information passed by the NIC module to the OSI
+ * layer when forwarding data to Octeon device software.
+ */
+struct octnic_data_pkt {
+	/** Pointer to information maintained by NIC module for this packet. The
+	 *  OSI layer passes this as-is to the driver.
+	 */
+	void *buf;
+
+	/** Type of buffer passed in "buf" above. */
+	u32 reqtype;
+
+	/** Total data bytes to be transferred in this command. */
+	u32 datasize;
+
+	/** Command to be passed to the Octeon device software. */
+	union octeon_instr_64B cmd;
+
+	/** Input queue to use to send this command. */
+	u32 q_no;
+
+};
+
+/** Structure passed by NIC module to OSI layer to prepare a command to send
+ * network data to Octeon.
+ */
+union octnic_cmd_setup {
+	struct {
+		u32 iq_no:8;
+		u32 gather:1;
+		u32 timestamp:1;
+		u32 ip_csum:1;
+		u32 transport_csum:1;
+		u32 tnl_csum:1;
+		u32 rsvd:19;
+
+		union {
+			u32 datasize;
+			u32 gatherptrs;
+		} u;
+	} s;
+
+	u64 u64;
+
+};
+
+static inline int octnet_iq_is_full(struct octeon_device *oct, u32 q_no)
+{
+	return ((u32)atomic_read(&oct->instr_queue[q_no]->instr_pending)
+		>= (oct->instr_queue[q_no]->max_count - 2));
+}
+
+static inline void
+octnet_prepare_pci_cmd_o2(struct octeon_device *oct,
+			  union octeon_instr_64B *cmd,
+			  union octnic_cmd_setup *setup, u32 tag)
+{
+	struct octeon_instr_ih2 *ih2;
+	struct octeon_instr_irh *irh;
+	union octnic_packet_params packet_params;
+	int port;
+
+	memset(cmd, 0, sizeof(union octeon_instr_64B));
+
+	ih2 = (struct octeon_instr_ih2 *)&cmd->cmd2.ih2;
+
+	/* assume that rflag is cleared so therefore front data will only have
+	 * irh and ossp[0], ossp[1] for a total of 32 bytes
+	 */
+	ih2->fsz = LIO_PCICMD_O2;
+
+	ih2->tagtype = ORDERED_TAG;
+	ih2->grp = DEFAULT_POW_GRP;
+
+	port = (int)oct->instr_queue[setup->s.iq_no]->txpciq.s.port;
+
+	if (tag)
+		ih2->tag = tag;
+	else
+		ih2->tag = LIO_DATA(port);
+
+	ih2->raw = 1;
+	ih2->qos = (port & 3) + 4;	/* map qos based on interface */
+
+	if (!setup->s.gather) {
+		ih2->dlengsz = setup->s.u.datasize;
+	} else {
+		ih2->gather = 1;
+		ih2->dlengsz = setup->s.u.gatherptrs;
+	}
+
+	irh = (struct octeon_instr_irh *)&cmd->cmd2.irh;
+
+	irh->opcode = OPCODE_NIC;
+	irh->subcode = OPCODE_NIC_NW_DATA;
+
+	packet_params.u32 = 0;
+
+	packet_params.s.ip_csum = setup->s.ip_csum;
+	packet_params.s.transport_csum = setup->s.transport_csum;
+	packet_params.s.tnl_csum = setup->s.tnl_csum;
+	packet_params.s.tsflag = setup->s.timestamp;
+
+	irh->ossp = packet_params.u32;
+}
+
+static inline void
+octnet_prepare_pci_cmd_o3(struct octeon_device *oct,
+			  union octeon_instr_64B *cmd,
+			  union octnic_cmd_setup *setup, u32 tag)
+{
+	struct octeon_instr_irh *irh;
+	struct octeon_instr_ih3     *ih3;
+	struct octeon_instr_pki_ih3 *pki_ih3;
+	union octnic_packet_params packet_params;
+	int port;
+
+	memset(cmd, 0, sizeof(union octeon_instr_64B));
+
+	ih3 = (struct octeon_instr_ih3 *)&cmd->cmd3.ih3;
+	pki_ih3 = (struct octeon_instr_pki_ih3 *)&cmd->cmd3.pki_ih3;
+
+	/* assume that rflag is cleared so therefore front data will only have
+	 * irh and ossp[1] and ossp[2] for a total of 24 bytes
+	 */
+	ih3->pkind       = oct->instr_queue[setup->s.iq_no]->txpciq.s.pkind;
+	/*PKI IH*/
+	ih3->fsz = LIO_PCICMD_O3;
+
+	if (!setup->s.gather) {
+		ih3->dlengsz = setup->s.u.datasize;
+	} else {
+		ih3->gather = 1;
+		ih3->dlengsz = setup->s.u.gatherptrs;
+	}
+
+	pki_ih3->w       = 1;
+	pki_ih3->raw     = 1;
+	pki_ih3->utag    = 1;
+	pki_ih3->utt     = 1;
+	pki_ih3->uqpg    = oct->instr_queue[setup->s.iq_no]->txpciq.s.use_qpg;
+
+	port = (int)oct->instr_queue[setup->s.iq_no]->txpciq.s.port;
+
+	if (tag)
+		pki_ih3->tag = tag;
+	else
+		pki_ih3->tag     = LIO_DATA(port);
+
+	pki_ih3->tagtype = ORDERED_TAG;
+	pki_ih3->qpg     = oct->instr_queue[setup->s.iq_no]->txpciq.s.qpg;
+	pki_ih3->pm      = 0x7; /*0x7 - meant for Parse nothing, uninterpreted*/
+	pki_ih3->sl      = 8;   /* sl will be sizeof(pki_ih3)*/
+
+	irh = (struct octeon_instr_irh *)&cmd->cmd3.irh;
+
+	irh->opcode = OPCODE_NIC;
+	irh->subcode = OPCODE_NIC_NW_DATA;
+
+	packet_params.u32 = 0;
+
+	packet_params.s.ip_csum = setup->s.ip_csum;
+	packet_params.s.transport_csum = setup->s.transport_csum;
+	packet_params.s.tnl_csum = setup->s.tnl_csum;
+	packet_params.s.tsflag = setup->s.timestamp;
+
+	irh->ossp = packet_params.u32;
+}
+
+/** Utility function to prepare a 64B NIC instruction based on a setup command
+ * @param cmd - pointer to instruction to be filled in.
+ * @param setup - pointer to the setup structure
+ * @param q_no - which queue for back pressure
+ *
+ * Assumes the cmd instruction is pre-allocated, but no fields are filled in.
+ */
+static inline void
+octnet_prepare_pci_cmd(struct octeon_device *oct, union octeon_instr_64B *cmd,
+		       union octnic_cmd_setup *setup, u32 tag)
+{
+	if (OCTEON_CN6XXX(oct))
+		octnet_prepare_pci_cmd_o2(oct, cmd, setup, tag);
+	else
+		octnet_prepare_pci_cmd_o3(oct, cmd, setup, tag);
+}
+
+/** Allocate and a soft command with space for a response immediately following
+ * the commnad.
+ * @param oct - octeon device pointer
+ * @param cmd - pointer to the command structure, pre-filled for everything
+ * except the response.
+ * @param rdatasize - size in bytes of the response.
+ *
+ * @returns pointer to allocated buffer with command copied into it, and
+ * response space immediately following.
+ */
+void *
+octeon_alloc_soft_command_resp(struct octeon_device    *oct,
+			       union octeon_instr_64B *cmd,
+			       u32		       rdatasize);
+
+/** Send a NIC data packet to the device
+ * @param oct - octeon device pointer
+ * @param ndata - control structure with queueing, and buffer information
+ *
+ * @returns IQ_FAILED if it failed to add to the input queue. IQ_STOP if it the
+ * queue should be stopped, and IQ_SEND_OK if it sent okay.
+ */
+int octnet_send_nic_data_pkt(struct octeon_device *oct,
+			     struct octnic_data_pkt *ndata,
+			     int xmit_more);
+
+/** Send a NIC control packet to the device
+ * @param oct - octeon device pointer
+ * @param nctrl - control structure with command, timout, and callback info
+ * @returns IQ_FAILED if it failed to add to the input queue. IQ_STOP if it the
+ * queue should be stopped, and IQ_SEND_OK if it sent okay.
+ */
+int
+octnet_send_nic_ctrl_pkt(struct octeon_device *oct,
+			 struct octnic_ctrl_pkt *nctrl);
+
+#endif
diff --git a/drivers/net/ethernet/cavium/liquidio/request_manager.c b/drivers/net/ethernet/cavium/liquidio/request_manager.c
new file mode 100644
index 000000000..8e59c2825
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/request_manager.c
@@ -0,0 +1,936 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more
+ * details.
+ **********************************************************************/
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include <linux/vmalloc.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_main.h"
+#include "octeon_network.h"
+#include "cn66xx_device.h"
+#include "cn23xx_pf_device.h"
+#include "cn23xx_vf_device.h"
+
+struct iq_post_status {
+	int status;
+	int index;
+};
+
+static void check_db_timeout(struct work_struct *work);
+static void  __check_db_timeout(struct octeon_device *oct, u64 iq_no);
+
+static void (*reqtype_free_fn[MAX_OCTEON_DEVICES][REQTYPE_LAST + 1]) (void *);
+
+static inline int IQ_INSTR_MODE_64B(struct octeon_device *oct, int iq_no)
+{
+	struct octeon_instr_queue *iq =
+	    (struct octeon_instr_queue *)oct->instr_queue[iq_no];
+	return iq->iqcmd_64B;
+}
+
+#define IQ_INSTR_MODE_32B(oct, iq_no)  (!IQ_INSTR_MODE_64B(oct, iq_no))
+
+/* Define this to return the request status comaptible to old code */
+/*#define OCTEON_USE_OLD_REQ_STATUS*/
+
+/* Return 0 on success, 1 on failure */
+int octeon_init_instr_queue(struct octeon_device *oct,
+			    union oct_txpciq txpciq,
+			    u32 num_descs)
+{
+	struct octeon_instr_queue *iq;
+	struct octeon_iq_config *conf = NULL;
+	u32 iq_no = (u32)txpciq.s.q_no;
+	u32 q_size;
+	struct cavium_wq *db_wq;
+	int numa_node = dev_to_node(&oct->pci_dev->dev);
+
+	if (OCTEON_CN6XXX(oct))
+		conf = &(CFG_GET_IQ_CFG(CHIP_CONF(oct, cn6xxx)));
+	else if (OCTEON_CN23XX_PF(oct))
+		conf = &(CFG_GET_IQ_CFG(CHIP_CONF(oct, cn23xx_pf)));
+	else if (OCTEON_CN23XX_VF(oct))
+		conf = &(CFG_GET_IQ_CFG(CHIP_CONF(oct, cn23xx_vf)));
+
+	if (!conf) {
+		dev_err(&oct->pci_dev->dev, "Unsupported Chip %x\n",
+			oct->chip_id);
+		return 1;
+	}
+
+	q_size = (u32)conf->instr_type * num_descs;
+
+	iq = oct->instr_queue[iq_no];
+
+	iq->oct_dev = oct;
+
+	iq->base_addr = lio_dma_alloc(oct, q_size, &iq->base_addr_dma);
+	if (!iq->base_addr) {
+		dev_err(&oct->pci_dev->dev, "Cannot allocate memory for instr queue %d\n",
+			iq_no);
+		return 1;
+	}
+
+	iq->max_count = num_descs;
+
+	/* Initialize a list to holds requests that have been posted to Octeon
+	 * but has yet to be fetched by octeon
+	 */
+	iq->request_list = vzalloc_node(array_size(num_descs, sizeof(*iq->request_list)),
+					numa_node);
+	if (!iq->request_list)
+		iq->request_list = vzalloc(array_size(num_descs, sizeof(*iq->request_list)));
+	if (!iq->request_list) {
+		lio_dma_free(oct, q_size, iq->base_addr, iq->base_addr_dma);
+		dev_err(&oct->pci_dev->dev, "Alloc failed for IQ[%d] nr free list\n",
+			iq_no);
+		return 1;
+	}
+
+	dev_dbg(&oct->pci_dev->dev, "IQ[%d]: base: %p basedma: %pad count: %d\n",
+		iq_no, iq->base_addr, &iq->base_addr_dma, iq->max_count);
+
+	iq->txpciq.u64 = txpciq.u64;
+	iq->fill_threshold = (u32)conf->db_min;
+	iq->fill_cnt = 0;
+	iq->host_write_index = 0;
+	iq->octeon_read_index = 0;
+	iq->flush_index = 0;
+	iq->last_db_time = 0;
+	iq->do_auto_flush = 1;
+	iq->db_timeout = (u32)conf->db_timeout;
+	atomic_set(&iq->instr_pending, 0);
+	iq->pkts_processed = 0;
+
+	/* Initialize the spinlock for this instruction queue */
+	spin_lock_init(&iq->lock);
+	if (iq_no == 0) {
+		iq->allow_soft_cmds = true;
+		spin_lock_init(&iq->post_lock);
+	} else {
+		iq->allow_soft_cmds = false;
+	}
+
+	spin_lock_init(&iq->iq_flush_running_lock);
+
+	oct->io_qmask.iq |= BIT_ULL(iq_no);
+
+	/* Set the 32B/64B mode for each input queue */
+	oct->io_qmask.iq64B |= ((conf->instr_type == 64) << iq_no);
+	iq->iqcmd_64B = (conf->instr_type == 64);
+
+	oct->fn_list.setup_iq_regs(oct, iq_no);
+
+	oct->check_db_wq[iq_no].wq = alloc_workqueue("check_iq_db",
+						     WQ_MEM_RECLAIM,
+						     0);
+	if (!oct->check_db_wq[iq_no].wq) {
+		vfree(iq->request_list);
+		iq->request_list = NULL;
+		lio_dma_free(oct, q_size, iq->base_addr, iq->base_addr_dma);
+		dev_err(&oct->pci_dev->dev, "check db wq create failed for iq %d\n",
+			iq_no);
+		return 1;
+	}
+
+	db_wq = &oct->check_db_wq[iq_no];
+
+	INIT_DELAYED_WORK(&db_wq->wk.work, check_db_timeout);
+	db_wq->wk.ctxptr = oct;
+	db_wq->wk.ctxul = iq_no;
+	queue_delayed_work(db_wq->wq, &db_wq->wk.work, msecs_to_jiffies(1));
+
+	return 0;
+}
+
+int octeon_delete_instr_queue(struct octeon_device *oct, u32 iq_no)
+{
+	u64 desc_size = 0, q_size;
+	struct octeon_instr_queue *iq = oct->instr_queue[iq_no];
+
+	cancel_delayed_work_sync(&oct->check_db_wq[iq_no].wk.work);
+	destroy_workqueue(oct->check_db_wq[iq_no].wq);
+
+	if (OCTEON_CN6XXX(oct))
+		desc_size =
+		    CFG_GET_IQ_INSTR_TYPE(CHIP_CONF(oct, cn6xxx));
+	else if (OCTEON_CN23XX_PF(oct))
+		desc_size =
+		    CFG_GET_IQ_INSTR_TYPE(CHIP_CONF(oct, cn23xx_pf));
+	else if (OCTEON_CN23XX_VF(oct))
+		desc_size =
+		    CFG_GET_IQ_INSTR_TYPE(CHIP_CONF(oct, cn23xx_vf));
+
+	vfree(iq->request_list);
+
+	if (iq->base_addr) {
+		q_size = iq->max_count * desc_size;
+		lio_dma_free(oct, (u32)q_size, iq->base_addr,
+			     iq->base_addr_dma);
+		oct->io_qmask.iq &= ~(1ULL << iq_no);
+		vfree(oct->instr_queue[iq_no]);
+		oct->instr_queue[iq_no] = NULL;
+		oct->num_iqs--;
+		return 0;
+	}
+	return 1;
+}
+
+/* Return 0 on success, 1 on failure */
+int octeon_setup_iq(struct octeon_device *oct,
+		    int ifidx,
+		    int q_index,
+		    union oct_txpciq txpciq,
+		    u32 num_descs,
+		    void *app_ctx)
+{
+	u32 iq_no = (u32)txpciq.s.q_no;
+	int numa_node = dev_to_node(&oct->pci_dev->dev);
+
+	if (oct->instr_queue[iq_no]) {
+		dev_dbg(&oct->pci_dev->dev, "IQ is in use. Cannot create the IQ: %d again\n",
+			iq_no);
+		oct->instr_queue[iq_no]->txpciq.u64 = txpciq.u64;
+		oct->instr_queue[iq_no]->app_ctx = app_ctx;
+		return 0;
+	}
+	oct->instr_queue[iq_no] =
+	    vzalloc_node(sizeof(struct octeon_instr_queue), numa_node);
+	if (!oct->instr_queue[iq_no])
+		oct->instr_queue[iq_no] =
+		    vzalloc(sizeof(struct octeon_instr_queue));
+	if (!oct->instr_queue[iq_no])
+		return 1;
+
+
+	oct->instr_queue[iq_no]->q_index = q_index;
+	oct->instr_queue[iq_no]->app_ctx = app_ctx;
+	oct->instr_queue[iq_no]->ifidx = ifidx;
+
+	if (octeon_init_instr_queue(oct, txpciq, num_descs)) {
+		vfree(oct->instr_queue[iq_no]);
+		oct->instr_queue[iq_no] = NULL;
+		return 1;
+	}
+
+	oct->num_iqs++;
+	if (oct->fn_list.enable_io_queues(oct)) {
+		octeon_delete_instr_queue(oct, iq_no);
+		return 1;
+	}
+
+	return 0;
+}
+
+int lio_wait_for_instr_fetch(struct octeon_device *oct)
+{
+	int i, retry = 1000, pending, instr_cnt = 0;
+
+	do {
+		instr_cnt = 0;
+
+		for (i = 0; i < MAX_OCTEON_INSTR_QUEUES(oct); i++) {
+			if (!(oct->io_qmask.iq & BIT_ULL(i)))
+				continue;
+			pending =
+			    atomic_read(&oct->instr_queue[i]->instr_pending);
+			if (pending)
+				__check_db_timeout(oct, i);
+			instr_cnt += pending;
+		}
+
+		if (instr_cnt == 0)
+			break;
+
+		schedule_timeout_uninterruptible(1);
+
+	} while (retry-- && instr_cnt);
+
+	return instr_cnt;
+}
+
+static inline void
+ring_doorbell(struct octeon_device *oct, struct octeon_instr_queue *iq)
+{
+	if (atomic_read(&oct->status) == OCT_DEV_RUNNING) {
+		writel(iq->fill_cnt, iq->doorbell_reg);
+		/* make sure doorbell write goes through */
+		iq->fill_cnt = 0;
+		iq->last_db_time = jiffies;
+		return;
+	}
+}
+
+void
+octeon_ring_doorbell_locked(struct octeon_device *oct, u32 iq_no)
+{
+	struct octeon_instr_queue *iq;
+
+	iq = oct->instr_queue[iq_no];
+	spin_lock(&iq->post_lock);
+	if (iq->fill_cnt)
+		ring_doorbell(oct, iq);
+	spin_unlock(&iq->post_lock);
+}
+
+static inline void __copy_cmd_into_iq(struct octeon_instr_queue *iq,
+				      u8 *cmd)
+{
+	u8 *iqptr, cmdsize;
+
+	cmdsize = ((iq->iqcmd_64B) ? 64 : 32);
+	iqptr = iq->base_addr + (cmdsize * iq->host_write_index);
+
+	memcpy(iqptr, cmd, cmdsize);
+}
+
+static inline struct iq_post_status
+__post_command2(struct octeon_instr_queue *iq, u8 *cmd)
+{
+	struct iq_post_status st;
+
+	st.status = IQ_SEND_OK;
+
+	/* This ensures that the read index does not wrap around to the same
+	 * position if queue gets full before Octeon could fetch any instr.
+	 */
+	if (atomic_read(&iq->instr_pending) >= (s32)(iq->max_count - 1)) {
+		st.status = IQ_SEND_FAILED;
+		st.index = -1;
+		return st;
+	}
+
+	if (atomic_read(&iq->instr_pending) >= (s32)(iq->max_count - 2))
+		st.status = IQ_SEND_STOP;
+
+	__copy_cmd_into_iq(iq, cmd);
+
+	/* "index" is returned, host_write_index is modified. */
+	st.index = iq->host_write_index;
+	iq->host_write_index = incr_index(iq->host_write_index, 1,
+					  iq->max_count);
+	iq->fill_cnt++;
+
+	/* Flush the command into memory. We need to be sure the data is in
+	 * memory before indicating that the instruction is pending.
+	 */
+	wmb();
+
+	atomic_inc(&iq->instr_pending);
+
+	return st;
+}
+
+int
+octeon_register_reqtype_free_fn(struct octeon_device *oct, int reqtype,
+				void (*fn)(void *))
+{
+	if (reqtype > REQTYPE_LAST) {
+		dev_err(&oct->pci_dev->dev, "%s: Invalid reqtype: %d\n",
+			__func__, reqtype);
+		return -EINVAL;
+	}
+
+	reqtype_free_fn[oct->octeon_id][reqtype] = fn;
+
+	return 0;
+}
+
+static inline void
+__add_to_request_list(struct octeon_instr_queue *iq,
+		      int idx, void *buf, int reqtype)
+{
+	iq->request_list[idx].buf = buf;
+	iq->request_list[idx].reqtype = reqtype;
+}
+
+/* Can only run in process context */
+int
+lio_process_iq_request_list(struct octeon_device *oct,
+			    struct octeon_instr_queue *iq, u32 napi_budget)
+{
+	struct cavium_wq *cwq = &oct->dma_comp_wq;
+	int reqtype;
+	void *buf;
+	u32 old = iq->flush_index;
+	u32 inst_count = 0;
+	unsigned int pkts_compl = 0, bytes_compl = 0;
+	struct octeon_soft_command *sc;
+	unsigned long flags;
+
+	while (old != iq->octeon_read_index) {
+		reqtype = iq->request_list[old].reqtype;
+		buf     = iq->request_list[old].buf;
+
+		if (reqtype == REQTYPE_NONE)
+			goto skip_this;
+
+		octeon_update_tx_completion_counters(buf, reqtype, &pkts_compl,
+						     &bytes_compl);
+
+		switch (reqtype) {
+		case REQTYPE_NORESP_NET:
+		case REQTYPE_NORESP_NET_SG:
+		case REQTYPE_RESP_NET_SG:
+			reqtype_free_fn[oct->octeon_id][reqtype](buf);
+			break;
+		case REQTYPE_RESP_NET:
+		case REQTYPE_SOFT_COMMAND:
+			sc = buf;
+			/* We're expecting a response from Octeon.
+			 * It's up to lio_process_ordered_list() to
+			 * process  sc. Add sc to the ordered soft
+			 * command response list because we expect
+			 * a response from Octeon.
+			 */
+			spin_lock_irqsave(&oct->response_list
+					  [OCTEON_ORDERED_SC_LIST].lock, flags);
+			atomic_inc(&oct->response_list
+				   [OCTEON_ORDERED_SC_LIST].pending_req_count);
+			list_add_tail(&sc->node, &oct->response_list
+				[OCTEON_ORDERED_SC_LIST].head);
+			spin_unlock_irqrestore(&oct->response_list
+					       [OCTEON_ORDERED_SC_LIST].lock,
+					       flags);
+			break;
+		default:
+			dev_err(&oct->pci_dev->dev,
+				"%s Unknown reqtype: %d buf: %p at idx %d\n",
+				__func__, reqtype, buf, old);
+		}
+
+		iq->request_list[old].buf = NULL;
+		iq->request_list[old].reqtype = 0;
+
+ skip_this:
+		inst_count++;
+		old = incr_index(old, 1, iq->max_count);
+
+		if ((napi_budget) && (inst_count >= napi_budget))
+			break;
+	}
+	if (bytes_compl)
+		octeon_report_tx_completion_to_bql(iq->app_ctx, pkts_compl,
+						   bytes_compl);
+	iq->flush_index = old;
+
+	if (atomic_read(&oct->response_list
+			[OCTEON_ORDERED_SC_LIST].pending_req_count))
+		queue_work(cwq->wq, &cwq->wk.work.work);
+
+	return inst_count;
+}
+
+/* Can only be called from process context */
+int
+octeon_flush_iq(struct octeon_device *oct, struct octeon_instr_queue *iq,
+		u32 napi_budget)
+{
+	u32 inst_processed = 0;
+	u32 tot_inst_processed = 0;
+	int tx_done = 1;
+
+	if (!spin_trylock(&iq->iq_flush_running_lock))
+		return tx_done;
+
+	spin_lock_bh(&iq->lock);
+
+	iq->octeon_read_index = oct->fn_list.update_iq_read_idx(iq);
+
+	do {
+		/* Process any outstanding IQ packets. */
+		if (iq->flush_index == iq->octeon_read_index)
+			break;
+
+		if (napi_budget)
+			inst_processed =
+				lio_process_iq_request_list(oct, iq,
+							    napi_budget -
+							    tot_inst_processed);
+		else
+			inst_processed =
+				lio_process_iq_request_list(oct, iq, 0);
+
+		if (inst_processed) {
+			iq->pkts_processed += inst_processed;
+			atomic_sub(inst_processed, &iq->instr_pending);
+			iq->stats.instr_processed += inst_processed;
+		}
+
+		tot_inst_processed += inst_processed;
+	} while (tot_inst_processed < napi_budget);
+
+	if (napi_budget && (tot_inst_processed >= napi_budget))
+		tx_done = 0;
+
+	iq->last_db_time = jiffies;
+
+	spin_unlock_bh(&iq->lock);
+
+	spin_unlock(&iq->iq_flush_running_lock);
+
+	return tx_done;
+}
+
+/* Process instruction queue after timeout.
+ * This routine gets called from a workqueue or when removing the module.
+ */
+static void __check_db_timeout(struct octeon_device *oct, u64 iq_no)
+{
+	struct octeon_instr_queue *iq;
+	u64 next_time;
+
+	if (!oct)
+		return;
+
+	iq = oct->instr_queue[iq_no];
+	if (!iq)
+		return;
+
+	/* return immediately, if no work pending */
+	if (!atomic_read(&iq->instr_pending))
+		return;
+	/* If jiffies - last_db_time < db_timeout do nothing  */
+	next_time = iq->last_db_time + iq->db_timeout;
+	if (!time_after(jiffies, (unsigned long)next_time))
+		return;
+	iq->last_db_time = jiffies;
+
+	/* Flush the instruction queue */
+	octeon_flush_iq(oct, iq, 0);
+
+	lio_enable_irq(NULL, iq);
+}
+
+/* Called by the Poll thread at regular intervals to check the instruction
+ * queue for commands to be posted and for commands that were fetched by Octeon.
+ */
+static void check_db_timeout(struct work_struct *work)
+{
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct octeon_device *oct = (struct octeon_device *)wk->ctxptr;
+	u64 iq_no = wk->ctxul;
+	struct cavium_wq *db_wq = &oct->check_db_wq[iq_no];
+	u32 delay = 10;
+
+	__check_db_timeout(oct, iq_no);
+	queue_delayed_work(db_wq->wq, &db_wq->wk.work, msecs_to_jiffies(delay));
+}
+
+int
+octeon_send_command(struct octeon_device *oct, u32 iq_no,
+		    u32 force_db, void *cmd, void *buf,
+		    u32 datasize, u32 reqtype)
+{
+	int xmit_stopped;
+	struct iq_post_status st;
+	struct octeon_instr_queue *iq = oct->instr_queue[iq_no];
+
+	/* Get the lock and prevent other tasks and tx interrupt handler from
+	 * running.
+	 */
+	if (iq->allow_soft_cmds)
+		spin_lock_bh(&iq->post_lock);
+
+	st = __post_command2(iq, cmd);
+
+	if (st.status != IQ_SEND_FAILED) {
+		xmit_stopped = octeon_report_sent_bytes_to_bql(buf, reqtype);
+		__add_to_request_list(iq, st.index, buf, reqtype);
+		INCR_INSTRQUEUE_PKT_COUNT(oct, iq_no, bytes_sent, datasize);
+		INCR_INSTRQUEUE_PKT_COUNT(oct, iq_no, instr_posted, 1);
+
+		if (iq->fill_cnt >= MAX_OCTEON_FILL_COUNT || force_db ||
+		    xmit_stopped || st.status == IQ_SEND_STOP)
+			ring_doorbell(oct, iq);
+	} else {
+		INCR_INSTRQUEUE_PKT_COUNT(oct, iq_no, instr_dropped, 1);
+	}
+
+	if (iq->allow_soft_cmds)
+		spin_unlock_bh(&iq->post_lock);
+
+	/* This is only done here to expedite packets being flushed
+	 * for cases where there are no IQ completion interrupts.
+	 */
+
+	return st.status;
+}
+
+void
+octeon_prepare_soft_command(struct octeon_device *oct,
+			    struct octeon_soft_command *sc,
+			    u8 opcode,
+			    u8 subcode,
+			    u32 irh_ossp,
+			    u64 ossp0,
+			    u64 ossp1)
+{
+	struct octeon_config *oct_cfg;
+	struct octeon_instr_ih2 *ih2;
+	struct octeon_instr_ih3 *ih3;
+	struct octeon_instr_pki_ih3 *pki_ih3;
+	struct octeon_instr_irh *irh;
+	struct octeon_instr_rdp *rdp;
+
+	WARN_ON(opcode > 15);
+	WARN_ON(subcode > 127);
+
+	oct_cfg = octeon_get_conf(oct);
+
+	if (OCTEON_CN23XX_PF(oct) || OCTEON_CN23XX_VF(oct)) {
+		ih3 = (struct octeon_instr_ih3 *)&sc->cmd.cmd3.ih3;
+
+		ih3->pkind = oct->instr_queue[sc->iq_no]->txpciq.s.pkind;
+
+		pki_ih3 = (struct octeon_instr_pki_ih3 *)&sc->cmd.cmd3.pki_ih3;
+
+		pki_ih3->w           = 1;
+		pki_ih3->raw         = 1;
+		pki_ih3->utag        = 1;
+		pki_ih3->uqpg        =
+			oct->instr_queue[sc->iq_no]->txpciq.s.use_qpg;
+		pki_ih3->utt         = 1;
+		pki_ih3->tag     = LIO_CONTROL;
+		pki_ih3->tagtype = ATOMIC_TAG;
+		pki_ih3->qpg         =
+			oct->instr_queue[sc->iq_no]->txpciq.s.ctrl_qpg;
+
+		pki_ih3->pm          = 0x7;
+		pki_ih3->sl          = 8;
+
+		if (sc->datasize)
+			ih3->dlengsz = sc->datasize;
+
+		irh            = (struct octeon_instr_irh *)&sc->cmd.cmd3.irh;
+		irh->opcode    = opcode;
+		irh->subcode   = subcode;
+
+		/* opcode/subcode specific parameters (ossp) */
+		irh->ossp       = irh_ossp;
+		sc->cmd.cmd3.ossp[0] = ossp0;
+		sc->cmd.cmd3.ossp[1] = ossp1;
+
+		if (sc->rdatasize) {
+			rdp = (struct octeon_instr_rdp *)&sc->cmd.cmd3.rdp;
+			rdp->pcie_port = oct->pcie_port;
+			rdp->rlen      = sc->rdatasize;
+
+			irh->rflag =  1;
+			/*PKI IH3*/
+			/* pki_ih3 irh+ossp[0]+ossp[1]+rdp+rptr = 48 bytes */
+			ih3->fsz    = LIO_SOFTCMDRESP_IH3;
+		} else {
+			irh->rflag =  0;
+			/*PKI IH3*/
+			/* pki_h3 + irh + ossp[0] + ossp[1] = 32 bytes */
+			ih3->fsz    = LIO_PCICMD_O3;
+		}
+
+	} else {
+		ih2          = (struct octeon_instr_ih2 *)&sc->cmd.cmd2.ih2;
+		ih2->tagtype = ATOMIC_TAG;
+		ih2->tag     = LIO_CONTROL;
+		ih2->raw     = 1;
+		ih2->grp     = CFG_GET_CTRL_Q_GRP(oct_cfg);
+
+		if (sc->datasize) {
+			ih2->dlengsz = sc->datasize;
+			ih2->rs = 1;
+		}
+
+		irh            = (struct octeon_instr_irh *)&sc->cmd.cmd2.irh;
+		irh->opcode    = opcode;
+		irh->subcode   = subcode;
+
+		/* opcode/subcode specific parameters (ossp) */
+		irh->ossp       = irh_ossp;
+		sc->cmd.cmd2.ossp[0] = ossp0;
+		sc->cmd.cmd2.ossp[1] = ossp1;
+
+		if (sc->rdatasize) {
+			rdp = (struct octeon_instr_rdp *)&sc->cmd.cmd2.rdp;
+			rdp->pcie_port = oct->pcie_port;
+			rdp->rlen      = sc->rdatasize;
+
+			irh->rflag =  1;
+			/* irh+ossp[0]+ossp[1]+rdp+rptr = 40 bytes */
+			ih2->fsz   = LIO_SOFTCMDRESP_IH2;
+		} else {
+			irh->rflag =  0;
+			/* irh + ossp[0] + ossp[1] = 24 bytes */
+			ih2->fsz   = LIO_PCICMD_O2;
+		}
+	}
+}
+
+int octeon_send_soft_command(struct octeon_device *oct,
+			     struct octeon_soft_command *sc)
+{
+	struct octeon_instr_queue *iq;
+	struct octeon_instr_ih2 *ih2;
+	struct octeon_instr_ih3 *ih3;
+	struct octeon_instr_irh *irh;
+	u32 len;
+
+	iq = oct->instr_queue[sc->iq_no];
+	if (!iq->allow_soft_cmds) {
+		dev_err(&oct->pci_dev->dev, "Soft commands are not allowed on Queue %d\n",
+			sc->iq_no);
+		INCR_INSTRQUEUE_PKT_COUNT(oct, sc->iq_no, instr_dropped, 1);
+		return IQ_SEND_FAILED;
+	}
+
+	if (OCTEON_CN23XX_PF(oct) || OCTEON_CN23XX_VF(oct)) {
+		ih3 =  (struct octeon_instr_ih3 *)&sc->cmd.cmd3.ih3;
+		if (ih3->dlengsz) {
+			WARN_ON(!sc->dmadptr);
+			sc->cmd.cmd3.dptr = sc->dmadptr;
+		}
+		irh = (struct octeon_instr_irh *)&sc->cmd.cmd3.irh;
+		if (irh->rflag) {
+			WARN_ON(!sc->dmarptr);
+			WARN_ON(!sc->status_word);
+			*sc->status_word = COMPLETION_WORD_INIT;
+			sc->cmd.cmd3.rptr = sc->dmarptr;
+		}
+		len = (u32)ih3->dlengsz;
+	} else {
+		ih2 = (struct octeon_instr_ih2 *)&sc->cmd.cmd2.ih2;
+		if (ih2->dlengsz) {
+			WARN_ON(!sc->dmadptr);
+			sc->cmd.cmd2.dptr = sc->dmadptr;
+		}
+		irh = (struct octeon_instr_irh *)&sc->cmd.cmd2.irh;
+		if (irh->rflag) {
+			WARN_ON(!sc->dmarptr);
+			WARN_ON(!sc->status_word);
+			*sc->status_word = COMPLETION_WORD_INIT;
+			sc->cmd.cmd2.rptr = sc->dmarptr;
+		}
+		len = (u32)ih2->dlengsz;
+	}
+
+	sc->expiry_time = jiffies + msecs_to_jiffies(LIO_SC_MAX_TMO_MS);
+
+	return (octeon_send_command(oct, sc->iq_no, 1, &sc->cmd, sc,
+				    len, REQTYPE_SOFT_COMMAND));
+}
+
+int octeon_setup_sc_buffer_pool(struct octeon_device *oct)
+{
+	int i;
+	u64 dma_addr;
+	struct octeon_soft_command *sc;
+
+	INIT_LIST_HEAD(&oct->sc_buf_pool.head);
+	spin_lock_init(&oct->sc_buf_pool.lock);
+	atomic_set(&oct->sc_buf_pool.alloc_buf_count, 0);
+
+	for (i = 0; i < MAX_SOFT_COMMAND_BUFFERS; i++) {
+		sc = (struct octeon_soft_command *)
+			lio_dma_alloc(oct,
+				      SOFT_COMMAND_BUFFER_SIZE,
+					  (dma_addr_t *)&dma_addr);
+		if (!sc) {
+			octeon_free_sc_buffer_pool(oct);
+			return 1;
+		}
+
+		sc->dma_addr = dma_addr;
+		sc->size = SOFT_COMMAND_BUFFER_SIZE;
+
+		list_add_tail(&sc->node, &oct->sc_buf_pool.head);
+	}
+
+	return 0;
+}
+
+int octeon_free_sc_done_list(struct octeon_device *oct)
+{
+	struct octeon_response_list *done_sc_list, *zombie_sc_list;
+	struct octeon_soft_command *sc;
+	struct list_head *tmp, *tmp2;
+	spinlock_t *sc_lists_lock; /* lock for response_list */
+
+	done_sc_list = &oct->response_list[OCTEON_DONE_SC_LIST];
+	zombie_sc_list = &oct->response_list[OCTEON_ZOMBIE_SC_LIST];
+
+	if (!atomic_read(&done_sc_list->pending_req_count))
+		return 0;
+
+	sc_lists_lock = &oct->response_list[OCTEON_ORDERED_SC_LIST].lock;
+
+	spin_lock_bh(sc_lists_lock);
+
+	list_for_each_safe(tmp, tmp2, &done_sc_list->head) {
+		sc = list_entry(tmp, struct octeon_soft_command, node);
+
+		if (READ_ONCE(sc->caller_is_done)) {
+			list_del(&sc->node);
+			atomic_dec(&done_sc_list->pending_req_count);
+
+			if (*sc->status_word == COMPLETION_WORD_INIT) {
+				/* timeout; move sc to zombie list */
+				list_add_tail(&sc->node, &zombie_sc_list->head);
+				atomic_inc(&zombie_sc_list->pending_req_count);
+			} else {
+				octeon_free_soft_command(oct, sc);
+			}
+		}
+	}
+
+	spin_unlock_bh(sc_lists_lock);
+
+	return 0;
+}
+
+int octeon_free_sc_zombie_list(struct octeon_device *oct)
+{
+	struct octeon_response_list *zombie_sc_list;
+	struct octeon_soft_command *sc;
+	struct list_head *tmp, *tmp2;
+	spinlock_t *sc_lists_lock; /* lock for response_list */
+
+	zombie_sc_list = &oct->response_list[OCTEON_ZOMBIE_SC_LIST];
+	sc_lists_lock = &oct->response_list[OCTEON_ORDERED_SC_LIST].lock;
+
+	spin_lock_bh(sc_lists_lock);
+
+	list_for_each_safe(tmp, tmp2, &zombie_sc_list->head) {
+		list_del(tmp);
+		atomic_dec(&zombie_sc_list->pending_req_count);
+		sc = list_entry(tmp, struct octeon_soft_command, node);
+		octeon_free_soft_command(oct, sc);
+	}
+
+	spin_unlock_bh(sc_lists_lock);
+
+	return 0;
+}
+
+int octeon_free_sc_buffer_pool(struct octeon_device *oct)
+{
+	struct list_head *tmp, *tmp2;
+	struct octeon_soft_command *sc;
+
+	octeon_free_sc_zombie_list(oct);
+
+	spin_lock_bh(&oct->sc_buf_pool.lock);
+
+	list_for_each_safe(tmp, tmp2, &oct->sc_buf_pool.head) {
+		list_del(tmp);
+
+		sc = (struct octeon_soft_command *)tmp;
+
+		lio_dma_free(oct, sc->size, sc, sc->dma_addr);
+	}
+
+	INIT_LIST_HEAD(&oct->sc_buf_pool.head);
+
+	spin_unlock_bh(&oct->sc_buf_pool.lock);
+
+	return 0;
+}
+
+struct octeon_soft_command *octeon_alloc_soft_command(struct octeon_device *oct,
+						      u32 datasize,
+						      u32 rdatasize,
+						      u32 ctxsize)
+{
+	u64 dma_addr;
+	u32 size;
+	u32 offset = sizeof(struct octeon_soft_command);
+	struct octeon_soft_command *sc = NULL;
+	struct list_head *tmp;
+
+	if (!rdatasize)
+		rdatasize = 16;
+
+	WARN_ON((offset + datasize + rdatasize + ctxsize) >
+	       SOFT_COMMAND_BUFFER_SIZE);
+
+	spin_lock_bh(&oct->sc_buf_pool.lock);
+
+	if (list_empty(&oct->sc_buf_pool.head)) {
+		spin_unlock_bh(&oct->sc_buf_pool.lock);
+		return NULL;
+	}
+
+	list_for_each(tmp, &oct->sc_buf_pool.head)
+		break;
+
+	list_del(tmp);
+
+	atomic_inc(&oct->sc_buf_pool.alloc_buf_count);
+
+	spin_unlock_bh(&oct->sc_buf_pool.lock);
+
+	sc = (struct octeon_soft_command *)tmp;
+
+	dma_addr = sc->dma_addr;
+	size = sc->size;
+
+	memset(sc, 0, sc->size);
+
+	sc->dma_addr = dma_addr;
+	sc->size = size;
+
+	if (ctxsize) {
+		sc->ctxptr = (u8 *)sc + offset;
+		sc->ctxsize = ctxsize;
+	}
+
+	/* Start data at 128 byte boundary */
+	offset = (offset + ctxsize + 127) & 0xffffff80;
+
+	if (datasize) {
+		sc->virtdptr = (u8 *)sc + offset;
+		sc->dmadptr = dma_addr + offset;
+		sc->datasize = datasize;
+	}
+
+	/* Start rdata at 128 byte boundary */
+	offset = (offset + datasize + 127) & 0xffffff80;
+
+	if (rdatasize) {
+		WARN_ON(rdatasize < 16);
+		sc->virtrptr = (u8 *)sc + offset;
+		sc->dmarptr = dma_addr + offset;
+		sc->rdatasize = rdatasize;
+		sc->status_word = (u64 *)((u8 *)(sc->virtrptr) + rdatasize - 8);
+	}
+
+	return sc;
+}
+
+void octeon_free_soft_command(struct octeon_device *oct,
+			      struct octeon_soft_command *sc)
+{
+	spin_lock_bh(&oct->sc_buf_pool.lock);
+
+	list_add_tail(&sc->node, &oct->sc_buf_pool.head);
+
+	atomic_dec(&oct->sc_buf_pool.alloc_buf_count);
+
+	spin_unlock_bh(&oct->sc_buf_pool.lock);
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/response_manager.c b/drivers/net/ethernet/cavium/liquidio/response_manager.c
new file mode 100644
index 000000000..ac7747ccf
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/response_manager.c
@@ -0,0 +1,234 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more
+ * details.
+ **********************************************************************/
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include "liquidio_common.h"
+#include "octeon_droq.h"
+#include "octeon_iq.h"
+#include "response_manager.h"
+#include "octeon_device.h"
+#include "octeon_main.h"
+
+static void oct_poll_req_completion(struct work_struct *work);
+
+int octeon_setup_response_list(struct octeon_device *oct)
+{
+	int i, ret = 0;
+	struct cavium_wq *cwq;
+
+	for (i = 0; i < MAX_RESPONSE_LISTS; i++) {
+		INIT_LIST_HEAD(&oct->response_list[i].head);
+		spin_lock_init(&oct->response_list[i].lock);
+		atomic_set(&oct->response_list[i].pending_req_count, 0);
+	}
+	spin_lock_init(&oct->cmd_resp_wqlock);
+
+	oct->dma_comp_wq.wq = alloc_workqueue("dma-comp", WQ_MEM_RECLAIM, 0);
+	if (!oct->dma_comp_wq.wq) {
+		dev_err(&oct->pci_dev->dev, "failed to create wq thread\n");
+		return -ENOMEM;
+	}
+
+	cwq = &oct->dma_comp_wq;
+	INIT_DELAYED_WORK(&cwq->wk.work, oct_poll_req_completion);
+	cwq->wk.ctxptr = oct;
+	oct->cmd_resp_state = OCT_DRV_ONLINE;
+
+	return ret;
+}
+
+void octeon_delete_response_list(struct octeon_device *oct)
+{
+	cancel_delayed_work_sync(&oct->dma_comp_wq.wk.work);
+	destroy_workqueue(oct->dma_comp_wq.wq);
+}
+
+int lio_process_ordered_list(struct octeon_device *octeon_dev,
+			     u32 force_quit)
+{
+	struct octeon_response_list *ordered_sc_list;
+	struct octeon_soft_command *sc;
+	int request_complete = 0;
+	int resp_to_process = MAX_ORD_REQS_TO_PROCESS;
+	u32 status;
+	u64 status64;
+
+	octeon_free_sc_done_list(octeon_dev);
+
+	ordered_sc_list = &octeon_dev->response_list[OCTEON_ORDERED_SC_LIST];
+
+	do {
+		spin_lock_bh(&ordered_sc_list->lock);
+
+		if (list_empty(&ordered_sc_list->head)) {
+			spin_unlock_bh(&ordered_sc_list->lock);
+			return 1;
+		}
+
+		sc = list_first_entry(&ordered_sc_list->head,
+				      struct octeon_soft_command, node);
+
+		status = OCTEON_REQUEST_PENDING;
+
+		/* check if octeon has finished DMA'ing a response
+		 * to where rptr is pointing to
+		 */
+		status64 = *sc->status_word;
+
+		if (status64 != COMPLETION_WORD_INIT) {
+			/* This logic ensures that all 64b have been written.
+			 * 1. check byte 0 for non-FF
+			 * 2. if non-FF, then swap result from BE to host order
+			 * 3. check byte 7 (swapped to 0) for non-FF
+			 * 4. if non-FF, use the low 32-bit status code
+			 * 5. if either byte 0 or byte 7 is FF, don't use status
+			 */
+			if ((status64 & 0xff) != 0xff) {
+				octeon_swap_8B_data(&status64, 1);
+				if (((status64 & 0xff) != 0xff)) {
+					/* retrieve 16-bit firmware status */
+					status = (u32)(status64 & 0xffffULL);
+					if (status) {
+						status =
+						  FIRMWARE_STATUS_CODE(status);
+					} else {
+						/* i.e. no error */
+						status = OCTEON_REQUEST_DONE;
+					}
+				}
+			}
+		} else if (unlikely(force_quit) || (sc->expiry_time &&
+			time_after(jiffies, (unsigned long)sc->expiry_time))) {
+			struct octeon_instr_irh *irh =
+				(struct octeon_instr_irh *)&sc->cmd.cmd3.irh;
+
+			dev_err(&octeon_dev->pci_dev->dev, "%s: ", __func__);
+			dev_err(&octeon_dev->pci_dev->dev,
+				"cmd %x/%x/%llx/%llx failed, ",
+				irh->opcode, irh->subcode,
+				sc->cmd.cmd3.ossp[0], sc->cmd.cmd3.ossp[1]);
+			dev_err(&octeon_dev->pci_dev->dev,
+				"timeout (%ld, %ld)\n",
+				(long)jiffies, (long)sc->expiry_time);
+			status = OCTEON_REQUEST_TIMEOUT;
+		}
+
+		if (status != OCTEON_REQUEST_PENDING) {
+			sc->sc_status = status;
+
+			/* we have received a response or we have timed out */
+			/* remove node from linked list */
+			list_del(&sc->node);
+			atomic_dec(&octeon_dev->response_list
+				   [OCTEON_ORDERED_SC_LIST].
+				   pending_req_count);
+
+			if (!sc->callback) {
+				atomic_inc(&octeon_dev->response_list
+					   [OCTEON_DONE_SC_LIST].
+					   pending_req_count);
+				list_add_tail(&sc->node,
+					      &octeon_dev->response_list
+					      [OCTEON_DONE_SC_LIST].head);
+
+				if (unlikely(READ_ONCE(sc->caller_is_done))) {
+					/* caller does not wait for response
+					 * from firmware
+					 */
+					if (status != OCTEON_REQUEST_DONE) {
+						struct octeon_instr_irh *irh;
+
+						irh =
+						    (struct octeon_instr_irh *)
+						    &sc->cmd.cmd3.irh;
+						dev_dbg
+						    (&octeon_dev->pci_dev->dev,
+						    "%s: sc failed: opcode=%x, ",
+						    __func__, irh->opcode);
+						dev_dbg
+						    (&octeon_dev->pci_dev->dev,
+						    "subcode=%x, ossp[0]=%llx, ",
+						    irh->subcode,
+						    sc->cmd.cmd3.ossp[0]);
+						dev_dbg
+						    (&octeon_dev->pci_dev->dev,
+						    "ossp[1]=%llx, status=%d\n",
+						    sc->cmd.cmd3.ossp[1],
+						    status);
+					}
+				} else {
+					complete(&sc->complete);
+				}
+
+				spin_unlock_bh(&ordered_sc_list->lock);
+			} else {
+				/* sc with callback function */
+				if (status == OCTEON_REQUEST_TIMEOUT) {
+					atomic_inc(&octeon_dev->response_list
+						   [OCTEON_ZOMBIE_SC_LIST].
+						   pending_req_count);
+					list_add_tail(&sc->node,
+						      &octeon_dev->response_list
+						      [OCTEON_ZOMBIE_SC_LIST].
+						      head);
+				}
+
+				spin_unlock_bh(&ordered_sc_list->lock);
+
+				sc->callback(octeon_dev, status,
+					     sc->callback_arg);
+				/* sc is freed by caller */
+			}
+
+			request_complete++;
+
+		} else {
+			/* no response yet */
+			request_complete = 0;
+			spin_unlock_bh
+			    (&ordered_sc_list->lock);
+		}
+
+		/* If we hit the Max Ordered requests to process every loop,
+		 * we quit
+		 * and let this function be invoked the next time the poll
+		 * thread runs
+		 * to process the remaining requests. This function can take up
+		 * the entire CPU if there is no upper limit to the requests
+		 * processed.
+		 */
+		if (request_complete >= resp_to_process)
+			break;
+	} while (request_complete);
+
+	return 0;
+}
+
+static void oct_poll_req_completion(struct work_struct *work)
+{
+	struct cavium_wk *wk = (struct cavium_wk *)work;
+	struct octeon_device *oct = (struct octeon_device *)wk->ctxptr;
+	struct cavium_wq *cwq = &oct->dma_comp_wq;
+
+	lio_process_ordered_list(oct, 0);
+
+	if (atomic_read(&oct->response_list
+			[OCTEON_ORDERED_SC_LIST].pending_req_count))
+		queue_delayed_work(cwq->wq, &cwq->wk.work, msecs_to_jiffies(1));
+}
diff --git a/drivers/net/ethernet/cavium/liquidio/response_manager.h b/drivers/net/ethernet/cavium/liquidio/response_manager.h
new file mode 100644
index 000000000..ed4020d26
--- /dev/null
+++ b/drivers/net/ethernet/cavium/liquidio/response_manager.h
@@ -0,0 +1,143 @@
+/**********************************************************************
+ * Author: Cavium, Inc.
+ *
+ * Contact: support@cavium.com
+ *          Please include "LiquidIO" in the subject.
+ *
+ * Copyright (c) 2003-2016 Cavium, Inc.
+ *
+ * This file is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, Version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This file is distributed in the hope that it will be useful, but
+ * AS-IS and WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE, TITLE, or
+ * NONINFRINGEMENT.  See the GNU General Public License for more
+ * details.
+ **********************************************************************/
+
+/*! \file response_manager.h
+ *  \brief Host Driver:  Response queues for host instructions.
+ */
+
+#ifndef __RESPONSE_MANAGER_H__
+#define __RESPONSE_MANAGER_H__
+
+/** Maximum ordered requests to process in every invocation of
+ * lio_process_ordered_list(). The function will continue to process requests
+ * as long as it can find one that has finished processing. If it keeps
+ * finding requests that have completed, the function can run for ever. The
+ * value defined here sets an upper limit on the number of requests it can
+ * process before it returns control to the poll thread.
+ */
+#define  MAX_ORD_REQS_TO_PROCESS   4096
+
+/** Head of a response list. There are several response lists in the
+ *  system. One for each response order- Unordered, ordered
+ *  and 1 for noresponse entries on each instruction queue.
+ */
+struct octeon_response_list {
+	/** List structure to add delete pending entries to */
+	struct list_head head;
+
+	/** A lock for this response list */
+	spinlock_t lock;
+
+	atomic_t pending_req_count;
+};
+
+/** The type of response list.
+ */
+enum {
+	OCTEON_ORDERED_LIST = 0,
+	OCTEON_UNORDERED_NONBLOCKING_LIST = 1,
+	OCTEON_UNORDERED_BLOCKING_LIST = 2,
+	OCTEON_ORDERED_SC_LIST = 3,
+	OCTEON_DONE_SC_LIST = 4,
+	OCTEON_ZOMBIE_SC_LIST = 5
+};
+
+/** Response Order values for a Octeon Request. */
+enum {
+	OCTEON_RESP_ORDERED = 0,
+	OCTEON_RESP_UNORDERED = 1,
+	OCTEON_RESP_NORESPONSE = 2
+};
+
+/** Error codes  used in Octeon Host-Core communication.
+ *
+ *   31            16 15            0
+ *   ---------------------------------
+ *   |               |               |
+ *   ---------------------------------
+ *   Error codes are 32-bit wide. The upper 16-bits, called Major Error Number,
+ *   are reserved to identify the group to which the error code belongs. The
+ *   lower 16-bits, called Minor Error Number, carry the actual code.
+ *
+ *   So error codes are (MAJOR NUMBER << 16)| MINOR_NUMBER.
+ */
+
+/*------------   Error codes used by host driver   -----------------*/
+#define DRIVER_MAJOR_ERROR_CODE           0x0000
+/*------   Error codes used by firmware (bits 15..0 set by firmware */
+#define FIRMWARE_MAJOR_ERROR_CODE         0x0001
+
+/**  A value of 0x00000000 indicates no error i.e. success */
+#define DRIVER_ERROR_NONE                 0x00000000
+
+#define DRIVER_ERROR_REQ_PENDING          0x00000001
+#define DRIVER_ERROR_REQ_TIMEOUT          0x00000003
+#define DRIVER_ERROR_REQ_EINTR            0x00000004
+#define DRIVER_ERROR_REQ_ENXIO            0x00000006
+#define DRIVER_ERROR_REQ_ENOMEM           0x0000000C
+#define DRIVER_ERROR_REQ_EINVAL           0x00000016
+#define DRIVER_ERROR_REQ_FAILED           0x000000ff
+
+/** Status for a request.
+ * If a request is not queued to Octeon by the driver, the driver returns
+ * an error condition that's describe by one of the OCTEON_REQ_ERR_* value
+ * below. If the request is successfully queued, the driver will return
+ * a OCTEON_REQUEST_PENDING status. OCTEON_REQUEST_TIMEOUT and
+ * OCTEON_REQUEST_INTERRUPTED are only returned by the driver if the
+ * response for request failed to arrive before a time-out period or if
+ * the request processing * got interrupted due to a signal respectively.
+ */
+enum {
+	OCTEON_REQUEST_DONE = (DRIVER_ERROR_NONE),
+	OCTEON_REQUEST_PENDING = (DRIVER_ERROR_REQ_PENDING),
+	OCTEON_REQUEST_TIMEOUT = (DRIVER_ERROR_REQ_TIMEOUT),
+	OCTEON_REQUEST_INTERRUPTED = (DRIVER_ERROR_REQ_EINTR),
+	OCTEON_REQUEST_NO_DEVICE = (0x00000021),
+	OCTEON_REQUEST_NOT_RUNNING,
+	OCTEON_REQUEST_INVALID_IQ,
+	OCTEON_REQUEST_INVALID_BUFCNT,
+	OCTEON_REQUEST_INVALID_RESP_ORDER,
+	OCTEON_REQUEST_NO_MEMORY,
+	OCTEON_REQUEST_INVALID_BUFSIZE,
+	OCTEON_REQUEST_NO_PENDING_ENTRY,
+	OCTEON_REQUEST_NO_IQ_SPACE = (0x7FFFFFFF)
+
+};
+
+#define FIRMWARE_STATUS_CODE(status) \
+	((FIRMWARE_MAJOR_ERROR_CODE << 16) | (status))
+
+/** Initialize the response lists. The number of response lists to create is
+ * given by count.
+ * @param octeon_dev      - the octeon device structure.
+ */
+int octeon_setup_response_list(struct octeon_device *octeon_dev);
+
+void octeon_delete_response_list(struct octeon_device *octeon_dev);
+
+/** Check the status of first entry in the ordered list. If the instruction at
+ * that entry finished processing or has timed-out, the entry is cleaned.
+ * @param octeon_dev  - the octeon device structure.
+ * @param force_quit - the request is forced to timeout if this is 1
+ * @return 1 if the ordered list is empty, 0 otherwise.
+ */
+int lio_process_ordered_list(struct octeon_device *octeon_dev,
+			     u32 force_quit);
+
+#endif
diff --git a/drivers/net/ethernet/cavium/octeon/Makefile b/drivers/net/ethernet/cavium/octeon/Makefile
new file mode 100644
index 000000000..4f5098f6b
--- /dev/null
+++ b/drivers/net/ethernet/cavium/octeon/Makefile
@@ -0,0 +1,6 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# Makefile for the Cavium network device drivers.
+#
+
+obj-$(CONFIG_OCTEON_MGMT_ETHERNET)	+= octeon_mgmt.o
diff --git a/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c b/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c
new file mode 100644
index 000000000..edde0b8fa
--- /dev/null
+++ b/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c
@@ -0,0 +1,1557 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2009-2012 Cavium, Inc
+ */
+
+#include <linux/platform_device.h>
+#include <linux/dma-mapping.h>
+#include <linux/etherdevice.h>
+#include <linux/capability.h>
+#include <linux/net_tstamp.h>
+#include <linux/interrupt.h>
+#include <linux/netdevice.h>
+#include <linux/spinlock.h>
+#include <linux/if_vlan.h>
+#include <linux/of_mdio.h>
+#include <linux/module.h>
+#include <linux/of_net.h>
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/phy.h>
+#include <linux/io.h>
+
+#include <asm/octeon/octeon.h>
+#include <asm/octeon/cvmx-mixx-defs.h>
+#include <asm/octeon/cvmx-agl-defs.h>
+
+#define DRV_NAME "octeon_mgmt"
+#define DRV_DESCRIPTION \
+	"Cavium Networks Octeon MII (management) port Network Driver"
+
+#define OCTEON_MGMT_NAPI_WEIGHT 16
+
+/* Ring sizes that are powers of two allow for more efficient modulo
+ * opertions.
+ */
+#define OCTEON_MGMT_RX_RING_SIZE 512
+#define OCTEON_MGMT_TX_RING_SIZE 128
+
+/* Allow 8 bytes for vlan and FCS. */
+#define OCTEON_MGMT_RX_HEADROOM (ETH_HLEN + ETH_FCS_LEN + VLAN_HLEN)
+
+union mgmt_port_ring_entry {
+	u64 d64;
+	struct {
+#define RING_ENTRY_CODE_DONE 0xf
+#define RING_ENTRY_CODE_MORE 0x10
+#ifdef __BIG_ENDIAN_BITFIELD
+		u64 reserved_62_63:2;
+		/* Length of the buffer/packet in bytes */
+		u64 len:14;
+		/* For TX, signals that the packet should be timestamped */
+		u64 tstamp:1;
+		/* The RX error code */
+		u64 code:7;
+		/* Physical address of the buffer */
+		u64 addr:40;
+#else
+		u64 addr:40;
+		u64 code:7;
+		u64 tstamp:1;
+		u64 len:14;
+		u64 reserved_62_63:2;
+#endif
+	} s;
+};
+
+#define MIX_ORING1	0x0
+#define MIX_ORING2	0x8
+#define MIX_IRING1	0x10
+#define MIX_IRING2	0x18
+#define MIX_CTL		0x20
+#define MIX_IRHWM	0x28
+#define MIX_IRCNT	0x30
+#define MIX_ORHWM	0x38
+#define MIX_ORCNT	0x40
+#define MIX_ISR		0x48
+#define MIX_INTENA	0x50
+#define MIX_REMCNT	0x58
+#define MIX_BIST	0x78
+
+#define AGL_GMX_PRT_CFG			0x10
+#define AGL_GMX_RX_FRM_CTL		0x18
+#define AGL_GMX_RX_FRM_MAX		0x30
+#define AGL_GMX_RX_JABBER		0x38
+#define AGL_GMX_RX_STATS_CTL		0x50
+
+#define AGL_GMX_RX_STATS_PKTS_DRP	0xb0
+#define AGL_GMX_RX_STATS_OCTS_DRP	0xb8
+#define AGL_GMX_RX_STATS_PKTS_BAD	0xc0
+
+#define AGL_GMX_RX_ADR_CTL		0x100
+#define AGL_GMX_RX_ADR_CAM_EN		0x108
+#define AGL_GMX_RX_ADR_CAM0		0x180
+#define AGL_GMX_RX_ADR_CAM1		0x188
+#define AGL_GMX_RX_ADR_CAM2		0x190
+#define AGL_GMX_RX_ADR_CAM3		0x198
+#define AGL_GMX_RX_ADR_CAM4		0x1a0
+#define AGL_GMX_RX_ADR_CAM5		0x1a8
+
+#define AGL_GMX_TX_CLK			0x208
+#define AGL_GMX_TX_STATS_CTL		0x268
+#define AGL_GMX_TX_CTL			0x270
+#define AGL_GMX_TX_STAT0		0x280
+#define AGL_GMX_TX_STAT1		0x288
+#define AGL_GMX_TX_STAT2		0x290
+#define AGL_GMX_TX_STAT3		0x298
+#define AGL_GMX_TX_STAT4		0x2a0
+#define AGL_GMX_TX_STAT5		0x2a8
+#define AGL_GMX_TX_STAT6		0x2b0
+#define AGL_GMX_TX_STAT7		0x2b8
+#define AGL_GMX_TX_STAT8		0x2c0
+#define AGL_GMX_TX_STAT9		0x2c8
+
+struct octeon_mgmt {
+	struct net_device *netdev;
+	u64 mix;
+	u64 agl;
+	u64 agl_prt_ctl;
+	int port;
+	int irq;
+	bool has_rx_tstamp;
+	u64 *tx_ring;
+	dma_addr_t tx_ring_handle;
+	unsigned int tx_next;
+	unsigned int tx_next_clean;
+	unsigned int tx_current_fill;
+	/* The tx_list lock also protects the ring related variables */
+	struct sk_buff_head tx_list;
+
+	/* RX variables only touched in napi_poll.  No locking necessary. */
+	u64 *rx_ring;
+	dma_addr_t rx_ring_handle;
+	unsigned int rx_next;
+	unsigned int rx_next_fill;
+	unsigned int rx_current_fill;
+	struct sk_buff_head rx_list;
+
+	spinlock_t lock;
+	unsigned int last_duplex;
+	unsigned int last_link;
+	unsigned int last_speed;
+	struct device *dev;
+	struct napi_struct napi;
+	struct tasklet_struct tx_clean_tasklet;
+	struct device_node *phy_np;
+	resource_size_t mix_phys;
+	resource_size_t mix_size;
+	resource_size_t agl_phys;
+	resource_size_t agl_size;
+	resource_size_t agl_prt_ctl_phys;
+	resource_size_t agl_prt_ctl_size;
+};
+
+static void octeon_mgmt_set_rx_irq(struct octeon_mgmt *p, int enable)
+{
+	union cvmx_mixx_intena mix_intena;
+	unsigned long flags;
+
+	spin_lock_irqsave(&p->lock, flags);
+	mix_intena.u64 = cvmx_read_csr(p->mix + MIX_INTENA);
+	mix_intena.s.ithena = enable ? 1 : 0;
+	cvmx_write_csr(p->mix + MIX_INTENA, mix_intena.u64);
+	spin_unlock_irqrestore(&p->lock, flags);
+}
+
+static void octeon_mgmt_set_tx_irq(struct octeon_mgmt *p, int enable)
+{
+	union cvmx_mixx_intena mix_intena;
+	unsigned long flags;
+
+	spin_lock_irqsave(&p->lock, flags);
+	mix_intena.u64 = cvmx_read_csr(p->mix + MIX_INTENA);
+	mix_intena.s.othena = enable ? 1 : 0;
+	cvmx_write_csr(p->mix + MIX_INTENA, mix_intena.u64);
+	spin_unlock_irqrestore(&p->lock, flags);
+}
+
+static void octeon_mgmt_enable_rx_irq(struct octeon_mgmt *p)
+{
+	octeon_mgmt_set_rx_irq(p, 1);
+}
+
+static void octeon_mgmt_disable_rx_irq(struct octeon_mgmt *p)
+{
+	octeon_mgmt_set_rx_irq(p, 0);
+}
+
+static void octeon_mgmt_enable_tx_irq(struct octeon_mgmt *p)
+{
+	octeon_mgmt_set_tx_irq(p, 1);
+}
+
+static void octeon_mgmt_disable_tx_irq(struct octeon_mgmt *p)
+{
+	octeon_mgmt_set_tx_irq(p, 0);
+}
+
+static unsigned int ring_max_fill(unsigned int ring_size)
+{
+	return ring_size - 8;
+}
+
+static unsigned int ring_size_to_bytes(unsigned int ring_size)
+{
+	return ring_size * sizeof(union mgmt_port_ring_entry);
+}
+
+static void octeon_mgmt_rx_fill_ring(struct net_device *netdev)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+
+	while (p->rx_current_fill < ring_max_fill(OCTEON_MGMT_RX_RING_SIZE)) {
+		unsigned int size;
+		union mgmt_port_ring_entry re;
+		struct sk_buff *skb;
+
+		/* CN56XX pass 1 needs 8 bytes of padding.  */
+		size = netdev->mtu + OCTEON_MGMT_RX_HEADROOM + 8 + NET_IP_ALIGN;
+
+		skb = netdev_alloc_skb(netdev, size);
+		if (!skb)
+			break;
+		skb_reserve(skb, NET_IP_ALIGN);
+		__skb_queue_tail(&p->rx_list, skb);
+
+		re.d64 = 0;
+		re.s.len = size;
+		re.s.addr = dma_map_single(p->dev, skb->data,
+					   size,
+					   DMA_FROM_DEVICE);
+
+		/* Put it in the ring.  */
+		p->rx_ring[p->rx_next_fill] = re.d64;
+		/* Make sure there is no reorder of filling the ring and ringing
+		 * the bell
+		 */
+		wmb();
+
+		dma_sync_single_for_device(p->dev, p->rx_ring_handle,
+					   ring_size_to_bytes(OCTEON_MGMT_RX_RING_SIZE),
+					   DMA_BIDIRECTIONAL);
+		p->rx_next_fill =
+			(p->rx_next_fill + 1) % OCTEON_MGMT_RX_RING_SIZE;
+		p->rx_current_fill++;
+		/* Ring the bell.  */
+		cvmx_write_csr(p->mix + MIX_IRING2, 1);
+	}
+}
+
+static void octeon_mgmt_clean_tx_buffers(struct octeon_mgmt *p)
+{
+	union cvmx_mixx_orcnt mix_orcnt;
+	union mgmt_port_ring_entry re;
+	struct sk_buff *skb;
+	int cleaned = 0;
+	unsigned long flags;
+
+	mix_orcnt.u64 = cvmx_read_csr(p->mix + MIX_ORCNT);
+	while (mix_orcnt.s.orcnt) {
+		spin_lock_irqsave(&p->tx_list.lock, flags);
+
+		mix_orcnt.u64 = cvmx_read_csr(p->mix + MIX_ORCNT);
+
+		if (mix_orcnt.s.orcnt == 0) {
+			spin_unlock_irqrestore(&p->tx_list.lock, flags);
+			break;
+		}
+
+		dma_sync_single_for_cpu(p->dev, p->tx_ring_handle,
+					ring_size_to_bytes(OCTEON_MGMT_TX_RING_SIZE),
+					DMA_BIDIRECTIONAL);
+
+		re.d64 = p->tx_ring[p->tx_next_clean];
+		p->tx_next_clean =
+			(p->tx_next_clean + 1) % OCTEON_MGMT_TX_RING_SIZE;
+		skb = __skb_dequeue(&p->tx_list);
+
+		mix_orcnt.u64 = 0;
+		mix_orcnt.s.orcnt = 1;
+
+		/* Acknowledge to hardware that we have the buffer.  */
+		cvmx_write_csr(p->mix + MIX_ORCNT, mix_orcnt.u64);
+		p->tx_current_fill--;
+
+		spin_unlock_irqrestore(&p->tx_list.lock, flags);
+
+		dma_unmap_single(p->dev, re.s.addr, re.s.len,
+				 DMA_TO_DEVICE);
+
+		/* Read the hardware TX timestamp if one was recorded */
+		if (unlikely(re.s.tstamp)) {
+			struct skb_shared_hwtstamps ts;
+			u64 ns;
+
+			memset(&ts, 0, sizeof(ts));
+			/* Read the timestamp */
+			ns = cvmx_read_csr(CVMX_MIXX_TSTAMP(p->port));
+			/* Remove the timestamp from the FIFO */
+			cvmx_write_csr(CVMX_MIXX_TSCTL(p->port), 0);
+			/* Tell the kernel about the timestamp */
+			ts.hwtstamp = ns_to_ktime(ns);
+			skb_tstamp_tx(skb, &ts);
+		}
+
+		dev_kfree_skb_any(skb);
+		cleaned++;
+
+		mix_orcnt.u64 = cvmx_read_csr(p->mix + MIX_ORCNT);
+	}
+
+	if (cleaned && netif_queue_stopped(p->netdev))
+		netif_wake_queue(p->netdev);
+}
+
+static void octeon_mgmt_clean_tx_tasklet(struct tasklet_struct *t)
+{
+	struct octeon_mgmt *p = from_tasklet(p, t, tx_clean_tasklet);
+	octeon_mgmt_clean_tx_buffers(p);
+	octeon_mgmt_enable_tx_irq(p);
+}
+
+static void octeon_mgmt_update_rx_stats(struct net_device *netdev)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+	unsigned long flags;
+	u64 drop, bad;
+
+	/* These reads also clear the count registers.  */
+	drop = cvmx_read_csr(p->agl + AGL_GMX_RX_STATS_PKTS_DRP);
+	bad = cvmx_read_csr(p->agl + AGL_GMX_RX_STATS_PKTS_BAD);
+
+	if (drop || bad) {
+		/* Do an atomic update. */
+		spin_lock_irqsave(&p->lock, flags);
+		netdev->stats.rx_errors += bad;
+		netdev->stats.rx_dropped += drop;
+		spin_unlock_irqrestore(&p->lock, flags);
+	}
+}
+
+static void octeon_mgmt_update_tx_stats(struct net_device *netdev)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+	unsigned long flags;
+
+	union cvmx_agl_gmx_txx_stat0 s0;
+	union cvmx_agl_gmx_txx_stat1 s1;
+
+	/* These reads also clear the count registers.  */
+	s0.u64 = cvmx_read_csr(p->agl + AGL_GMX_TX_STAT0);
+	s1.u64 = cvmx_read_csr(p->agl + AGL_GMX_TX_STAT1);
+
+	if (s0.s.xsdef || s0.s.xscol || s1.s.scol || s1.s.mcol) {
+		/* Do an atomic update. */
+		spin_lock_irqsave(&p->lock, flags);
+		netdev->stats.tx_errors += s0.s.xsdef + s0.s.xscol;
+		netdev->stats.collisions += s1.s.scol + s1.s.mcol;
+		spin_unlock_irqrestore(&p->lock, flags);
+	}
+}
+
+/*
+ * Dequeue a receive skb and its corresponding ring entry.  The ring
+ * entry is returned, *pskb is updated to point to the skb.
+ */
+static u64 octeon_mgmt_dequeue_rx_buffer(struct octeon_mgmt *p,
+					 struct sk_buff **pskb)
+{
+	union mgmt_port_ring_entry re;
+
+	dma_sync_single_for_cpu(p->dev, p->rx_ring_handle,
+				ring_size_to_bytes(OCTEON_MGMT_RX_RING_SIZE),
+				DMA_BIDIRECTIONAL);
+
+	re.d64 = p->rx_ring[p->rx_next];
+	p->rx_next = (p->rx_next + 1) % OCTEON_MGMT_RX_RING_SIZE;
+	p->rx_current_fill--;
+	*pskb = __skb_dequeue(&p->rx_list);
+
+	dma_unmap_single(p->dev, re.s.addr,
+			 ETH_FRAME_LEN + OCTEON_MGMT_RX_HEADROOM,
+			 DMA_FROM_DEVICE);
+
+	return re.d64;
+}
+
+
+static int octeon_mgmt_receive_one(struct octeon_mgmt *p)
+{
+	struct net_device *netdev = p->netdev;
+	union cvmx_mixx_ircnt mix_ircnt;
+	union mgmt_port_ring_entry re;
+	struct sk_buff *skb;
+	struct sk_buff *skb2;
+	struct sk_buff *skb_new;
+	union mgmt_port_ring_entry re2;
+	int rc = 1;
+
+
+	re.d64 = octeon_mgmt_dequeue_rx_buffer(p, &skb);
+	if (likely(re.s.code == RING_ENTRY_CODE_DONE)) {
+		/* A good packet, send it up. */
+		skb_put(skb, re.s.len);
+good:
+		/* Process the RX timestamp if it was recorded */
+		if (p->has_rx_tstamp) {
+			/* The first 8 bytes are the timestamp */
+			u64 ns = *(u64 *)skb->data;
+			struct skb_shared_hwtstamps *ts;
+			ts = skb_hwtstamps(skb);
+			ts->hwtstamp = ns_to_ktime(ns);
+			__skb_pull(skb, 8);
+		}
+		skb->protocol = eth_type_trans(skb, netdev);
+		netdev->stats.rx_packets++;
+		netdev->stats.rx_bytes += skb->len;
+		netif_receive_skb(skb);
+		rc = 0;
+	} else if (re.s.code == RING_ENTRY_CODE_MORE) {
+		/* Packet split across skbs.  This can happen if we
+		 * increase the MTU.  Buffers that are already in the
+		 * rx ring can then end up being too small.  As the rx
+		 * ring is refilled, buffers sized for the new MTU
+		 * will be used and we should go back to the normal
+		 * non-split case.
+		 */
+		skb_put(skb, re.s.len);
+		do {
+			re2.d64 = octeon_mgmt_dequeue_rx_buffer(p, &skb2);
+			if (re2.s.code != RING_ENTRY_CODE_MORE
+				&& re2.s.code != RING_ENTRY_CODE_DONE)
+				goto split_error;
+			skb_put(skb2,  re2.s.len);
+			skb_new = skb_copy_expand(skb, 0, skb2->len,
+						  GFP_ATOMIC);
+			if (!skb_new)
+				goto split_error;
+			if (skb_copy_bits(skb2, 0, skb_tail_pointer(skb_new),
+					  skb2->len))
+				goto split_error;
+			skb_put(skb_new, skb2->len);
+			dev_kfree_skb_any(skb);
+			dev_kfree_skb_any(skb2);
+			skb = skb_new;
+		} while (re2.s.code == RING_ENTRY_CODE_MORE);
+		goto good;
+	} else {
+		/* Some other error, discard it. */
+		dev_kfree_skb_any(skb);
+		/* Error statistics are accumulated in
+		 * octeon_mgmt_update_rx_stats.
+		 */
+	}
+	goto done;
+split_error:
+	/* Discard the whole mess. */
+	dev_kfree_skb_any(skb);
+	dev_kfree_skb_any(skb2);
+	while (re2.s.code == RING_ENTRY_CODE_MORE) {
+		re2.d64 = octeon_mgmt_dequeue_rx_buffer(p, &skb2);
+		dev_kfree_skb_any(skb2);
+	}
+	netdev->stats.rx_errors++;
+
+done:
+	/* Tell the hardware we processed a packet.  */
+	mix_ircnt.u64 = 0;
+	mix_ircnt.s.ircnt = 1;
+	cvmx_write_csr(p->mix + MIX_IRCNT, mix_ircnt.u64);
+	return rc;
+}
+
+static int octeon_mgmt_receive_packets(struct octeon_mgmt *p, int budget)
+{
+	unsigned int work_done = 0;
+	union cvmx_mixx_ircnt mix_ircnt;
+	int rc;
+
+	mix_ircnt.u64 = cvmx_read_csr(p->mix + MIX_IRCNT);
+	while (work_done < budget && mix_ircnt.s.ircnt) {
+
+		rc = octeon_mgmt_receive_one(p);
+		if (!rc)
+			work_done++;
+
+		/* Check for more packets. */
+		mix_ircnt.u64 = cvmx_read_csr(p->mix + MIX_IRCNT);
+	}
+
+	octeon_mgmt_rx_fill_ring(p->netdev);
+
+	return work_done;
+}
+
+static int octeon_mgmt_napi_poll(struct napi_struct *napi, int budget)
+{
+	struct octeon_mgmt *p = container_of(napi, struct octeon_mgmt, napi);
+	struct net_device *netdev = p->netdev;
+	unsigned int work_done = 0;
+
+	work_done = octeon_mgmt_receive_packets(p, budget);
+
+	if (work_done < budget) {
+		/* We stopped because no more packets were available. */
+		napi_complete_done(napi, work_done);
+		octeon_mgmt_enable_rx_irq(p);
+	}
+	octeon_mgmt_update_rx_stats(netdev);
+
+	return work_done;
+}
+
+/* Reset the hardware to clean state.  */
+static void octeon_mgmt_reset_hw(struct octeon_mgmt *p)
+{
+	union cvmx_mixx_ctl mix_ctl;
+	union cvmx_mixx_bist mix_bist;
+	union cvmx_agl_gmx_bist agl_gmx_bist;
+
+	mix_ctl.u64 = 0;
+	cvmx_write_csr(p->mix + MIX_CTL, mix_ctl.u64);
+	do {
+		mix_ctl.u64 = cvmx_read_csr(p->mix + MIX_CTL);
+	} while (mix_ctl.s.busy);
+	mix_ctl.s.reset = 1;
+	cvmx_write_csr(p->mix + MIX_CTL, mix_ctl.u64);
+	cvmx_read_csr(p->mix + MIX_CTL);
+	octeon_io_clk_delay(64);
+
+	mix_bist.u64 = cvmx_read_csr(p->mix + MIX_BIST);
+	if (mix_bist.u64)
+		dev_warn(p->dev, "MIX failed BIST (0x%016llx)\n",
+			(unsigned long long)mix_bist.u64);
+
+	agl_gmx_bist.u64 = cvmx_read_csr(CVMX_AGL_GMX_BIST);
+	if (agl_gmx_bist.u64)
+		dev_warn(p->dev, "AGL failed BIST (0x%016llx)\n",
+			 (unsigned long long)agl_gmx_bist.u64);
+}
+
+struct octeon_mgmt_cam_state {
+	u64 cam[6];
+	u64 cam_mask;
+	int cam_index;
+};
+
+static void octeon_mgmt_cam_state_add(struct octeon_mgmt_cam_state *cs,
+				      const unsigned char *addr)
+{
+	int i;
+
+	for (i = 0; i < 6; i++)
+		cs->cam[i] |= (u64)addr[i] << (8 * (cs->cam_index));
+	cs->cam_mask |= (1ULL << cs->cam_index);
+	cs->cam_index++;
+}
+
+static void octeon_mgmt_set_rx_filtering(struct net_device *netdev)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+	union cvmx_agl_gmx_rxx_adr_ctl adr_ctl;
+	union cvmx_agl_gmx_prtx_cfg agl_gmx_prtx;
+	unsigned long flags;
+	unsigned int prev_packet_enable;
+	unsigned int cam_mode = 1; /* 1 - Accept on CAM match */
+	unsigned int multicast_mode = 1; /* 1 - Reject all multicast.  */
+	struct octeon_mgmt_cam_state cam_state;
+	struct netdev_hw_addr *ha;
+	int available_cam_entries;
+
+	memset(&cam_state, 0, sizeof(cam_state));
+
+	if ((netdev->flags & IFF_PROMISC) || netdev->uc.count > 7) {
+		cam_mode = 0;
+		available_cam_entries = 8;
+	} else {
+		/* One CAM entry for the primary address, leaves seven
+		 * for the secondary addresses.
+		 */
+		available_cam_entries = 7 - netdev->uc.count;
+	}
+
+	if (netdev->flags & IFF_MULTICAST) {
+		if (cam_mode == 0 || (netdev->flags & IFF_ALLMULTI) ||
+		    netdev_mc_count(netdev) > available_cam_entries)
+			multicast_mode = 2; /* 2 - Accept all multicast.  */
+		else
+			multicast_mode = 0; /* 0 - Use CAM.  */
+	}
+
+	if (cam_mode == 1) {
+		/* Add primary address. */
+		octeon_mgmt_cam_state_add(&cam_state, netdev->dev_addr);
+		netdev_for_each_uc_addr(ha, netdev)
+			octeon_mgmt_cam_state_add(&cam_state, ha->addr);
+	}
+	if (multicast_mode == 0) {
+		netdev_for_each_mc_addr(ha, netdev)
+			octeon_mgmt_cam_state_add(&cam_state, ha->addr);
+	}
+
+	spin_lock_irqsave(&p->lock, flags);
+
+	/* Disable packet I/O. */
+	agl_gmx_prtx.u64 = cvmx_read_csr(p->agl + AGL_GMX_PRT_CFG);
+	prev_packet_enable = agl_gmx_prtx.s.en;
+	agl_gmx_prtx.s.en = 0;
+	cvmx_write_csr(p->agl + AGL_GMX_PRT_CFG, agl_gmx_prtx.u64);
+
+	adr_ctl.u64 = 0;
+	adr_ctl.s.cam_mode = cam_mode;
+	adr_ctl.s.mcst = multicast_mode;
+	adr_ctl.s.bcst = 1;     /* Allow broadcast */
+
+	cvmx_write_csr(p->agl + AGL_GMX_RX_ADR_CTL, adr_ctl.u64);
+
+	cvmx_write_csr(p->agl + AGL_GMX_RX_ADR_CAM0, cam_state.cam[0]);
+	cvmx_write_csr(p->agl + AGL_GMX_RX_ADR_CAM1, cam_state.cam[1]);
+	cvmx_write_csr(p->agl + AGL_GMX_RX_ADR_CAM2, cam_state.cam[2]);
+	cvmx_write_csr(p->agl + AGL_GMX_RX_ADR_CAM3, cam_state.cam[3]);
+	cvmx_write_csr(p->agl + AGL_GMX_RX_ADR_CAM4, cam_state.cam[4]);
+	cvmx_write_csr(p->agl + AGL_GMX_RX_ADR_CAM5, cam_state.cam[5]);
+	cvmx_write_csr(p->agl + AGL_GMX_RX_ADR_CAM_EN, cam_state.cam_mask);
+
+	/* Restore packet I/O. */
+	agl_gmx_prtx.s.en = prev_packet_enable;
+	cvmx_write_csr(p->agl + AGL_GMX_PRT_CFG, agl_gmx_prtx.u64);
+
+	spin_unlock_irqrestore(&p->lock, flags);
+}
+
+static int octeon_mgmt_set_mac_address(struct net_device *netdev, void *addr)
+{
+	int r = eth_mac_addr(netdev, addr);
+
+	if (r)
+		return r;
+
+	octeon_mgmt_set_rx_filtering(netdev);
+
+	return 0;
+}
+
+static int octeon_mgmt_change_mtu(struct net_device *netdev, int new_mtu)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+	int max_packet = new_mtu + ETH_HLEN + ETH_FCS_LEN;
+
+	netdev->mtu = new_mtu;
+
+	/* HW lifts the limit if the frame is VLAN tagged
+	 * (+4 bytes per each tag, up to two tags)
+	 */
+	cvmx_write_csr(p->agl + AGL_GMX_RX_FRM_MAX, max_packet);
+	/* Set the hardware to truncate packets larger than the MTU. The jabber
+	 * register must be set to a multiple of 8 bytes, so round up. JABBER is
+	 * an unconditional limit, so we need to account for two possible VLAN
+	 * tags.
+	 */
+	cvmx_write_csr(p->agl + AGL_GMX_RX_JABBER,
+		       (max_packet + 7 + VLAN_HLEN * 2) & 0xfff8);
+
+	return 0;
+}
+
+static irqreturn_t octeon_mgmt_interrupt(int cpl, void *dev_id)
+{
+	struct net_device *netdev = dev_id;
+	struct octeon_mgmt *p = netdev_priv(netdev);
+	union cvmx_mixx_isr mixx_isr;
+
+	mixx_isr.u64 = cvmx_read_csr(p->mix + MIX_ISR);
+
+	/* Clear any pending interrupts */
+	cvmx_write_csr(p->mix + MIX_ISR, mixx_isr.u64);
+	cvmx_read_csr(p->mix + MIX_ISR);
+
+	if (mixx_isr.s.irthresh) {
+		octeon_mgmt_disable_rx_irq(p);
+		napi_schedule(&p->napi);
+	}
+	if (mixx_isr.s.orthresh) {
+		octeon_mgmt_disable_tx_irq(p);
+		tasklet_schedule(&p->tx_clean_tasklet);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int octeon_mgmt_ioctl_hwtstamp(struct net_device *netdev,
+				      struct ifreq *rq, int cmd)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+	struct hwtstamp_config config;
+	union cvmx_mio_ptp_clock_cfg ptp;
+	union cvmx_agl_gmx_rxx_frm_ctl rxx_frm_ctl;
+	bool have_hw_timestamps = false;
+
+	if (copy_from_user(&config, rq->ifr_data, sizeof(config)))
+		return -EFAULT;
+
+	/* Check the status of hardware for tiemstamps */
+	if (OCTEON_IS_MODEL(OCTEON_CN6XXX)) {
+		/* Get the current state of the PTP clock */
+		ptp.u64 = cvmx_read_csr(CVMX_MIO_PTP_CLOCK_CFG);
+		if (!ptp.s.ext_clk_en) {
+			/* The clock has not been configured to use an
+			 * external source.  Program it to use the main clock
+			 * reference.
+			 */
+			u64 clock_comp = (NSEC_PER_SEC << 32) /	octeon_get_io_clock_rate();
+			if (!ptp.s.ptp_en)
+				cvmx_write_csr(CVMX_MIO_PTP_CLOCK_COMP, clock_comp);
+			netdev_info(netdev,
+				    "PTP Clock using sclk reference @ %lldHz\n",
+				    (NSEC_PER_SEC << 32) / clock_comp);
+		} else {
+			/* The clock is already programmed to use a GPIO */
+			u64 clock_comp = cvmx_read_csr(CVMX_MIO_PTP_CLOCK_COMP);
+			netdev_info(netdev,
+				    "PTP Clock using GPIO%d @ %lld Hz\n",
+				    ptp.s.ext_clk_in, (NSEC_PER_SEC << 32) / clock_comp);
+		}
+
+		/* Enable the clock if it wasn't done already */
+		if (!ptp.s.ptp_en) {
+			ptp.s.ptp_en = 1;
+			cvmx_write_csr(CVMX_MIO_PTP_CLOCK_CFG, ptp.u64);
+		}
+		have_hw_timestamps = true;
+	}
+
+	if (!have_hw_timestamps)
+		return -EINVAL;
+
+	switch (config.tx_type) {
+	case HWTSTAMP_TX_OFF:
+	case HWTSTAMP_TX_ON:
+		break;
+	default:
+		return -ERANGE;
+	}
+
+	switch (config.rx_filter) {
+	case HWTSTAMP_FILTER_NONE:
+		p->has_rx_tstamp = false;
+		rxx_frm_ctl.u64 = cvmx_read_csr(p->agl + AGL_GMX_RX_FRM_CTL);
+		rxx_frm_ctl.s.ptp_mode = 0;
+		cvmx_write_csr(p->agl + AGL_GMX_RX_FRM_CTL, rxx_frm_ctl.u64);
+		break;
+	case HWTSTAMP_FILTER_ALL:
+	case HWTSTAMP_FILTER_SOME:
+	case HWTSTAMP_FILTER_PTP_V1_L4_EVENT:
+	case HWTSTAMP_FILTER_PTP_V1_L4_SYNC:
+	case HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_L4_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_L4_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_L2_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_L2_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_L2_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_DELAY_REQ:
+	case HWTSTAMP_FILTER_NTP_ALL:
+		p->has_rx_tstamp = have_hw_timestamps;
+		config.rx_filter = HWTSTAMP_FILTER_ALL;
+		if (p->has_rx_tstamp) {
+			rxx_frm_ctl.u64 = cvmx_read_csr(p->agl + AGL_GMX_RX_FRM_CTL);
+			rxx_frm_ctl.s.ptp_mode = 1;
+			cvmx_write_csr(p->agl + AGL_GMX_RX_FRM_CTL, rxx_frm_ctl.u64);
+		}
+		break;
+	default:
+		return -ERANGE;
+	}
+
+	if (copy_to_user(rq->ifr_data, &config, sizeof(config)))
+		return -EFAULT;
+
+	return 0;
+}
+
+static int octeon_mgmt_ioctl(struct net_device *netdev,
+			     struct ifreq *rq, int cmd)
+{
+	switch (cmd) {
+	case SIOCSHWTSTAMP:
+		return octeon_mgmt_ioctl_hwtstamp(netdev, rq, cmd);
+	default:
+		return phy_do_ioctl(netdev, rq, cmd);
+	}
+}
+
+static void octeon_mgmt_disable_link(struct octeon_mgmt *p)
+{
+	union cvmx_agl_gmx_prtx_cfg prtx_cfg;
+
+	/* Disable GMX before we make any changes. */
+	prtx_cfg.u64 = cvmx_read_csr(p->agl + AGL_GMX_PRT_CFG);
+	prtx_cfg.s.en = 0;
+	prtx_cfg.s.tx_en = 0;
+	prtx_cfg.s.rx_en = 0;
+	cvmx_write_csr(p->agl + AGL_GMX_PRT_CFG, prtx_cfg.u64);
+
+	if (OCTEON_IS_MODEL(OCTEON_CN6XXX)) {
+		int i;
+		for (i = 0; i < 10; i++) {
+			prtx_cfg.u64 = cvmx_read_csr(p->agl + AGL_GMX_PRT_CFG);
+			if (prtx_cfg.s.tx_idle == 1 || prtx_cfg.s.rx_idle == 1)
+				break;
+			mdelay(1);
+			i++;
+		}
+	}
+}
+
+static void octeon_mgmt_enable_link(struct octeon_mgmt *p)
+{
+	union cvmx_agl_gmx_prtx_cfg prtx_cfg;
+
+	/* Restore the GMX enable state only if link is set */
+	prtx_cfg.u64 = cvmx_read_csr(p->agl + AGL_GMX_PRT_CFG);
+	prtx_cfg.s.tx_en = 1;
+	prtx_cfg.s.rx_en = 1;
+	prtx_cfg.s.en = 1;
+	cvmx_write_csr(p->agl + AGL_GMX_PRT_CFG, prtx_cfg.u64);
+}
+
+static void octeon_mgmt_update_link(struct octeon_mgmt *p)
+{
+	struct net_device *ndev = p->netdev;
+	struct phy_device *phydev = ndev->phydev;
+	union cvmx_agl_gmx_prtx_cfg prtx_cfg;
+
+	prtx_cfg.u64 = cvmx_read_csr(p->agl + AGL_GMX_PRT_CFG);
+
+	if (!phydev->link)
+		prtx_cfg.s.duplex = 1;
+	else
+		prtx_cfg.s.duplex = phydev->duplex;
+
+	switch (phydev->speed) {
+	case 10:
+		prtx_cfg.s.speed = 0;
+		prtx_cfg.s.slottime = 0;
+
+		if (OCTEON_IS_MODEL(OCTEON_CN6XXX)) {
+			prtx_cfg.s.burst = 1;
+			prtx_cfg.s.speed_msb = 1;
+		}
+		break;
+	case 100:
+		prtx_cfg.s.speed = 0;
+		prtx_cfg.s.slottime = 0;
+
+		if (OCTEON_IS_MODEL(OCTEON_CN6XXX)) {
+			prtx_cfg.s.burst = 1;
+			prtx_cfg.s.speed_msb = 0;
+		}
+		break;
+	case 1000:
+		/* 1000 MBits is only supported on 6XXX chips */
+		if (OCTEON_IS_MODEL(OCTEON_CN6XXX)) {
+			prtx_cfg.s.speed = 1;
+			prtx_cfg.s.speed_msb = 0;
+			/* Only matters for half-duplex */
+			prtx_cfg.s.slottime = 1;
+			prtx_cfg.s.burst = phydev->duplex;
+		}
+		break;
+	case 0:  /* No link */
+	default:
+		break;
+	}
+
+	/* Write the new GMX setting with the port still disabled. */
+	cvmx_write_csr(p->agl + AGL_GMX_PRT_CFG, prtx_cfg.u64);
+
+	/* Read GMX CFG again to make sure the config is completed. */
+	prtx_cfg.u64 = cvmx_read_csr(p->agl + AGL_GMX_PRT_CFG);
+
+	if (OCTEON_IS_MODEL(OCTEON_CN6XXX)) {
+		union cvmx_agl_gmx_txx_clk agl_clk;
+		union cvmx_agl_prtx_ctl prtx_ctl;
+
+		prtx_ctl.u64 = cvmx_read_csr(p->agl_prt_ctl);
+		agl_clk.u64 = cvmx_read_csr(p->agl + AGL_GMX_TX_CLK);
+		/* MII (both speeds) and RGMII 1000 speed. */
+		agl_clk.s.clk_cnt = 1;
+		if (prtx_ctl.s.mode == 0) { /* RGMII mode */
+			if (phydev->speed == 10)
+				agl_clk.s.clk_cnt = 50;
+			else if (phydev->speed == 100)
+				agl_clk.s.clk_cnt = 5;
+		}
+		cvmx_write_csr(p->agl + AGL_GMX_TX_CLK, agl_clk.u64);
+	}
+}
+
+static void octeon_mgmt_adjust_link(struct net_device *netdev)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+	struct phy_device *phydev = netdev->phydev;
+	unsigned long flags;
+	int link_changed = 0;
+
+	if (!phydev)
+		return;
+
+	spin_lock_irqsave(&p->lock, flags);
+
+
+	if (!phydev->link && p->last_link)
+		link_changed = -1;
+
+	if (phydev->link &&
+	    (p->last_duplex != phydev->duplex ||
+	     p->last_link != phydev->link ||
+	     p->last_speed != phydev->speed)) {
+		octeon_mgmt_disable_link(p);
+		link_changed = 1;
+		octeon_mgmt_update_link(p);
+		octeon_mgmt_enable_link(p);
+	}
+
+	p->last_link = phydev->link;
+	p->last_speed = phydev->speed;
+	p->last_duplex = phydev->duplex;
+
+	spin_unlock_irqrestore(&p->lock, flags);
+
+	if (link_changed != 0) {
+		if (link_changed > 0)
+			netdev_info(netdev, "Link is up - %d/%s\n",
+				    phydev->speed, phydev->duplex == DUPLEX_FULL ? "Full" : "Half");
+		else
+			netdev_info(netdev, "Link is down\n");
+	}
+}
+
+static int octeon_mgmt_init_phy(struct net_device *netdev)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+	struct phy_device *phydev = NULL;
+
+	if (octeon_is_simulation() || p->phy_np == NULL) {
+		/* No PHYs in the simulator. */
+		netif_carrier_on(netdev);
+		return 0;
+	}
+
+	phydev = of_phy_connect(netdev, p->phy_np,
+				octeon_mgmt_adjust_link, 0,
+				PHY_INTERFACE_MODE_MII);
+
+	if (!phydev)
+		return -EPROBE_DEFER;
+
+	return 0;
+}
+
+static int octeon_mgmt_open(struct net_device *netdev)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+	union cvmx_mixx_ctl mix_ctl;
+	union cvmx_agl_gmx_inf_mode agl_gmx_inf_mode;
+	union cvmx_mixx_oring1 oring1;
+	union cvmx_mixx_iring1 iring1;
+	union cvmx_agl_gmx_rxx_frm_ctl rxx_frm_ctl;
+	union cvmx_mixx_irhwm mix_irhwm;
+	union cvmx_mixx_orhwm mix_orhwm;
+	union cvmx_mixx_intena mix_intena;
+	struct sockaddr sa;
+
+	/* Allocate ring buffers.  */
+	p->tx_ring = kzalloc(ring_size_to_bytes(OCTEON_MGMT_TX_RING_SIZE),
+			     GFP_KERNEL);
+	if (!p->tx_ring)
+		return -ENOMEM;
+	p->tx_ring_handle =
+		dma_map_single(p->dev, p->tx_ring,
+			       ring_size_to_bytes(OCTEON_MGMT_TX_RING_SIZE),
+			       DMA_BIDIRECTIONAL);
+	p->tx_next = 0;
+	p->tx_next_clean = 0;
+	p->tx_current_fill = 0;
+
+
+	p->rx_ring = kzalloc(ring_size_to_bytes(OCTEON_MGMT_RX_RING_SIZE),
+			     GFP_KERNEL);
+	if (!p->rx_ring)
+		goto err_nomem;
+	p->rx_ring_handle =
+		dma_map_single(p->dev, p->rx_ring,
+			       ring_size_to_bytes(OCTEON_MGMT_RX_RING_SIZE),
+			       DMA_BIDIRECTIONAL);
+
+	p->rx_next = 0;
+	p->rx_next_fill = 0;
+	p->rx_current_fill = 0;
+
+	octeon_mgmt_reset_hw(p);
+
+	mix_ctl.u64 = cvmx_read_csr(p->mix + MIX_CTL);
+
+	/* Bring it out of reset if needed. */
+	if (mix_ctl.s.reset) {
+		mix_ctl.s.reset = 0;
+		cvmx_write_csr(p->mix + MIX_CTL, mix_ctl.u64);
+		do {
+			mix_ctl.u64 = cvmx_read_csr(p->mix + MIX_CTL);
+		} while (mix_ctl.s.reset);
+	}
+
+	if (OCTEON_IS_MODEL(OCTEON_CN5XXX)) {
+		agl_gmx_inf_mode.u64 = 0;
+		agl_gmx_inf_mode.s.en = 1;
+		cvmx_write_csr(CVMX_AGL_GMX_INF_MODE, agl_gmx_inf_mode.u64);
+	}
+	if (OCTEON_IS_MODEL(OCTEON_CN56XX_PASS1_X)
+		|| OCTEON_IS_MODEL(OCTEON_CN52XX_PASS1_X)) {
+		/* Force compensation values, as they are not
+		 * determined properly by HW
+		 */
+		union cvmx_agl_gmx_drv_ctl drv_ctl;
+
+		drv_ctl.u64 = cvmx_read_csr(CVMX_AGL_GMX_DRV_CTL);
+		if (p->port) {
+			drv_ctl.s.byp_en1 = 1;
+			drv_ctl.s.nctl1 = 6;
+			drv_ctl.s.pctl1 = 6;
+		} else {
+			drv_ctl.s.byp_en = 1;
+			drv_ctl.s.nctl = 6;
+			drv_ctl.s.pctl = 6;
+		}
+		cvmx_write_csr(CVMX_AGL_GMX_DRV_CTL, drv_ctl.u64);
+	}
+
+	oring1.u64 = 0;
+	oring1.s.obase = p->tx_ring_handle >> 3;
+	oring1.s.osize = OCTEON_MGMT_TX_RING_SIZE;
+	cvmx_write_csr(p->mix + MIX_ORING1, oring1.u64);
+
+	iring1.u64 = 0;
+	iring1.s.ibase = p->rx_ring_handle >> 3;
+	iring1.s.isize = OCTEON_MGMT_RX_RING_SIZE;
+	cvmx_write_csr(p->mix + MIX_IRING1, iring1.u64);
+
+	memcpy(sa.sa_data, netdev->dev_addr, ETH_ALEN);
+	octeon_mgmt_set_mac_address(netdev, &sa);
+
+	octeon_mgmt_change_mtu(netdev, netdev->mtu);
+
+	/* Enable the port HW. Packets are not allowed until
+	 * cvmx_mgmt_port_enable() is called.
+	 */
+	mix_ctl.u64 = 0;
+	mix_ctl.s.crc_strip = 1;    /* Strip the ending CRC */
+	mix_ctl.s.en = 1;           /* Enable the port */
+	mix_ctl.s.nbtarb = 0;       /* Arbitration mode */
+	/* MII CB-request FIFO programmable high watermark */
+	mix_ctl.s.mrq_hwm = 1;
+#ifdef __LITTLE_ENDIAN
+	mix_ctl.s.lendian = 1;
+#endif
+	cvmx_write_csr(p->mix + MIX_CTL, mix_ctl.u64);
+
+	/* Read the PHY to find the mode of the interface. */
+	if (octeon_mgmt_init_phy(netdev)) {
+		dev_err(p->dev, "Cannot initialize PHY on MIX%d.\n", p->port);
+		goto err_noirq;
+	}
+
+	/* Set the mode of the interface, RGMII/MII. */
+	if (OCTEON_IS_MODEL(OCTEON_CN6XXX) && netdev->phydev) {
+		union cvmx_agl_prtx_ctl agl_prtx_ctl;
+		int rgmii_mode =
+			(linkmode_test_bit(ETHTOOL_LINK_MODE_1000baseT_Half_BIT,
+					   netdev->phydev->supported) |
+			 linkmode_test_bit(ETHTOOL_LINK_MODE_1000baseT_Full_BIT,
+					   netdev->phydev->supported)) != 0;
+
+		agl_prtx_ctl.u64 = cvmx_read_csr(p->agl_prt_ctl);
+		agl_prtx_ctl.s.mode = rgmii_mode ? 0 : 1;
+		cvmx_write_csr(p->agl_prt_ctl,	agl_prtx_ctl.u64);
+
+		/* MII clocks counts are based on the 125Mhz
+		 * reference, which has an 8nS period. So our delays
+		 * need to be multiplied by this factor.
+		 */
+#define NS_PER_PHY_CLK 8
+
+		/* Take the DLL and clock tree out of reset */
+		agl_prtx_ctl.u64 = cvmx_read_csr(p->agl_prt_ctl);
+		agl_prtx_ctl.s.clkrst = 0;
+		if (rgmii_mode) {
+			agl_prtx_ctl.s.dllrst = 0;
+			agl_prtx_ctl.s.clktx_byp = 0;
+		}
+		cvmx_write_csr(p->agl_prt_ctl,	agl_prtx_ctl.u64);
+		cvmx_read_csr(p->agl_prt_ctl); /* Force write out before wait */
+
+		/* Wait for the DLL to lock. External 125 MHz
+		 * reference clock must be stable at this point.
+		 */
+		ndelay(256 * NS_PER_PHY_CLK);
+
+		/* Enable the interface */
+		agl_prtx_ctl.u64 = cvmx_read_csr(p->agl_prt_ctl);
+		agl_prtx_ctl.s.enable = 1;
+		cvmx_write_csr(p->agl_prt_ctl, agl_prtx_ctl.u64);
+
+		/* Read the value back to force the previous write */
+		agl_prtx_ctl.u64 = cvmx_read_csr(p->agl_prt_ctl);
+
+		/* Enable the compensation controller */
+		agl_prtx_ctl.s.comp = 1;
+		agl_prtx_ctl.s.drv_byp = 0;
+		cvmx_write_csr(p->agl_prt_ctl,	agl_prtx_ctl.u64);
+		/* Force write out before wait. */
+		cvmx_read_csr(p->agl_prt_ctl);
+
+		/* For compensation state to lock. */
+		ndelay(1040 * NS_PER_PHY_CLK);
+
+		/* Default Interframe Gaps are too small.  Recommended
+		 * workaround is.
+		 *
+		 * AGL_GMX_TX_IFG[IFG1]=14
+		 * AGL_GMX_TX_IFG[IFG2]=10
+		 */
+		cvmx_write_csr(CVMX_AGL_GMX_TX_IFG, 0xae);
+	}
+
+	octeon_mgmt_rx_fill_ring(netdev);
+
+	/* Clear statistics. */
+	/* Clear on read. */
+	cvmx_write_csr(p->agl + AGL_GMX_RX_STATS_CTL, 1);
+	cvmx_write_csr(p->agl + AGL_GMX_RX_STATS_PKTS_DRP, 0);
+	cvmx_write_csr(p->agl + AGL_GMX_RX_STATS_PKTS_BAD, 0);
+
+	cvmx_write_csr(p->agl + AGL_GMX_TX_STATS_CTL, 1);
+	cvmx_write_csr(p->agl + AGL_GMX_TX_STAT0, 0);
+	cvmx_write_csr(p->agl + AGL_GMX_TX_STAT1, 0);
+
+	/* Clear any pending interrupts */
+	cvmx_write_csr(p->mix + MIX_ISR, cvmx_read_csr(p->mix + MIX_ISR));
+
+	if (request_irq(p->irq, octeon_mgmt_interrupt, 0, netdev->name,
+			netdev)) {
+		dev_err(p->dev, "request_irq(%d) failed.\n", p->irq);
+		goto err_noirq;
+	}
+
+	/* Interrupt every single RX packet */
+	mix_irhwm.u64 = 0;
+	mix_irhwm.s.irhwm = 0;
+	cvmx_write_csr(p->mix + MIX_IRHWM, mix_irhwm.u64);
+
+	/* Interrupt when we have 1 or more packets to clean.  */
+	mix_orhwm.u64 = 0;
+	mix_orhwm.s.orhwm = 0;
+	cvmx_write_csr(p->mix + MIX_ORHWM, mix_orhwm.u64);
+
+	/* Enable receive and transmit interrupts */
+	mix_intena.u64 = 0;
+	mix_intena.s.ithena = 1;
+	mix_intena.s.othena = 1;
+	cvmx_write_csr(p->mix + MIX_INTENA, mix_intena.u64);
+
+	/* Enable packet I/O. */
+
+	rxx_frm_ctl.u64 = 0;
+	rxx_frm_ctl.s.ptp_mode = p->has_rx_tstamp ? 1 : 0;
+	rxx_frm_ctl.s.pre_align = 1;
+	/* When set, disables the length check for non-min sized pkts
+	 * with padding in the client data.
+	 */
+	rxx_frm_ctl.s.pad_len = 1;
+	/* When set, disables the length check for VLAN pkts */
+	rxx_frm_ctl.s.vlan_len = 1;
+	/* When set, PREAMBLE checking is  less strict */
+	rxx_frm_ctl.s.pre_free = 1;
+	/* Control Pause Frames can match station SMAC */
+	rxx_frm_ctl.s.ctl_smac = 0;
+	/* Control Pause Frames can match globally assign Multicast address */
+	rxx_frm_ctl.s.ctl_mcst = 1;
+	/* Forward pause information to TX block */
+	rxx_frm_ctl.s.ctl_bck = 1;
+	/* Drop Control Pause Frames */
+	rxx_frm_ctl.s.ctl_drp = 1;
+	/* Strip off the preamble */
+	rxx_frm_ctl.s.pre_strp = 1;
+	/* This port is configured to send PREAMBLE+SFD to begin every
+	 * frame.  GMX checks that the PREAMBLE is sent correctly.
+	 */
+	rxx_frm_ctl.s.pre_chk = 1;
+	cvmx_write_csr(p->agl + AGL_GMX_RX_FRM_CTL, rxx_frm_ctl.u64);
+
+	/* Configure the port duplex, speed and enables */
+	octeon_mgmt_disable_link(p);
+	if (netdev->phydev)
+		octeon_mgmt_update_link(p);
+	octeon_mgmt_enable_link(p);
+
+	p->last_link = 0;
+	p->last_speed = 0;
+	/* PHY is not present in simulator. The carrier is enabled
+	 * while initializing the phy for simulator, leave it enabled.
+	 */
+	if (netdev->phydev) {
+		netif_carrier_off(netdev);
+		phy_start(netdev->phydev);
+	}
+
+	netif_wake_queue(netdev);
+	napi_enable(&p->napi);
+
+	return 0;
+err_noirq:
+	octeon_mgmt_reset_hw(p);
+	dma_unmap_single(p->dev, p->rx_ring_handle,
+			 ring_size_to_bytes(OCTEON_MGMT_RX_RING_SIZE),
+			 DMA_BIDIRECTIONAL);
+	kfree(p->rx_ring);
+err_nomem:
+	dma_unmap_single(p->dev, p->tx_ring_handle,
+			 ring_size_to_bytes(OCTEON_MGMT_TX_RING_SIZE),
+			 DMA_BIDIRECTIONAL);
+	kfree(p->tx_ring);
+	return -ENOMEM;
+}
+
+static int octeon_mgmt_stop(struct net_device *netdev)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+
+	napi_disable(&p->napi);
+	netif_stop_queue(netdev);
+
+	if (netdev->phydev) {
+		phy_stop(netdev->phydev);
+		phy_disconnect(netdev->phydev);
+	}
+
+	netif_carrier_off(netdev);
+
+	octeon_mgmt_reset_hw(p);
+
+	free_irq(p->irq, netdev);
+
+	/* dma_unmap is a nop on Octeon, so just free everything.  */
+	skb_queue_purge(&p->tx_list);
+	skb_queue_purge(&p->rx_list);
+
+	dma_unmap_single(p->dev, p->rx_ring_handle,
+			 ring_size_to_bytes(OCTEON_MGMT_RX_RING_SIZE),
+			 DMA_BIDIRECTIONAL);
+	kfree(p->rx_ring);
+
+	dma_unmap_single(p->dev, p->tx_ring_handle,
+			 ring_size_to_bytes(OCTEON_MGMT_TX_RING_SIZE),
+			 DMA_BIDIRECTIONAL);
+	kfree(p->tx_ring);
+
+	return 0;
+}
+
+static netdev_tx_t
+octeon_mgmt_xmit(struct sk_buff *skb, struct net_device *netdev)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+	union mgmt_port_ring_entry re;
+	unsigned long flags;
+	netdev_tx_t rv = NETDEV_TX_BUSY;
+
+	re.d64 = 0;
+	re.s.tstamp = ((skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) != 0);
+	re.s.len = skb->len;
+	re.s.addr = dma_map_single(p->dev, skb->data,
+				   skb->len,
+				   DMA_TO_DEVICE);
+
+	spin_lock_irqsave(&p->tx_list.lock, flags);
+
+	if (unlikely(p->tx_current_fill >= ring_max_fill(OCTEON_MGMT_TX_RING_SIZE) - 1)) {
+		spin_unlock_irqrestore(&p->tx_list.lock, flags);
+		netif_stop_queue(netdev);
+		spin_lock_irqsave(&p->tx_list.lock, flags);
+	}
+
+	if (unlikely(p->tx_current_fill >=
+		     ring_max_fill(OCTEON_MGMT_TX_RING_SIZE))) {
+		spin_unlock_irqrestore(&p->tx_list.lock, flags);
+		dma_unmap_single(p->dev, re.s.addr, re.s.len,
+				 DMA_TO_DEVICE);
+		goto out;
+	}
+
+	__skb_queue_tail(&p->tx_list, skb);
+
+	/* Put it in the ring.  */
+	p->tx_ring[p->tx_next] = re.d64;
+	p->tx_next = (p->tx_next + 1) % OCTEON_MGMT_TX_RING_SIZE;
+	p->tx_current_fill++;
+
+	spin_unlock_irqrestore(&p->tx_list.lock, flags);
+
+	dma_sync_single_for_device(p->dev, p->tx_ring_handle,
+				   ring_size_to_bytes(OCTEON_MGMT_TX_RING_SIZE),
+				   DMA_BIDIRECTIONAL);
+
+	netdev->stats.tx_packets++;
+	netdev->stats.tx_bytes += skb->len;
+
+	/* Ring the bell.  */
+	cvmx_write_csr(p->mix + MIX_ORING2, 1);
+
+	netif_trans_update(netdev);
+	rv = NETDEV_TX_OK;
+out:
+	octeon_mgmt_update_tx_stats(netdev);
+	return rv;
+}
+
+#ifdef CONFIG_NET_POLL_CONTROLLER
+static void octeon_mgmt_poll_controller(struct net_device *netdev)
+{
+	struct octeon_mgmt *p = netdev_priv(netdev);
+
+	octeon_mgmt_receive_packets(p, 16);
+	octeon_mgmt_update_rx_stats(netdev);
+}
+#endif
+
+static void octeon_mgmt_get_drvinfo(struct net_device *netdev,
+				    struct ethtool_drvinfo *info)
+{
+	strscpy(info->driver, DRV_NAME, sizeof(info->driver));
+}
+
+static int octeon_mgmt_nway_reset(struct net_device *dev)
+{
+	if (!capable(CAP_NET_ADMIN))
+		return -EPERM;
+
+	if (dev->phydev)
+		return phy_start_aneg(dev->phydev);
+
+	return -EOPNOTSUPP;
+}
+
+static const struct ethtool_ops octeon_mgmt_ethtool_ops = {
+	.get_drvinfo = octeon_mgmt_get_drvinfo,
+	.nway_reset = octeon_mgmt_nway_reset,
+	.get_link = ethtool_op_get_link,
+	.get_link_ksettings = phy_ethtool_get_link_ksettings,
+	.set_link_ksettings = phy_ethtool_set_link_ksettings,
+};
+
+static const struct net_device_ops octeon_mgmt_ops = {
+	.ndo_open =			octeon_mgmt_open,
+	.ndo_stop =			octeon_mgmt_stop,
+	.ndo_start_xmit =		octeon_mgmt_xmit,
+	.ndo_set_rx_mode =		octeon_mgmt_set_rx_filtering,
+	.ndo_set_mac_address =		octeon_mgmt_set_mac_address,
+	.ndo_eth_ioctl =			octeon_mgmt_ioctl,
+	.ndo_change_mtu =		octeon_mgmt_change_mtu,
+#ifdef CONFIG_NET_POLL_CONTROLLER
+	.ndo_poll_controller =		octeon_mgmt_poll_controller,
+#endif
+};
+
+static int octeon_mgmt_probe(struct platform_device *pdev)
+{
+	struct net_device *netdev;
+	struct octeon_mgmt *p;
+	const __be32 *data;
+	struct resource *res_mix;
+	struct resource *res_agl;
+	struct resource *res_agl_prt_ctl;
+	int len;
+	int result;
+
+	netdev = alloc_etherdev(sizeof(struct octeon_mgmt));
+	if (netdev == NULL)
+		return -ENOMEM;
+
+	SET_NETDEV_DEV(netdev, &pdev->dev);
+
+	platform_set_drvdata(pdev, netdev);
+	p = netdev_priv(netdev);
+	netif_napi_add_weight(netdev, &p->napi, octeon_mgmt_napi_poll,
+			      OCTEON_MGMT_NAPI_WEIGHT);
+
+	p->netdev = netdev;
+	p->dev = &pdev->dev;
+	p->has_rx_tstamp = false;
+
+	data = of_get_property(pdev->dev.of_node, "cell-index", &len);
+	if (data && len == sizeof(*data)) {
+		p->port = be32_to_cpup(data);
+	} else {
+		dev_err(&pdev->dev, "no 'cell-index' property\n");
+		result = -ENXIO;
+		goto err;
+	}
+
+	snprintf(netdev->name, IFNAMSIZ, "mgmt%d", p->port);
+
+	result = platform_get_irq(pdev, 0);
+	if (result < 0)
+		goto err;
+
+	p->irq = result;
+
+	res_mix = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (res_mix == NULL) {
+		dev_err(&pdev->dev, "no 'reg' resource\n");
+		result = -ENXIO;
+		goto err;
+	}
+
+	res_agl = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+	if (res_agl == NULL) {
+		dev_err(&pdev->dev, "no 'reg' resource\n");
+		result = -ENXIO;
+		goto err;
+	}
+
+	res_agl_prt_ctl = platform_get_resource(pdev, IORESOURCE_MEM, 3);
+	if (res_agl_prt_ctl == NULL) {
+		dev_err(&pdev->dev, "no 'reg' resource\n");
+		result = -ENXIO;
+		goto err;
+	}
+
+	p->mix_phys = res_mix->start;
+	p->mix_size = resource_size(res_mix);
+	p->agl_phys = res_agl->start;
+	p->agl_size = resource_size(res_agl);
+	p->agl_prt_ctl_phys = res_agl_prt_ctl->start;
+	p->agl_prt_ctl_size = resource_size(res_agl_prt_ctl);
+
+
+	if (!devm_request_mem_region(&pdev->dev, p->mix_phys, p->mix_size,
+				     res_mix->name)) {
+		dev_err(&pdev->dev, "request_mem_region (%s) failed\n",
+			res_mix->name);
+		result = -ENXIO;
+		goto err;
+	}
+
+	if (!devm_request_mem_region(&pdev->dev, p->agl_phys, p->agl_size,
+				     res_agl->name)) {
+		result = -ENXIO;
+		dev_err(&pdev->dev, "request_mem_region (%s) failed\n",
+			res_agl->name);
+		goto err;
+	}
+
+	if (!devm_request_mem_region(&pdev->dev, p->agl_prt_ctl_phys,
+				     p->agl_prt_ctl_size, res_agl_prt_ctl->name)) {
+		result = -ENXIO;
+		dev_err(&pdev->dev, "request_mem_region (%s) failed\n",
+			res_agl_prt_ctl->name);
+		goto err;
+	}
+
+	p->mix = (u64)devm_ioremap(&pdev->dev, p->mix_phys, p->mix_size);
+	p->agl = (u64)devm_ioremap(&pdev->dev, p->agl_phys, p->agl_size);
+	p->agl_prt_ctl = (u64)devm_ioremap(&pdev->dev, p->agl_prt_ctl_phys,
+					   p->agl_prt_ctl_size);
+	if (!p->mix || !p->agl || !p->agl_prt_ctl) {
+		dev_err(&pdev->dev, "failed to map I/O memory\n");
+		result = -ENOMEM;
+		goto err;
+	}
+
+	spin_lock_init(&p->lock);
+
+	skb_queue_head_init(&p->tx_list);
+	skb_queue_head_init(&p->rx_list);
+	tasklet_setup(&p->tx_clean_tasklet,
+		      octeon_mgmt_clean_tx_tasklet);
+
+	netdev->priv_flags |= IFF_UNICAST_FLT;
+
+	netdev->netdev_ops = &octeon_mgmt_ops;
+	netdev->ethtool_ops = &octeon_mgmt_ethtool_ops;
+
+	netdev->min_mtu = 64 - OCTEON_MGMT_RX_HEADROOM;
+	netdev->max_mtu = 16383 - OCTEON_MGMT_RX_HEADROOM - VLAN_HLEN;
+
+	result = of_get_ethdev_address(pdev->dev.of_node, netdev);
+	if (result)
+		eth_hw_addr_random(netdev);
+
+	p->phy_np = of_parse_phandle(pdev->dev.of_node, "phy-handle", 0);
+
+	result = dma_coerce_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
+	if (result)
+		goto err;
+
+	netif_carrier_off(netdev);
+	result = register_netdev(netdev);
+	if (result)
+		goto err;
+
+	return 0;
+
+err:
+	of_node_put(p->phy_np);
+	free_netdev(netdev);
+	return result;
+}
+
+static int octeon_mgmt_remove(struct platform_device *pdev)
+{
+	struct net_device *netdev = platform_get_drvdata(pdev);
+	struct octeon_mgmt *p = netdev_priv(netdev);
+
+	unregister_netdev(netdev);
+	of_node_put(p->phy_np);
+	free_netdev(netdev);
+	return 0;
+}
+
+static const struct of_device_id octeon_mgmt_match[] = {
+	{
+		.compatible = "cavium,octeon-5750-mix",
+	},
+	{},
+};
+MODULE_DEVICE_TABLE(of, octeon_mgmt_match);
+
+static struct platform_driver octeon_mgmt_driver = {
+	.driver = {
+		.name		= "octeon_mgmt",
+		.of_match_table = octeon_mgmt_match,
+	},
+	.probe		= octeon_mgmt_probe,
+	.remove		= octeon_mgmt_remove,
+};
+
+module_platform_driver(octeon_mgmt_driver);
+
+MODULE_SOFTDEP("pre: mdio-cavium");
+MODULE_DESCRIPTION(DRV_DESCRIPTION);
+MODULE_AUTHOR("David Daney");
+MODULE_LICENSE("GPL");
diff --git a/drivers/net/ethernet/cavium/thunder/Makefile b/drivers/net/ethernet/cavium/thunder/Makefile
new file mode 100644
index 000000000..2fc6142d1
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/Makefile
@@ -0,0 +1,13 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# Makefile for Cavium's Thunder ethernet device
+#
+
+obj-$(CONFIG_THUNDER_NIC_RGX) += thunder_xcv.o
+obj-$(CONFIG_THUNDER_NIC_BGX) += thunder_bgx.o
+obj-$(CONFIG_THUNDER_NIC_PF) += nicpf.o
+obj-$(CONFIG_THUNDER_NIC_VF) += nicvf.o
+
+nicpf-y := nic_main.o
+nicvf-y := nicvf_main.o nicvf_queues.o
+nicvf-y += nicvf_ethtool.o
diff --git a/drivers/net/ethernet/cavium/thunder/nic.h b/drivers/net/ethernet/cavium/thunder/nic.h
new file mode 100644
index 000000000..090d6b839
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/nic.h
@@ -0,0 +1,638 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (C) 2015 Cavium, Inc.
+ */
+
+#ifndef NIC_H
+#define	NIC_H
+
+#include <linux/netdevice.h>
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include "thunder_bgx.h"
+
+/* PCI device IDs */
+#define	PCI_DEVICE_ID_THUNDER_NIC_PF		0xA01E
+#define	PCI_DEVICE_ID_THUNDER_PASS1_NIC_VF	0x0011
+#define	PCI_DEVICE_ID_THUNDER_NIC_VF		0xA034
+#define	PCI_DEVICE_ID_THUNDER_BGX		0xA026
+
+/* Subsystem device IDs */
+#define PCI_SUBSYS_DEVID_88XX_NIC_PF		0xA11E
+#define PCI_SUBSYS_DEVID_81XX_NIC_PF		0xA21E
+#define PCI_SUBSYS_DEVID_83XX_NIC_PF		0xA31E
+
+#define PCI_SUBSYS_DEVID_88XX_PASS1_NIC_VF	0xA11E
+#define PCI_SUBSYS_DEVID_88XX_NIC_VF		0xA134
+#define PCI_SUBSYS_DEVID_81XX_NIC_VF		0xA234
+#define PCI_SUBSYS_DEVID_83XX_NIC_VF		0xA334
+
+
+/* PCI BAR nos */
+#define	PCI_CFG_REG_BAR_NUM		0
+#define	PCI_MSIX_REG_BAR_NUM		4
+
+/* NIC SRIOV VF count */
+#define	MAX_NUM_VFS_SUPPORTED		128
+#define	DEFAULT_NUM_VF_ENABLED		8
+
+#define	NIC_TNS_BYPASS_MODE		0
+#define	NIC_TNS_MODE			1
+
+/* NIC priv flags */
+#define	NIC_SRIOV_ENABLED		BIT(0)
+
+/* Min/Max packet size */
+#define	NIC_HW_MIN_FRS			64
+#define	NIC_HW_MAX_FRS			9190 /* Excluding L2 header and FCS */
+
+/* Max pkinds */
+#define	NIC_MAX_PKIND			16
+
+/* Max when CPI_ALG is IP diffserv */
+#define	NIC_MAX_CPI_PER_LMAC		64
+
+/* NIC VF Interrupts */
+#define	NICVF_INTR_CQ			0
+#define	NICVF_INTR_SQ			1
+#define	NICVF_INTR_RBDR			2
+#define	NICVF_INTR_PKT_DROP		3
+#define	NICVF_INTR_TCP_TIMER		4
+#define	NICVF_INTR_MBOX			5
+#define	NICVF_INTR_QS_ERR		6
+
+#define	NICVF_INTR_CQ_SHIFT		0
+#define	NICVF_INTR_SQ_SHIFT		8
+#define	NICVF_INTR_RBDR_SHIFT		16
+#define	NICVF_INTR_PKT_DROP_SHIFT	20
+#define	NICVF_INTR_TCP_TIMER_SHIFT	21
+#define	NICVF_INTR_MBOX_SHIFT		22
+#define	NICVF_INTR_QS_ERR_SHIFT		23
+
+#define	NICVF_INTR_CQ_MASK		(0xFF << NICVF_INTR_CQ_SHIFT)
+#define	NICVF_INTR_SQ_MASK		(0xFF << NICVF_INTR_SQ_SHIFT)
+#define	NICVF_INTR_RBDR_MASK		(0x03 << NICVF_INTR_RBDR_SHIFT)
+#define	NICVF_INTR_PKT_DROP_MASK	BIT(NICVF_INTR_PKT_DROP_SHIFT)
+#define	NICVF_INTR_TCP_TIMER_MASK	BIT(NICVF_INTR_TCP_TIMER_SHIFT)
+#define	NICVF_INTR_MBOX_MASK		BIT(NICVF_INTR_MBOX_SHIFT)
+#define	NICVF_INTR_QS_ERR_MASK		BIT(NICVF_INTR_QS_ERR_SHIFT)
+
+/* MSI-X interrupts */
+#define	NIC_PF_MSIX_VECTORS		10
+#define	NIC_VF_MSIX_VECTORS		20
+
+#define NIC_PF_INTR_ID_ECC0_SBE		0
+#define NIC_PF_INTR_ID_ECC0_DBE		1
+#define NIC_PF_INTR_ID_ECC1_SBE		2
+#define NIC_PF_INTR_ID_ECC1_DBE		3
+#define NIC_PF_INTR_ID_ECC2_SBE		4
+#define NIC_PF_INTR_ID_ECC2_DBE		5
+#define NIC_PF_INTR_ID_ECC3_SBE		6
+#define NIC_PF_INTR_ID_ECC3_DBE		7
+#define NIC_PF_INTR_ID_MBOX0		8
+#define NIC_PF_INTR_ID_MBOX1		9
+
+/* Minimum FIFO level before all packets for the CQ are dropped
+ *
+ * This value ensures that once a packet has been "accepted"
+ * for reception it will not get dropped due to non-availability
+ * of CQ descriptor. An errata in HW mandates this value to be
+ * atleast 0x100.
+ */
+#define NICPF_CQM_MIN_DROP_LEVEL       0x100
+
+/* Global timer for CQ timer thresh interrupts
+ * Calculated for SCLK of 700Mhz
+ * value written should be a 1/16th of what is expected
+ *
+ * 1 tick per 0.025usec
+ */
+#define NICPF_CLK_PER_INT_TICK		1
+
+/* Time to wait before we decide that a SQ is stuck.
+ *
+ * Since both pkt rx and tx notifications are done with same CQ,
+ * when packets are being received at very high rate (eg: L2 forwarding)
+ * then freeing transmitted skbs will be delayed and watchdog
+ * will kick in, resetting interface. Hence keeping this value high.
+ */
+#define	NICVF_TX_TIMEOUT		(50 * HZ)
+
+struct nicvf_cq_poll {
+	struct  nicvf *nicvf;
+	u8	cq_idx;		/* Completion queue index */
+	struct	napi_struct napi;
+};
+
+#define NIC_MAX_RSS_HASH_BITS		8
+#define NIC_MAX_RSS_IDR_TBL_SIZE	(1 << NIC_MAX_RSS_HASH_BITS)
+#define RSS_HASH_KEY_SIZE		5 /* 320 bit key */
+
+struct nicvf_rss_info {
+	bool enable;
+#define	RSS_L2_EXTENDED_HASH_ENA	BIT(0)
+#define	RSS_IP_HASH_ENA			BIT(1)
+#define	RSS_TCP_HASH_ENA		BIT(2)
+#define	RSS_TCP_SYN_DIS			BIT(3)
+#define	RSS_UDP_HASH_ENA		BIT(4)
+#define RSS_L4_EXTENDED_HASH_ENA	BIT(5)
+#define	RSS_ROCE_ENA			BIT(6)
+#define	RSS_L3_BI_DIRECTION_ENA		BIT(7)
+#define	RSS_L4_BI_DIRECTION_ENA		BIT(8)
+	u64 cfg;
+	u8  hash_bits;
+	u16 rss_size;
+	u8  ind_tbl[NIC_MAX_RSS_IDR_TBL_SIZE];
+	u64 key[RSS_HASH_KEY_SIZE];
+} ____cacheline_aligned_in_smp;
+
+struct nicvf_pfc {
+	u8    autoneg;
+	u8    fc_rx;
+	u8    fc_tx;
+};
+
+enum rx_stats_reg_offset {
+	RX_OCTS = 0x0,
+	RX_UCAST = 0x1,
+	RX_BCAST = 0x2,
+	RX_MCAST = 0x3,
+	RX_RED = 0x4,
+	RX_RED_OCTS = 0x5,
+	RX_ORUN = 0x6,
+	RX_ORUN_OCTS = 0x7,
+	RX_FCS = 0x8,
+	RX_L2ERR = 0x9,
+	RX_DRP_BCAST = 0xa,
+	RX_DRP_MCAST = 0xb,
+	RX_DRP_L3BCAST = 0xc,
+	RX_DRP_L3MCAST = 0xd,
+	RX_STATS_ENUM_LAST,
+};
+
+enum tx_stats_reg_offset {
+	TX_OCTS = 0x0,
+	TX_UCAST = 0x1,
+	TX_BCAST = 0x2,
+	TX_MCAST = 0x3,
+	TX_DROP = 0x4,
+	TX_STATS_ENUM_LAST,
+};
+
+struct nicvf_hw_stats {
+	u64 rx_bytes;
+	u64 rx_frames;
+	u64 rx_ucast_frames;
+	u64 rx_bcast_frames;
+	u64 rx_mcast_frames;
+	u64 rx_drops;
+	u64 rx_drop_red;
+	u64 rx_drop_red_bytes;
+	u64 rx_drop_overrun;
+	u64 rx_drop_overrun_bytes;
+	u64 rx_drop_bcast;
+	u64 rx_drop_mcast;
+	u64 rx_drop_l3_bcast;
+	u64 rx_drop_l3_mcast;
+	u64 rx_fcs_errors;
+	u64 rx_l2_errors;
+
+	u64 tx_bytes;
+	u64 tx_frames;
+	u64 tx_ucast_frames;
+	u64 tx_bcast_frames;
+	u64 tx_mcast_frames;
+	u64 tx_drops;
+};
+
+struct nicvf_drv_stats {
+	/* CQE Rx errs */
+	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;
+
+	/* CQE Tx errs */
+	u64 tx_desc_fault;
+	u64 tx_hdr_cons_err;
+	u64 tx_subdesc_err;
+	u64 tx_max_size_exceeded;
+	u64 tx_imm_size_oflow;
+	u64 tx_data_seq_err;
+	u64 tx_mem_seq_err;
+	u64 tx_lock_viol;
+	u64 tx_data_fault;
+	u64 tx_tstmp_conflict;
+	u64 tx_tstmp_timeout;
+	u64 tx_mem_fault;
+	u64 tx_csum_overlap;
+	u64 tx_csum_overflow;
+
+	/* driver debug stats */
+	u64 tx_tso;
+	u64 tx_timeout;
+	u64 txq_stop;
+	u64 txq_wake;
+
+	u64 rcv_buffer_alloc_failures;
+	u64 page_alloc;
+
+	struct u64_stats_sync   syncp;
+};
+
+struct cavium_ptp;
+
+struct xcast_addr_list {
+	int              count;
+	u64              mc[];
+};
+
+struct nicvf_work {
+	struct work_struct     work;
+	u8                     mode;
+	struct xcast_addr_list *mc;
+};
+
+struct nicvf {
+	struct nicvf		*pnicvf;
+	struct net_device	*netdev;
+	struct pci_dev		*pdev;
+	void __iomem		*reg_base;
+	struct bpf_prog         *xdp_prog;
+#define	MAX_QUEUES_PER_QSET			8
+	struct queue_set	*qs;
+	void			*iommu_domain;
+	u8			vf_id;
+	u8			sqs_id;
+	bool                    sqs_mode;
+	bool			hw_tso;
+	bool			t88;
+
+	/* Receive buffer alloc */
+	u32			rb_page_offset;
+	u16			rb_pageref;
+	bool			rb_alloc_fail;
+	bool			rb_work_scheduled;
+	struct page		*rb_page;
+	struct delayed_work	rbdr_work;
+	struct tasklet_struct	rbdr_task;
+
+	/* Secondary Qset */
+	u8			sqs_count;
+#define	MAX_SQS_PER_VF_SINGLE_NODE		5
+#define	MAX_SQS_PER_VF				11
+	struct nicvf		*snicvf[MAX_SQS_PER_VF];
+
+	/* Queue count */
+	u8			rx_queues;
+	u8			tx_queues;
+	u8			xdp_tx_queues;
+	u8			max_queues;
+
+	u8			node;
+	u8			cpi_alg;
+	bool			link_up;
+	u8			mac_type;
+	u8			duplex;
+	u32			speed;
+	bool			tns_mode;
+	bool			loopback_supported;
+	struct nicvf_rss_info	rss_info;
+	struct nicvf_pfc	pfc;
+	struct tasklet_struct	qs_err_task;
+	struct work_struct	reset_task;
+	struct nicvf_work       rx_mode_work;
+	/* spinlock to protect workqueue arguments from concurrent access */
+	spinlock_t              rx_mode_wq_lock;
+	/* workqueue for handling kernel ndo_set_rx_mode() calls */
+	struct workqueue_struct *nicvf_rx_mode_wq;
+	/* mutex to protect VF's mailbox contents from concurrent access */
+	struct mutex            rx_mode_mtx;
+	struct delayed_work	link_change_work;
+	/* PTP timestamp */
+	struct cavium_ptp	*ptp_clock;
+	/* Inbound timestamping is on */
+	bool			hw_rx_tstamp;
+	/* When the packet that requires timestamping is sent, hardware inserts
+	 * two entries to the completion queue.  First is the regular
+	 * CQE_TYPE_SEND entry that signals that the packet was sent.
+	 * The second is CQE_TYPE_SEND_PTP that contains the actual timestamp
+	 * for that packet.
+	 * `ptp_skb` is initialized in the handler for the CQE_TYPE_SEND
+	 * entry and is used and zeroed in the handler for the CQE_TYPE_SEND_PTP
+	 * entry.
+	 * So `ptp_skb` is used to hold the pointer to the packet between
+	 * the calls to CQE_TYPE_SEND and CQE_TYPE_SEND_PTP handlers.
+	 */
+	struct sk_buff		*ptp_skb;
+	/* `tx_ptp_skbs` is set when the hardware is sending a packet that
+	 * requires timestamping.  Cavium hardware can not process more than one
+	 * such packet at once so this is set each time the driver submits
+	 * a packet that requires timestamping to the send queue and clears
+	 * each time it receives the entry on the completion queue saying
+	 * that such packet was sent.
+	 * So `tx_ptp_skbs` prevents driver from submitting more than one
+	 * packet that requires timestamping to the hardware for transmitting.
+	 */
+	atomic_t		tx_ptp_skbs;
+
+	/* Interrupt coalescing settings */
+	u32			cq_coalesce_usecs;
+	u32			msg_enable;
+
+	/* Stats */
+	struct nicvf_hw_stats   hw_stats;
+	struct nicvf_drv_stats  __percpu *drv_stats;
+	struct bgx_stats	bgx_stats;
+
+	/* Napi */
+	struct nicvf_cq_poll	*napi[8];
+
+	/* MSI-X  */
+	u8			num_vec;
+	char			irq_name[NIC_VF_MSIX_VECTORS][IFNAMSIZ + 15];
+	bool			irq_allocated[NIC_VF_MSIX_VECTORS];
+	cpumask_var_t		affinity_mask[NIC_VF_MSIX_VECTORS];
+
+	/* VF <-> PF mailbox communication */
+	bool			pf_acked;
+	bool			pf_nacked;
+	bool			set_mac_pending;
+} ____cacheline_aligned_in_smp;
+
+/* PF <--> VF Mailbox communication
+ * Eight 64bit registers are shared between PF and VF.
+ * Separate set for each VF.
+ * Writing '1' into last register mbx7 means end of message.
+ */
+
+/* PF <--> VF mailbox communication */
+#define	NIC_PF_VF_MAILBOX_SIZE		2
+#define	NIC_MBOX_MSG_TIMEOUT		2000 /* ms */
+
+/* Mailbox message types */
+#define	NIC_MBOX_MSG_READY		0x01	/* Is PF ready to rcv msgs */
+#define	NIC_MBOX_MSG_ACK		0x02	/* ACK the message received */
+#define	NIC_MBOX_MSG_NACK		0x03	/* NACK the message received */
+#define	NIC_MBOX_MSG_QS_CFG		0x04	/* Configure Qset */
+#define	NIC_MBOX_MSG_RQ_CFG		0x05	/* Configure receive queue */
+#define	NIC_MBOX_MSG_SQ_CFG		0x06	/* Configure Send queue */
+#define	NIC_MBOX_MSG_RQ_DROP_CFG	0x07	/* Configure receive queue */
+#define	NIC_MBOX_MSG_SET_MAC		0x08	/* Add MAC ID to DMAC filter */
+#define	NIC_MBOX_MSG_SET_MAX_FRS	0x09	/* Set max frame size */
+#define	NIC_MBOX_MSG_CPI_CFG		0x0A	/* Config CPI, RSSI */
+#define	NIC_MBOX_MSG_RSS_SIZE		0x0B	/* Get RSS indir_tbl size */
+#define	NIC_MBOX_MSG_RSS_CFG		0x0C	/* Config RSS table */
+#define	NIC_MBOX_MSG_RSS_CFG_CONT	0x0D	/* RSS config continuation */
+#define	NIC_MBOX_MSG_RQ_BP_CFG		0x0E	/* RQ backpressure config */
+#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_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_RESET_STAT_COUNTER 0x17	/* Reset statistics counters */
+#define	NIC_MBOX_MSG_PFC		0x18	/* Pause frame control */
+#define	NIC_MBOX_MSG_PTP_CFG		0x19	/* HW packet timestamp */
+#define	NIC_MBOX_MSG_CFG_DONE		0xF0	/* VF configuration done */
+#define	NIC_MBOX_MSG_SHUTDOWN		0xF1	/* VF is being shutdown */
+#define	NIC_MBOX_MSG_RESET_XCAST	0xF2    /* Reset DCAM filtering mode */
+#define	NIC_MBOX_MSG_ADD_MCAST		0xF3    /* Add MAC to DCAM filters */
+#define	NIC_MBOX_MSG_SET_XCAST		0xF4    /* Set MCAST/BCAST RX mode */
+
+struct nic_cfg_msg {
+	u8    msg;
+	u8    vf_id;
+	u8    node_id;
+	u8    tns_mode:1;
+	u8    sqs_mode:1;
+	u8    loopback_supported:1;
+	u8    mac_addr[ETH_ALEN];
+};
+
+/* Qset configuration */
+struct qs_cfg_msg {
+	u8    msg;
+	u8    num;
+	u8    sqs_count;
+	u64   cfg;
+};
+
+/* Receive queue configuration */
+struct rq_cfg_msg {
+	u8    msg;
+	u8    qs_num;
+	u8    rq_num;
+	u64   cfg;
+};
+
+/* Send queue configuration */
+struct sq_cfg_msg {
+	u8    msg;
+	u8    qs_num;
+	u8    sq_num;
+	bool  sqs_mode;
+	u64   cfg;
+};
+
+/* Set VF's MAC address */
+struct set_mac_msg {
+	u8    msg;
+	u8    vf_id;
+	u8    mac_addr[ETH_ALEN];
+};
+
+/* Set Maximum frame size */
+struct set_frs_msg {
+	u8    msg;
+	u8    vf_id;
+	u16   max_frs;
+};
+
+/* Set CPI algorithm type */
+struct cpi_cfg_msg {
+	u8    msg;
+	u8    vf_id;
+	u8    rq_cnt;
+	u8    cpi_alg;
+};
+
+/* Get RSS table size */
+struct rss_sz_msg {
+	u8    msg;
+	u8    vf_id;
+	u16   ind_tbl_size;
+};
+
+/* Set RSS configuration */
+struct rss_cfg_msg {
+	u8    msg;
+	u8    vf_id;
+	u8    hash_bits;
+	u8    tbl_len;
+	u8    tbl_offset;
+#define RSS_IND_TBL_LEN_PER_MBX_MSG	8
+	u8    ind_tbl[RSS_IND_TBL_LEN_PER_MBX_MSG];
+};
+
+struct bgx_stats_msg {
+	u8    msg;
+	u8    vf_id;
+	u8    rx;
+	u8    idx;
+	u64   stats;
+};
+
+/* Physical interface link status */
+struct bgx_link_status {
+	u8    msg;
+	u8    mac_type;
+	u8    link_up;
+	u8    duplex;
+	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;
+};
+
+/* Reset statistics counters */
+struct reset_stat_cfg {
+	u8    msg;
+	/* Bitmap to select NIC_PF_VNIC(vf_id)_RX_STAT(0..13) */
+	u16   rx_stat_mask;
+	/* Bitmap to select NIC_PF_VNIC(vf_id)_TX_STAT(0..4) */
+	u8    tx_stat_mask;
+	/* Bitmap to select NIC_PF_QS(0..127)_RQ(0..7)_STAT(0..1)
+	 * bit14, bit15 NIC_PF_QS(vf_id)_RQ7_STAT(0..1)
+	 * bit12, bit13 NIC_PF_QS(vf_id)_RQ6_STAT(0..1)
+	 * ..
+	 * bit2, bit3 NIC_PF_QS(vf_id)_RQ1_STAT(0..1)
+	 * bit0, bit1 NIC_PF_QS(vf_id)_RQ0_STAT(0..1)
+	 */
+	u16   rq_stat_mask;
+	/* Bitmap to select NIC_PF_QS(0..127)_SQ(0..7)_STAT(0..1)
+	 * bit14, bit15 NIC_PF_QS(vf_id)_SQ7_STAT(0..1)
+	 * bit12, bit13 NIC_PF_QS(vf_id)_SQ6_STAT(0..1)
+	 * ..
+	 * bit2, bit3 NIC_PF_QS(vf_id)_SQ1_STAT(0..1)
+	 * bit0, bit1 NIC_PF_QS(vf_id)_SQ0_STAT(0..1)
+	 */
+	u16   sq_stat_mask;
+};
+
+struct pfc {
+	u8    msg;
+	u8    get; /* Get or set PFC settings */
+	u8    autoneg;
+	u8    fc_rx;
+	u8    fc_tx;
+};
+
+struct set_ptp {
+	u8    msg;
+	bool  enable;
+};
+
+struct xcast {
+	u8    msg;
+	u8    mode;
+	u64   mac:48;
+};
+
+/* 128 bit shared memory between PF and each VF */
+union nic_mbx {
+	struct { u8 msg; }	msg;
+	struct nic_cfg_msg	nic_cfg;
+	struct qs_cfg_msg	qs;
+	struct rq_cfg_msg	rq;
+	struct sq_cfg_msg	sq;
+	struct set_mac_msg	mac;
+	struct set_frs_msg	frs;
+	struct cpi_cfg_msg	cpi_cfg;
+	struct rss_sz_msg	rss_size;
+	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;
+	struct reset_stat_cfg	reset_stat;
+	struct pfc		pfc;
+	struct set_ptp		ptp;
+	struct xcast            xcast;
+};
+
+#define NIC_NODE_ID_MASK	0x03
+#define NIC_NODE_ID_SHIFT	44
+
+static inline int nic_get_node_id(struct pci_dev *pdev)
+{
+	u64 addr = pci_resource_start(pdev, PCI_CFG_REG_BAR_NUM);
+	return ((addr >> NIC_NODE_ID_SHIFT) & NIC_NODE_ID_MASK);
+}
+
+static inline bool pass1_silicon(struct pci_dev *pdev)
+{
+	return (pdev->revision < 8) &&
+		(pdev->subsystem_device == PCI_SUBSYS_DEVID_88XX_NIC_PF);
+}
+
+static inline bool pass2_silicon(struct pci_dev *pdev)
+{
+	return (pdev->revision >= 8) &&
+		(pdev->subsystem_device == PCI_SUBSYS_DEVID_88XX_NIC_PF);
+}
+
+int nicvf_set_real_num_queues(struct net_device *netdev,
+			      int tx_queues, int rx_queues);
+int nicvf_open(struct net_device *netdev);
+int nicvf_stop(struct net_device *netdev);
+int nicvf_send_msg_to_pf(struct nicvf *vf, union nic_mbx *mbx);
+void nicvf_config_rss(struct nicvf *nic);
+void nicvf_set_rss_key(struct nicvf *nic);
+void nicvf_set_ethtool_ops(struct net_device *netdev);
+void nicvf_update_stats(struct nicvf *nic);
+void nicvf_update_lmac_stats(struct nicvf *nic);
+
+#endif /* NIC_H */
diff --git a/drivers/net/ethernet/cavium/thunder/nic_main.c b/drivers/net/ethernet/cavium/thunder/nic_main.c
new file mode 100644
index 000000000..0ec65ec63
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/nic_main.c
@@ -0,0 +1,1416 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2015 Cavium, Inc.
+ */
+
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include <linux/etherdevice.h>
+#include <linux/of.h>
+#include <linux/if_vlan.h>
+
+#include "nic_reg.h"
+#include "nic.h"
+#include "q_struct.h"
+#include "thunder_bgx.h"
+
+#define DRV_NAME	"nicpf"
+#define DRV_VERSION	"1.0"
+
+#define NIC_VF_PER_MBX_REG      64
+
+struct hw_info {
+	u8		bgx_cnt;
+	u8		chans_per_lmac;
+	u8		chans_per_bgx; /* Rx/Tx chans */
+	u8		chans_per_rgx;
+	u8		chans_per_lbk;
+	u16		cpi_cnt;
+	u16		rssi_cnt;
+	u16		rss_ind_tbl_size;
+	u16		tl4_cnt;
+	u16		tl3_cnt;
+	u8		tl2_cnt;
+	u8		tl1_cnt;
+	bool		tl1_per_bgx; /* TL1 per BGX or per LMAC */
+};
+
+struct nicpf {
+	struct pci_dev		*pdev;
+	struct hw_info          *hw;
+	u8			node;
+	unsigned int		flags;
+	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)
+#define	NIC_GET_LMAC_FROM_VF_LMAC_MAP(map)	(map & 0xF)
+	u8			*vf_lmac_map;
+	u16			cpi_base[MAX_NUM_VFS_SUPPORTED];
+	u16			rssi_base[MAX_NUM_VFS_SUPPORTED];
+
+	/* MSI-X */
+	u8			num_vec;
+	unsigned int		irq_allocated[NIC_PF_MSIX_VECTORS];
+	char			irq_name[NIC_PF_MSIX_VECTORS][20];
+};
+
+/* Supported devices */
+static const struct pci_device_id nic_id_table[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_CAVIUM, PCI_DEVICE_ID_THUNDER_NIC_PF) },
+	{ 0, }  /* end of table */
+};
+
+MODULE_AUTHOR("Sunil Goutham");
+MODULE_DESCRIPTION("Cavium Thunder NIC Physical Function Driver");
+MODULE_LICENSE("GPL v2");
+MODULE_VERSION(DRV_VERSION);
+MODULE_DEVICE_TABLE(pci, nic_id_table);
+
+/* The Cavium ThunderX network controller can *only* be found in SoCs
+ * containing the ThunderX ARM64 CPU implementation.  All accesses to the device
+ * registers on this platform are implicitly strongly ordered with respect
+ * to memory accesses. So writeq_relaxed() and readq_relaxed() are safe to use
+ * with no memory barriers in this driver.  The readq()/writeq() functions add
+ * explicit ordering operation which in this case are redundant, and only
+ * add overhead.
+ */
+
+/* Register read/write APIs */
+static void nic_reg_write(struct nicpf *nic, u64 offset, u64 val)
+{
+	writeq_relaxed(val, nic->reg_base + offset);
+}
+
+static u64 nic_reg_read(struct nicpf *nic, u64 offset)
+{
+	return readq_relaxed(nic->reg_base + offset);
+}
+
+/* PF -> VF mailbox communication APIs */
+static void nic_enable_mbx_intr(struct nicpf *nic)
+{
+	int vf_cnt = pci_sriov_get_totalvfs(nic->pdev);
+
+#define INTR_MASK(vfs) ((vfs < 64) ? (BIT_ULL(vfs) - 1) : (~0ull))
+
+	/* Clear it, to avoid spurious interrupts (if any) */
+	nic_reg_write(nic, NIC_PF_MAILBOX_INT, INTR_MASK(vf_cnt));
+
+	/* Enable mailbox interrupt for all VFs */
+	nic_reg_write(nic, NIC_PF_MAILBOX_ENA_W1S, INTR_MASK(vf_cnt));
+	/* One mailbox intr enable reg per 64 VFs */
+	if (vf_cnt > 64) {
+		nic_reg_write(nic, NIC_PF_MAILBOX_INT + sizeof(u64),
+			      INTR_MASK(vf_cnt - 64));
+		nic_reg_write(nic, NIC_PF_MAILBOX_ENA_W1S + sizeof(u64),
+			      INTR_MASK(vf_cnt - 64));
+	}
+}
+
+static void nic_clear_mbx_intr(struct nicpf *nic, int vf, int mbx_reg)
+{
+	nic_reg_write(nic, NIC_PF_MAILBOX_INT + (mbx_reg << 3), BIT_ULL(vf));
+}
+
+static u64 nic_get_mbx_addr(int vf)
+{
+	return NIC_PF_VF_0_127_MAILBOX_0_1 + (vf << NIC_VF_NUM_SHIFT);
+}
+
+/* Send a mailbox message to VF
+ * @vf: vf to which this message to be sent
+ * @mbx: Message to be sent
+ */
+static void nic_send_msg_to_vf(struct nicpf *nic, int vf, union nic_mbx *mbx)
+{
+	void __iomem *mbx_addr = nic->reg_base + nic_get_mbx_addr(vf);
+	u64 *msg = (u64 *)mbx;
+
+	/* In first revision HW, mbox interrupt is triggerred
+	 * when PF writes to MBOX(1), in next revisions when
+	 * PF writes to MBOX(0)
+	 */
+	if (pass1_silicon(nic->pdev)) {
+		/* see the comment for nic_reg_write()/nic_reg_read()
+		 * functions above
+		 */
+		writeq_relaxed(msg[0], mbx_addr);
+		writeq_relaxed(msg[1], mbx_addr + 8);
+	} else {
+		writeq_relaxed(msg[1], mbx_addr + 8);
+		writeq_relaxed(msg[0], mbx_addr);
+	}
+}
+
+/* Responds to VF's READY message with VF's
+ * ID, node, MAC address e.t.c
+ * @vf: VF which sent READY message
+ */
+static void nic_mbx_send_ready(struct nicpf *nic, int vf)
+{
+	union nic_mbx mbx = {};
+	int bgx_idx, lmac;
+	const char *mac;
+
+	mbx.nic_cfg.msg = NIC_MBOX_MSG_READY;
+	mbx.nic_cfg.vf_id = vf;
+
+	mbx.nic_cfg.tns_mode = NIC_TNS_BYPASS_MODE;
+
+	if (vf < nic->num_vf_en) {
+		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 < nic->num_vf_en;
+
+	nic_send_msg_to_vf(nic, vf, &mbx);
+}
+
+/* ACKs VF's mailbox message
+ * @vf: VF to which ACK to be sent
+ */
+static void nic_mbx_send_ack(struct nicpf *nic, int vf)
+{
+	union nic_mbx mbx = {};
+
+	mbx.msg.msg = NIC_MBOX_MSG_ACK;
+	nic_send_msg_to_vf(nic, vf, &mbx);
+}
+
+/* NACKs VF's mailbox message that PF is not able to
+ * complete the action
+ * @vf: VF to which ACK to be sent
+ */
+static void nic_mbx_send_nack(struct nicpf *nic, int vf)
+{
+	union nic_mbx mbx = {};
+
+	mbx.msg.msg = NIC_MBOX_MSG_NACK;
+	nic_send_msg_to_vf(nic, vf, &mbx);
+}
+
+/* Flush all in flight receive packets to memory and
+ * bring down an active RQ
+ */
+static int nic_rcv_queue_sw_sync(struct nicpf *nic)
+{
+	u16 timeout = ~0x00;
+
+	nic_reg_write(nic, NIC_PF_SW_SYNC_RX, 0x01);
+	/* Wait till sync cycle is finished */
+	while (timeout) {
+		if (nic_reg_read(nic, NIC_PF_SW_SYNC_RX_DONE) & 0x1)
+			break;
+		timeout--;
+	}
+	nic_reg_write(nic, NIC_PF_SW_SYNC_RX, 0x00);
+	if (!timeout) {
+		dev_err(&nic->pdev->dev, "Receive queue software sync failed");
+		return 1;
+	}
+	return 0;
+}
+
+/* Get BGX Rx/Tx stats and respond to VF's request */
+static void nic_get_bgx_stats(struct nicpf *nic, struct bgx_stats_msg *bgx)
+{
+	int bgx_idx, lmac;
+	union nic_mbx mbx = {};
+
+	bgx_idx = NIC_GET_BGX_FROM_VF_LMAC_MAP(nic->vf_lmac_map[bgx->vf_id]);
+	lmac = NIC_GET_LMAC_FROM_VF_LMAC_MAP(nic->vf_lmac_map[bgx->vf_id]);
+
+	mbx.bgx_stats.msg = NIC_MBOX_MSG_BGX_STATS;
+	mbx.bgx_stats.vf_id = bgx->vf_id;
+	mbx.bgx_stats.rx = bgx->rx;
+	mbx.bgx_stats.idx = bgx->idx;
+	if (bgx->rx)
+		mbx.bgx_stats.stats = bgx_get_rx_stats(nic->node, bgx_idx,
+							    lmac, bgx->idx);
+	else
+		mbx.bgx_stats.stats = bgx_get_tx_stats(nic->node, bgx_idx,
+							    lmac, bgx->idx);
+	nic_send_msg_to_vf(nic, bgx->vf_id, &mbx);
+}
+
+/* Update hardware min/max frame size */
+static int nic_update_hw_frs(struct nicpf *nic, int new_frs, int vf)
+{
+	int bgx, lmac, lmac_cnt;
+	u64 lmac_credits;
+
+	if ((new_frs > NIC_HW_MAX_FRS) || (new_frs < NIC_HW_MIN_FRS))
+		return 1;
+
+	bgx = 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]);
+	lmac += bgx * MAX_LMAC_PER_BGX;
+
+	new_frs += VLAN_ETH_HLEN + ETH_FCS_LEN + 4;
+
+	/* Update corresponding LMAC credits */
+	lmac_cnt = bgx_get_lmac_count(nic->node, bgx);
+	lmac_credits = nic_reg_read(nic, NIC_PF_LMAC_0_7_CREDIT + (lmac * 8));
+	lmac_credits &= ~(0xFFFFFULL << 12);
+	lmac_credits |= (((((48 * 1024) / lmac_cnt) - new_frs) / 16) << 12);
+	nic_reg_write(nic, NIC_PF_LMAC_0_7_CREDIT + (lmac * 8), lmac_credits);
+
+	/* Enforce MTU in HW
+	 * This config is supported only from 88xx pass 2.0 onwards.
+	 */
+	if (!pass1_silicon(nic->pdev))
+		nic_reg_write(nic,
+			      NIC_PF_LMAC_0_7_CFG2 + (lmac * 8), new_frs);
+	return 0;
+}
+
+/* Set minimum transmit packet size */
+static void nic_set_tx_pkt_pad(struct nicpf *nic, int size)
+{
+	int lmac, max_lmac;
+	u16 sdevid;
+	u64 lmac_cfg;
+
+	/* There is a issue in HW where-in while sending GSO sized
+	 * pkts as part of TSO, if pkt len falls below this size
+	 * NIC will zero PAD packet and also updates IP total length.
+	 * Hence set this value to lessthan min pkt size of MAC+IP+TCP
+	 * headers, BGX will do the padding to transmit 64 byte pkt.
+	 */
+	if (size > 52)
+		size = 52;
+
+	pci_read_config_word(nic->pdev, PCI_SUBSYSTEM_ID, &sdevid);
+	/* 81xx's RGX has only one LMAC */
+	if (sdevid == PCI_SUBSYS_DEVID_81XX_NIC_PF)
+		max_lmac = ((nic->hw->bgx_cnt - 1) * MAX_LMAC_PER_BGX) + 1;
+	else
+		max_lmac = nic->hw->bgx_cnt * MAX_LMAC_PER_BGX;
+
+	for (lmac = 0; lmac < max_lmac; lmac++) {
+		lmac_cfg = nic_reg_read(nic, NIC_PF_LMAC_0_7_CFG | (lmac << 3));
+		lmac_cfg &= ~(0xF << 2);
+		lmac_cfg |= ((size / 4) << 2);
+		nic_reg_write(nic, NIC_PF_LMAC_0_7_CFG | (lmac << 3), lmac_cfg);
+	}
+}
+
+/* Function to check number of LMACs present and set VF::LMAC mapping.
+ * Mapping will be used while initializing channels.
+ */
+static void nic_set_lmac_vf_mapping(struct nicpf *nic)
+{
+	unsigned bgx_map = bgx_get_map(nic->node);
+	int bgx, next_bgx_lmac = 0;
+	int lmac, lmac_cnt = 0;
+	u64 lmac_credit;
+
+	nic->num_vf_en = 0;
+
+	for (bgx = 0; bgx < nic->hw->bgx_cnt; bgx++) {
+		if (!(bgx_map & (1 << bgx)))
+			continue;
+		lmac_cnt = bgx_get_lmac_count(nic->node, bgx);
+		for (lmac = 0; lmac < lmac_cnt; lmac++)
+			nic->vf_lmac_map[next_bgx_lmac++] =
+						NIC_SET_VF_LMAC_MAP(bgx, lmac);
+		nic->num_vf_en += lmac_cnt;
+
+		/* Program LMAC credits */
+		lmac_credit = (1ull << 1); /* channel credit enable */
+		lmac_credit |= (0x1ff << 2); /* Max outstanding pkt count */
+		/* 48KB BGX Tx buffer size, each unit is of size 16bytes */
+		lmac_credit |= (((((48 * 1024) / lmac_cnt) -
+				NIC_HW_MAX_FRS) / 16) << 12);
+		lmac = bgx * MAX_LMAC_PER_BGX;
+		for (; lmac < lmac_cnt + (bgx * MAX_LMAC_PER_BGX); lmac++)
+			nic_reg_write(nic,
+				      NIC_PF_LMAC_0_7_CREDIT + (lmac * 8),
+				      lmac_credit);
+
+		/* On CN81XX there are only 8 VFs but max possible no of
+		 * interfaces are 9.
+		 */
+		if (nic->num_vf_en >= pci_sriov_get_totalvfs(nic->pdev)) {
+			nic->num_vf_en = pci_sriov_get_totalvfs(nic->pdev);
+			break;
+		}
+	}
+}
+
+static void nic_get_hw_info(struct nicpf *nic)
+{
+	u16 sdevid;
+	struct hw_info *hw = nic->hw;
+
+	pci_read_config_word(nic->pdev, PCI_SUBSYSTEM_ID, &sdevid);
+
+	switch (sdevid) {
+	case PCI_SUBSYS_DEVID_88XX_NIC_PF:
+		hw->bgx_cnt = MAX_BGX_PER_CN88XX;
+		hw->chans_per_lmac = 16;
+		hw->chans_per_bgx = 128;
+		hw->cpi_cnt = 2048;
+		hw->rssi_cnt = 4096;
+		hw->rss_ind_tbl_size = NIC_MAX_RSS_IDR_TBL_SIZE;
+		hw->tl3_cnt = 256;
+		hw->tl2_cnt = 64;
+		hw->tl1_cnt = 2;
+		hw->tl1_per_bgx = true;
+		break;
+	case PCI_SUBSYS_DEVID_81XX_NIC_PF:
+		hw->bgx_cnt = MAX_BGX_PER_CN81XX;
+		hw->chans_per_lmac = 8;
+		hw->chans_per_bgx = 32;
+		hw->chans_per_rgx = 8;
+		hw->chans_per_lbk = 24;
+		hw->cpi_cnt = 512;
+		hw->rssi_cnt = 256;
+		hw->rss_ind_tbl_size = 32; /* Max RSSI / Max interfaces */
+		hw->tl3_cnt = 64;
+		hw->tl2_cnt = 16;
+		hw->tl1_cnt = 10;
+		hw->tl1_per_bgx = false;
+		break;
+	case PCI_SUBSYS_DEVID_83XX_NIC_PF:
+		hw->bgx_cnt = MAX_BGX_PER_CN83XX;
+		hw->chans_per_lmac = 8;
+		hw->chans_per_bgx = 32;
+		hw->chans_per_lbk = 64;
+		hw->cpi_cnt = 2048;
+		hw->rssi_cnt = 1024;
+		hw->rss_ind_tbl_size = 64; /* Max RSSI / Max interfaces */
+		hw->tl3_cnt = 256;
+		hw->tl2_cnt = 64;
+		hw->tl1_cnt = 18;
+		hw->tl1_per_bgx = false;
+		break;
+	}
+	hw->tl4_cnt = MAX_QUEUES_PER_QSET * pci_sriov_get_totalvfs(nic->pdev);
+}
+
+#define BGX0_BLOCK 8
+#define BGX1_BLOCK 9
+
+static void nic_init_hw(struct nicpf *nic)
+{
+	int i;
+	u64 cqm_cfg;
+
+	/* Enable NIC HW block */
+	nic_reg_write(nic, NIC_PF_CFG, 0x3);
+
+	/* Enable backpressure */
+	nic_reg_write(nic, NIC_PF_BP_CFG, (1ULL << 6) | 0x03);
+
+	/* TNS and TNS bypass modes are present only on 88xx
+	 * Also offset of this CSR has changed in 81xx and 83xx.
+	 */
+	if (nic->pdev->subsystem_device == PCI_SUBSYS_DEVID_88XX_NIC_PF) {
+		/* Disable TNS mode on both interfaces */
+		nic_reg_write(nic, NIC_PF_INTF_0_1_SEND_CFG,
+			      (NIC_TNS_BYPASS_MODE << 7) |
+			      BGX0_BLOCK | (1ULL << 16));
+		nic_reg_write(nic, NIC_PF_INTF_0_1_SEND_CFG | (1 << 8),
+			      (NIC_TNS_BYPASS_MODE << 7) |
+			      BGX1_BLOCK | (1ULL << 16));
+	} else {
+		/* Configure timestamp generation timeout to 10us */
+		for (i = 0; i < nic->hw->bgx_cnt; i++)
+			nic_reg_write(nic, NIC_PF_INTFX_SEND_CFG | (i << 3),
+				      (1ULL << 16));
+	}
+
+	nic_reg_write(nic, NIC_PF_INTF_0_1_BP_CFG,
+		      (1ULL << 63) | BGX0_BLOCK);
+	nic_reg_write(nic, NIC_PF_INTF_0_1_BP_CFG + (1 << 8),
+		      (1ULL << 63) | BGX1_BLOCK);
+
+	/* PKIND configuration */
+	nic->pkind.minlen = 0;
+	nic->pkind.maxlen = NIC_HW_MAX_FRS + VLAN_ETH_HLEN + ETH_FCS_LEN + 4;
+	nic->pkind.lenerr_en = 1;
+	nic->pkind.rx_hdr = 0;
+	nic->pkind.hdr_sl = 0;
+
+	for (i = 0; i < NIC_MAX_PKIND; i++)
+		nic_reg_write(nic, NIC_PF_PKIND_0_15_CFG | (i << 3),
+			      *(u64 *)&nic->pkind);
+
+	nic_set_tx_pkt_pad(nic, NIC_HW_MIN_FRS);
+
+	/* 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);
+
+	/* Check if HW expected value is higher (could be in future chips) */
+	cqm_cfg = nic_reg_read(nic, NIC_PF_CQM_CFG);
+	if (cqm_cfg < NICPF_CQM_MIN_DROP_LEVEL)
+		nic_reg_write(nic, NIC_PF_CQM_CFG, NICPF_CQM_MIN_DROP_LEVEL);
+}
+
+/* Channel parse index configuration */
+static void nic_config_cpi(struct nicpf *nic, struct cpi_cfg_msg *cfg)
+{
+	struct hw_info *hw = nic->hw;
+	u32 vnic, bgx, lmac, chan;
+	u32 padd, cpi_count = 0;
+	u64 cpi_base, cpi, rssi_base, rssi;
+	u8  qset, rq_idx = 0;
+
+	vnic = cfg->vf_id;
+	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]);
+
+	chan = (lmac * hw->chans_per_lmac) + (bgx * hw->chans_per_bgx);
+	cpi_base = vnic * NIC_MAX_CPI_PER_LMAC;
+	rssi_base = vnic * hw->rss_ind_tbl_size;
+
+	/* Rx channel configuration */
+	nic_reg_write(nic, NIC_PF_CHAN_0_255_RX_BP_CFG | (chan << 3),
+		      (1ull << 63) | (vnic << 0));
+	nic_reg_write(nic, NIC_PF_CHAN_0_255_RX_CFG | (chan << 3),
+		      ((u64)cfg->cpi_alg << 62) | (cpi_base << 48));
+
+	if (cfg->cpi_alg == CPI_ALG_NONE)
+		cpi_count = 1;
+	else if (cfg->cpi_alg == CPI_ALG_VLAN) /* 3 bits of PCP */
+		cpi_count = 8;
+	else if (cfg->cpi_alg == CPI_ALG_VLAN16) /* 3 bits PCP + DEI */
+		cpi_count = 16;
+	else if (cfg->cpi_alg == CPI_ALG_DIFF) /* 6bits DSCP */
+		cpi_count = NIC_MAX_CPI_PER_LMAC;
+
+	/* RSS Qset, Qidx mapping */
+	qset = cfg->vf_id;
+	rssi = rssi_base;
+	for (; rssi < (rssi_base + cfg->rq_cnt); rssi++) {
+		nic_reg_write(nic, NIC_PF_RSSI_0_4097_RQ | (rssi << 3),
+			      (qset << 3) | rq_idx);
+		rq_idx++;
+	}
+
+	rssi = 0;
+	cpi = cpi_base;
+	for (; cpi < (cpi_base + cpi_count); cpi++) {
+		/* Determine port to channel adder */
+		if (cfg->cpi_alg != CPI_ALG_DIFF)
+			padd = cpi % cpi_count;
+		else
+			padd = cpi % 8; /* 3 bits CS out of 6bits DSCP */
+
+		/* Leave RSS_SIZE as '0' to disable RSS */
+		if (pass1_silicon(nic->pdev)) {
+			nic_reg_write(nic, NIC_PF_CPI_0_2047_CFG | (cpi << 3),
+				      (vnic << 24) | (padd << 16) |
+				      (rssi_base + rssi));
+		} else {
+			/* Set MPI_ALG to '0' to disable MCAM parsing */
+			nic_reg_write(nic, NIC_PF_CPI_0_2047_CFG | (cpi << 3),
+				      (padd << 16));
+			/* MPI index is same as CPI if MPI_ALG is not enabled */
+			nic_reg_write(nic, NIC_PF_MPI_0_2047_CFG | (cpi << 3),
+				      (vnic << 24) | (rssi_base + rssi));
+		}
+
+		if ((rssi + 1) >= cfg->rq_cnt)
+			continue;
+
+		if (cfg->cpi_alg == CPI_ALG_VLAN)
+			rssi++;
+		else if (cfg->cpi_alg == CPI_ALG_VLAN16)
+			rssi = ((cpi - cpi_base) & 0xe) >> 1;
+		else if (cfg->cpi_alg == CPI_ALG_DIFF)
+			rssi = ((cpi - cpi_base) & 0x38) >> 3;
+	}
+	nic->cpi_base[cfg->vf_id] = cpi_base;
+	nic->rssi_base[cfg->vf_id] = rssi_base;
+}
+
+/* Responsds to VF with its RSS indirection table size */
+static void nic_send_rss_size(struct nicpf *nic, int vf)
+{
+	union nic_mbx mbx = {};
+
+	mbx.rss_size.msg = NIC_MBOX_MSG_RSS_SIZE;
+	mbx.rss_size.ind_tbl_size = nic->hw->rss_ind_tbl_size;
+	nic_send_msg_to_vf(nic, vf, &mbx);
+}
+
+/* Receive side scaling configuration
+ * configure:
+ * - RSS index
+ * - indir table i.e hash::RQ mapping
+ * - no of hash bits to consider
+ */
+static void nic_config_rss(struct nicpf *nic, struct rss_cfg_msg *cfg)
+{
+	u8  qset, idx = 0;
+	u64 cpi_cfg, cpi_base, rssi_base, rssi;
+	u64 idx_addr;
+
+	rssi_base = nic->rssi_base[cfg->vf_id] + cfg->tbl_offset;
+
+	rssi = rssi_base;
+
+	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++;
+	}
+
+	cpi_base = nic->cpi_base[cfg->vf_id];
+	if (pass1_silicon(nic->pdev))
+		idx_addr = NIC_PF_CPI_0_2047_CFG;
+	else
+		idx_addr = NIC_PF_MPI_0_2047_CFG;
+	cpi_cfg = nic_reg_read(nic, idx_addr | (cpi_base << 3));
+	cpi_cfg &= ~(0xFULL << 20);
+	cpi_cfg |= (cfg->hash_bits << 20);
+	nic_reg_write(nic, idx_addr | (cpi_base << 3), cpi_cfg);
+}
+
+/* 4 level transmit side scheduler configutation
+ * for TNS bypass mode
+ *
+ * Sample configuration for SQ0 on 88xx
+ * VNIC0-SQ0 -> TL4(0)   -> TL3[0]   -> TL2[0]  -> TL1[0] -> BGX0
+ * VNIC1-SQ0 -> TL4(8)   -> TL3[2]   -> TL2[0]  -> TL1[0] -> BGX0
+ * VNIC2-SQ0 -> TL4(16)  -> TL3[4]   -> TL2[1]  -> TL1[0] -> BGX0
+ * VNIC3-SQ0 -> TL4(24)  -> TL3[6]   -> TL2[1]  -> TL1[0] -> BGX0
+ * VNIC4-SQ0 -> TL4(512) -> TL3[128] -> TL2[32] -> TL1[1] -> BGX1
+ * VNIC5-SQ0 -> TL4(520) -> TL3[130] -> TL2[32] -> TL1[1] -> BGX1
+ * 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,
+			       struct sq_cfg_msg *sq)
+{
+	struct hw_info *hw = nic->hw;
+	u32 bgx, lmac, chan;
+	u32 tl2, tl3, tl4;
+	u32 rr_quantum;
+	u8 sq_idx = sq->sq_num;
+	u8 pqs_vnic;
+	int svf;
+
+	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]);
+
+	/* 24 bytes for FCS, IPG and preamble */
+	rr_quantum = ((NIC_HW_MAX_FRS + 24) / 4);
+
+	/* For 88xx 0-511 TL4 transmits via BGX0 and
+	 * 512-1023 TL4s transmit via BGX1.
+	 */
+	if (hw->tl1_per_bgx) {
+		tl4 = bgx * (hw->tl4_cnt / hw->bgx_cnt);
+		if (!sq->sqs_mode) {
+			tl4 += (lmac * MAX_QUEUES_PER_QSET);
+		} else {
+			for (svf = 0; svf < MAX_SQS_PER_VF; svf++) {
+				if (nic->vf_sqs[pqs_vnic][svf] == vnic)
+					break;
+			}
+			tl4 += (MAX_LMAC_PER_BGX * MAX_QUEUES_PER_QSET);
+			tl4 += (lmac * MAX_QUEUES_PER_QSET * MAX_SQS_PER_VF);
+			tl4 += (svf * MAX_QUEUES_PER_QSET);
+		}
+	} else {
+		tl4 = (vnic * MAX_QUEUES_PER_QSET);
+	}
+	tl4 += sq_idx;
+
+	tl3 = tl4 / (hw->tl4_cnt / hw->tl3_cnt);
+	nic_reg_write(nic, NIC_PF_QSET_0_127_SQ_0_7_CFG2 |
+		      ((u64)vnic << NIC_QS_ID_SHIFT) |
+		      ((u32)sq_idx << NIC_Q_NUM_SHIFT), tl4);
+	nic_reg_write(nic, NIC_PF_TL4_0_1023_CFG | (tl4 << 3),
+		      ((u64)vnic << 27) | ((u32)sq_idx << 24) | rr_quantum);
+
+	nic_reg_write(nic, NIC_PF_TL3_0_255_CFG | (tl3 << 3), rr_quantum);
+
+	/* On 88xx 0-127 channels are for BGX0 and
+	 * 127-255 channels for BGX1.
+	 *
+	 * On 81xx/83xx TL3_CHAN reg should be configured with channel
+	 * within LMAC i.e 0-7 and not the actual channel number like on 88xx
+	 */
+	chan = (lmac * hw->chans_per_lmac) + (bgx * hw->chans_per_bgx);
+	if (hw->tl1_per_bgx)
+		nic_reg_write(nic, NIC_PF_TL3_0_255_CHAN | (tl3 << 3), chan);
+	else
+		nic_reg_write(nic, NIC_PF_TL3_0_255_CHAN | (tl3 << 3), 0);
+
+	/* Enable backpressure on the channel */
+	nic_reg_write(nic, NIC_PF_CHAN_0_255_TX_CFG | (chan << 3), 1);
+
+	tl2 = tl3 >> 2;
+	nic_reg_write(nic, NIC_PF_TL3A_0_63_CFG | (tl2 << 3), tl2);
+	nic_reg_write(nic, NIC_PF_TL2_0_63_CFG | (tl2 << 3), rr_quantum);
+	/* No priorities as of now */
+	nic_reg_write(nic, NIC_PF_TL2_0_63_PRI | (tl2 << 3), 0x00);
+
+	/* Unlike 88xx where TL2s 0-31 transmits to TL1 '0' and rest to TL1 '1'
+	 * on 81xx/83xx TL2 needs to be configured to transmit to one of the
+	 * possible LMACs.
+	 *
+	 * This register doesn't exist on 88xx.
+	 */
+	if (!hw->tl1_per_bgx)
+		nic_reg_write(nic, NIC_PF_TL2_LMAC | (tl2 << 3),
+			      lmac + (bgx * MAX_LMAC_PER_BGX));
+}
+
+/* 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 >= nic->num_vf_en)
+		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);
+
+	/* Enable moving average calculation.
+	 * Keep the LVL/AVG delay to HW enforced minimum so that, not too many
+	 * packets sneek in between average calculations.
+	 */
+	nic_reg_write(nic, NIC_PF_CQ_AVG_CFG,
+		      (BIT_ULL(20) | 0x2ull << 14 | 0x1));
+	nic_reg_write(nic, NIC_PF_RRM_AVG_CFG,
+		      (BIT_ULL(20) | 0x3ull << 14 | 0x1));
+
+	return 0;
+}
+
+/* Reset statistics counters */
+static int nic_reset_stat_counters(struct nicpf *nic,
+				   int vf, struct reset_stat_cfg *cfg)
+{
+	int i, stat, qnum;
+	u64 reg_addr;
+
+	for (i = 0; i < RX_STATS_ENUM_LAST; i++) {
+		if (cfg->rx_stat_mask & BIT(i)) {
+			reg_addr = NIC_PF_VNIC_0_127_RX_STAT_0_13 |
+				   (vf << NIC_QS_ID_SHIFT) |
+				   (i << 3);
+			nic_reg_write(nic, reg_addr, 0);
+		}
+	}
+
+	for (i = 0; i < TX_STATS_ENUM_LAST; i++) {
+		if (cfg->tx_stat_mask & BIT(i)) {
+			reg_addr = NIC_PF_VNIC_0_127_TX_STAT_0_4 |
+				   (vf << NIC_QS_ID_SHIFT) |
+				   (i << 3);
+			nic_reg_write(nic, reg_addr, 0);
+		}
+	}
+
+	for (i = 0; i <= 15; i++) {
+		qnum = i >> 1;
+		stat = i & 1 ? 1 : 0;
+		reg_addr = (vf << NIC_QS_ID_SHIFT) |
+			   (qnum << NIC_Q_NUM_SHIFT) | (stat << 3);
+		if (cfg->rq_stat_mask & BIT(i)) {
+			reg_addr |= NIC_PF_QSET_0_127_RQ_0_7_STAT_0_1;
+			nic_reg_write(nic, reg_addr, 0);
+		}
+		if (cfg->sq_stat_mask & BIT(i)) {
+			reg_addr |= NIC_PF_QSET_0_127_SQ_0_7_STAT_0_1;
+			nic_reg_write(nic, reg_addr, 0);
+		}
+	}
+
+	return 0;
+}
+
+static void nic_enable_tunnel_parsing(struct nicpf *nic, int vf)
+{
+	u64 prot_def = (IPV6_PROT << 32) | (IPV4_PROT << 16) | ET_PROT;
+	u64 vxlan_prot_def = (IPV6_PROT_DEF << 32) |
+			      (IPV4_PROT_DEF) << 16 | ET_PROT_DEF;
+
+	/* Configure tunnel parsing parameters */
+	nic_reg_write(nic, NIC_PF_RX_GENEVE_DEF,
+		      (1ULL << 63 | UDP_GENEVE_PORT_NUM));
+	nic_reg_write(nic, NIC_PF_RX_GENEVE_PROT_DEF,
+		      ((7ULL << 61) | prot_def));
+	nic_reg_write(nic, NIC_PF_RX_NVGRE_PROT_DEF,
+		      ((7ULL << 61) | prot_def));
+	nic_reg_write(nic, NIC_PF_RX_VXLAN_DEF_0_1,
+		      ((1ULL << 63) | UDP_VXLAN_PORT_NUM));
+	nic_reg_write(nic, NIC_PF_RX_VXLAN_PROT_DEF,
+		      ((0xfULL << 60) | vxlan_prot_def));
+}
+
+static void nic_enable_vf(struct nicpf *nic, int vf, bool enable)
+{
+	int bgx, lmac;
+
+	nic->vf_enabled[vf] = enable;
+
+	if (vf >= nic->num_vf_en)
+		return;
+
+	bgx = 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]);
+
+	bgx_lmac_rx_tx_enable(nic->node, bgx, lmac, enable);
+}
+
+static void nic_pause_frame(struct nicpf *nic, int vf, struct pfc *cfg)
+{
+	int bgx, lmac;
+	struct pfc pfc;
+	union nic_mbx mbx = {};
+
+	if (vf >= nic->num_vf_en)
+		return;
+	bgx = 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]);
+
+	if (cfg->get) {
+		bgx_lmac_get_pfc(nic->node, bgx, lmac, &pfc);
+		mbx.pfc.msg = NIC_MBOX_MSG_PFC;
+		mbx.pfc.autoneg = pfc.autoneg;
+		mbx.pfc.fc_rx = pfc.fc_rx;
+		mbx.pfc.fc_tx = pfc.fc_tx;
+		nic_send_msg_to_vf(nic, vf, &mbx);
+	} else {
+		bgx_lmac_set_pfc(nic->node, bgx, lmac, cfg);
+		nic_mbx_send_ack(nic, vf);
+	}
+}
+
+/* Enable or disable HW timestamping by BGX for pkts received on a LMAC */
+static void nic_config_timestamp(struct nicpf *nic, int vf, struct set_ptp *ptp)
+{
+	struct pkind_cfg *pkind;
+	u8 lmac, bgx_idx;
+	u64 pkind_val, pkind_idx;
+
+	if (vf >= nic->num_vf_en)
+		return;
+
+	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]);
+
+	pkind_idx = lmac + bgx_idx * MAX_LMAC_PER_BGX;
+	pkind_val = nic_reg_read(nic, NIC_PF_PKIND_0_15_CFG | (pkind_idx << 3));
+	pkind = (struct pkind_cfg *)&pkind_val;
+
+	if (ptp->enable && !pkind->hdr_sl) {
+		/* Skiplen to exclude 8byte timestamp while parsing pkt
+		 * If not configured, will result in L2 errors.
+		 */
+		pkind->hdr_sl = 4;
+		/* Adjust max packet length allowed */
+		pkind->maxlen += (pkind->hdr_sl * 2);
+		bgx_config_timestamping(nic->node, bgx_idx, lmac, true);
+		nic_reg_write(nic, NIC_PF_RX_ETYPE_0_7 | (1 << 3),
+			      (ETYPE_ALG_ENDPARSE << 16) | ETH_P_1588);
+	} else if (!ptp->enable && pkind->hdr_sl) {
+		pkind->maxlen -= (pkind->hdr_sl * 2);
+		pkind->hdr_sl = 0;
+		bgx_config_timestamping(nic->node, bgx_idx, lmac, false);
+		nic_reg_write(nic, NIC_PF_RX_ETYPE_0_7 | (1 << 3),
+			      (ETYPE_ALG_SKIP << 16) | ETH_P_8021Q);
+	}
+
+	nic_reg_write(nic, NIC_PF_PKIND_0_15_CFG | (pkind_idx << 3), pkind_val);
+}
+
+/* Get BGX LMAC link status and update corresponding VF
+ * if there is a change, valid only if internal L2 switch
+ * is not present otherwise VF link is always treated as up
+ */
+static void nic_link_status_get(struct nicpf *nic, u8 vf)
+{
+	union nic_mbx mbx = {};
+	struct bgx_link_status link;
+	u8 bgx, lmac;
+
+	mbx.link_status.msg = NIC_MBOX_MSG_BGX_LINK_CHANGE;
+
+	/* Get BGX, LMAC indices for the VF */
+	bgx = 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]);
+
+	/* Get interface link status */
+	bgx_get_lmac_link_state(nic->node, bgx, lmac, &link);
+
+	/* Send a mbox message to VF with current link status */
+	mbx.link_status.link_up = link.link_up;
+	mbx.link_status.duplex = link.duplex;
+	mbx.link_status.speed = link.speed;
+	mbx.link_status.mac_type = link.mac_type;
+
+	/* reply with link status */
+	nic_send_msg_to_vf(nic, vf, &mbx);
+}
+
+/* Interrupt handler to handle mailbox messages from VFs */
+static void nic_handle_mbx_intr(struct nicpf *nic, int vf)
+{
+	union nic_mbx mbx = {};
+	u64 *mbx_data;
+	u64 mbx_addr;
+	u64 reg_addr;
+	u64 cfg;
+	int bgx, lmac;
+	int i;
+	int ret = 0;
+
+	mbx_addr = nic_get_mbx_addr(vf);
+	mbx_data = (u64 *)&mbx;
+
+	for (i = 0; i < NIC_PF_VF_MAILBOX_SIZE; i++) {
+		*mbx_data = nic_reg_read(nic, mbx_addr);
+		mbx_data++;
+		mbx_addr += sizeof(u64);
+	}
+
+	dev_dbg(&nic->pdev->dev, "%s: Mailbox msg 0x%02x from VF%d\n",
+		__func__, mbx.msg.msg, vf);
+	switch (mbx.msg.msg) {
+	case NIC_MBOX_MSG_READY:
+		nic_mbx_send_ready(nic, vf);
+		return;
+	case NIC_MBOX_MSG_QS_CFG:
+		reg_addr = NIC_PF_QSET_0_127_CFG |
+			   (mbx.qs.num << NIC_QS_ID_SHIFT);
+		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 |
+			   (mbx.rq.qs_num << NIC_QS_ID_SHIFT) |
+			   (mbx.rq.rq_num << NIC_Q_NUM_SHIFT);
+		nic_reg_write(nic, reg_addr, mbx.rq.cfg);
+		/* Enable CQE_RX2_S extension in CQE_RX descriptor.
+		 * This gets appended by default on 81xx/83xx chips,
+		 * for consistency enabling the same on 88xx pass2
+		 * where this is introduced.
+		 */
+		if (pass2_silicon(nic->pdev))
+			nic_reg_write(nic, NIC_PF_RX_CFG, 0x01);
+		if (!pass1_silicon(nic->pdev))
+			nic_enable_tunnel_parsing(nic, vf);
+		break;
+	case NIC_MBOX_MSG_RQ_BP_CFG:
+		reg_addr = NIC_PF_QSET_0_127_RQ_0_7_BP_CFG |
+			   (mbx.rq.qs_num << NIC_QS_ID_SHIFT) |
+			   (mbx.rq.rq_num << NIC_Q_NUM_SHIFT);
+		nic_reg_write(nic, reg_addr, mbx.rq.cfg);
+		break;
+	case NIC_MBOX_MSG_RQ_SW_SYNC:
+		ret = nic_rcv_queue_sw_sync(nic);
+		break;
+	case NIC_MBOX_MSG_RQ_DROP_CFG:
+		reg_addr = NIC_PF_QSET_0_127_RQ_0_7_DROP_CFG |
+			   (mbx.rq.qs_num << NIC_QS_ID_SHIFT) |
+			   (mbx.rq.rq_num << NIC_Q_NUM_SHIFT);
+		nic_reg_write(nic, reg_addr, mbx.rq.cfg);
+		break;
+	case NIC_MBOX_MSG_SQ_CFG:
+		reg_addr = NIC_PF_QSET_0_127_SQ_0_7_CFG |
+			   (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);
+		break;
+	case NIC_MBOX_MSG_SET_MAC:
+		if (vf >= nic->num_vf_en) {
+			ret = -1; /* NACK */
+			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]);
+		bgx_set_lmac_mac(nic->node, bgx, lmac, mbx.mac.mac_addr);
+		break;
+	case NIC_MBOX_MSG_SET_MAX_FRS:
+		ret = nic_update_hw_frs(nic, mbx.frs.max_frs,
+					mbx.frs.vf_id);
+		break;
+	case NIC_MBOX_MSG_CPI_CFG:
+		nic_config_cpi(nic, &mbx.cpi_cfg);
+		break;
+	case NIC_MBOX_MSG_RSS_SIZE:
+		nic_send_rss_size(nic, vf);
+		return;
+	case NIC_MBOX_MSG_RSS_CFG:
+	case NIC_MBOX_MSG_RSS_CFG_CONT:
+		nic_config_rss(nic, &mbx.rss_cfg);
+		break;
+	case NIC_MBOX_MSG_CFG_DONE:
+		/* Last message of VF config msg sequence */
+		nic_enable_vf(nic, vf, true);
+		break;
+	case NIC_MBOX_MSG_SHUTDOWN:
+		/* First msg in VF teardown sequence */
+		if (vf >= nic->num_vf_en)
+			nic->sqs_used[vf - nic->num_vf_en] = false;
+		nic->pqs_vf[vf] = 0;
+		nic_enable_vf(nic, vf, false);
+		break;
+	case NIC_MBOX_MSG_ALLOC_SQS:
+		nic_alloc_sqs(nic, &mbx.sqs_alloc);
+		return;
+	case NIC_MBOX_MSG_NICVF_PTR:
+		nic->nicvf[vf] = mbx.nicvf.nicvf;
+		break;
+	case NIC_MBOX_MSG_PNICVF_PTR:
+		nic_send_pnicvf(nic, vf);
+		return;
+	case NIC_MBOX_MSG_SNICVF_PTR:
+		nic_send_snicvf(nic, &mbx.nicvf);
+		return;
+	case NIC_MBOX_MSG_BGX_STATS:
+		nic_get_bgx_stats(nic, &mbx.bgx_stats);
+		return;
+	case NIC_MBOX_MSG_LOOPBACK:
+		ret = nic_config_loopback(nic, &mbx.lbk);
+		break;
+	case NIC_MBOX_MSG_RESET_STAT_COUNTER:
+		ret = nic_reset_stat_counters(nic, vf, &mbx.reset_stat);
+		break;
+	case NIC_MBOX_MSG_PFC:
+		nic_pause_frame(nic, vf, &mbx.pfc);
+		return;
+	case NIC_MBOX_MSG_PTP_CFG:
+		nic_config_timestamp(nic, vf, &mbx.ptp);
+		break;
+	case NIC_MBOX_MSG_RESET_XCAST:
+		if (vf >= nic->num_vf_en) {
+			ret = -1; /* NACK */
+			break;
+		}
+		bgx = 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]);
+		bgx_reset_xcast_mode(nic->node, bgx, lmac,
+				     vf < NIC_VF_PER_MBX_REG ? vf :
+				     vf - NIC_VF_PER_MBX_REG);
+		break;
+
+	case NIC_MBOX_MSG_ADD_MCAST:
+		if (vf >= nic->num_vf_en) {
+			ret = -1; /* NACK */
+			break;
+		}
+		bgx = 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]);
+		bgx_set_dmac_cam_filter(nic->node, bgx, lmac,
+					mbx.xcast.mac,
+					vf < NIC_VF_PER_MBX_REG ? vf :
+					vf - NIC_VF_PER_MBX_REG);
+		break;
+
+	case NIC_MBOX_MSG_SET_XCAST:
+		if (vf >= nic->num_vf_en) {
+			ret = -1; /* NACK */
+			break;
+		}
+		bgx = 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]);
+		bgx_set_xcast_mode(nic->node, bgx, lmac, mbx.xcast.mode);
+		break;
+	case NIC_MBOX_MSG_BGX_LINK_CHANGE:
+		if (vf >= nic->num_vf_en) {
+			ret = -1; /* NACK */
+			break;
+		}
+		nic_link_status_get(nic, vf);
+		return;
+	default:
+		dev_err(&nic->pdev->dev,
+			"Invalid msg from VF%d, msg 0x%x\n", vf, mbx.msg.msg);
+		break;
+	}
+
+	if (!ret) {
+		nic_mbx_send_ack(nic, vf);
+	} else if (mbx.msg.msg != NIC_MBOX_MSG_READY) {
+		dev_err(&nic->pdev->dev, "NACK for MBOX 0x%02x from VF %d\n",
+			mbx.msg.msg, vf);
+		nic_mbx_send_nack(nic, vf);
+	}
+}
+
+static irqreturn_t nic_mbx_intr_handler(int irq, void *nic_irq)
+{
+	struct nicpf *nic = (struct nicpf *)nic_irq;
+	int mbx;
+	u64 intr;
+	u8  vf;
+
+	if (irq == nic->irq_allocated[NIC_PF_INTR_ID_MBOX0])
+		mbx = 0;
+	else
+		mbx = 1;
+
+	intr = nic_reg_read(nic, NIC_PF_MAILBOX_INT + (mbx << 3));
+	dev_dbg(&nic->pdev->dev, "PF interrupt Mbox%d 0x%llx\n", mbx, intr);
+	for (vf = 0; vf < NIC_VF_PER_MBX_REG; vf++) {
+		if (intr & (1ULL << vf)) {
+			dev_dbg(&nic->pdev->dev, "Intr from VF %d\n",
+				vf + (mbx * NIC_VF_PER_MBX_REG));
+
+			nic_handle_mbx_intr(nic, vf +
+					    (mbx * NIC_VF_PER_MBX_REG));
+			nic_clear_mbx_intr(nic, vf, mbx);
+		}
+	}
+	return IRQ_HANDLED;
+}
+
+static void nic_free_all_interrupts(struct nicpf *nic)
+{
+	int irq;
+
+	for (irq = 0; irq < nic->num_vec; irq++) {
+		if (nic->irq_allocated[irq])
+			free_irq(nic->irq_allocated[irq], nic);
+		nic->irq_allocated[irq] = 0;
+	}
+}
+
+static int nic_register_interrupts(struct nicpf *nic)
+{
+	int i, ret, irq;
+	nic->num_vec = pci_msix_vec_count(nic->pdev);
+
+	/* Enable MSI-X */
+	ret = pci_alloc_irq_vectors(nic->pdev, nic->num_vec, nic->num_vec,
+				    PCI_IRQ_MSIX);
+	if (ret < 0) {
+		dev_err(&nic->pdev->dev,
+			"Request for #%d msix vectors failed, returned %d\n",
+			   nic->num_vec, ret);
+		return ret;
+	}
+
+	/* Register mailbox interrupt handler */
+	for (i = NIC_PF_INTR_ID_MBOX0; i < nic->num_vec; i++) {
+		sprintf(nic->irq_name[i],
+			"NICPF Mbox%d", (i - NIC_PF_INTR_ID_MBOX0));
+
+		irq = pci_irq_vector(nic->pdev, i);
+		ret = request_irq(irq, nic_mbx_intr_handler, 0,
+				  nic->irq_name[i], nic);
+		if (ret)
+			goto fail;
+
+		nic->irq_allocated[i] = irq;
+	}
+
+	/* Enable mailbox interrupt */
+	nic_enable_mbx_intr(nic);
+	return 0;
+
+fail:
+	dev_err(&nic->pdev->dev, "Request irq failed\n");
+	nic_free_all_interrupts(nic);
+	pci_free_irq_vectors(nic->pdev);
+	nic->num_vec = 0;
+	return ret;
+}
+
+static void nic_unregister_interrupts(struct nicpf *nic)
+{
+	nic_free_all_interrupts(nic);
+	pci_free_irq_vectors(nic->pdev);
+	nic->num_vec = 0;
+}
+
+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;
+
+	/* Secondary Qsets are needed only if CPU count is
+	 * morethan MAX_QUEUES_PER_QSET.
+	 */
+	if (num_online_cpus() <= MAX_QUEUES_PER_QSET)
+		return 0;
+
+	/* 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;
+
+	pos = pci_find_ext_capability(pdev, PCI_EXT_CAP_ID_SRIOV);
+	if (!pos) {
+		dev_err(&pdev->dev, "SRIOV capability is not found in PCIe config space\n");
+		return -ENODEV;
+	}
+
+	pci_read_config_word(pdev, (pos + PCI_SRIOV_TOTAL_VF), &total_vf_cnt);
+	if (total_vf_cnt < nic->num_vf_en)
+		nic->num_vf_en = total_vf_cnt;
+
+	if (!total_vf_cnt)
+		return 0;
+
+	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",
+			vf_en);
+		nic->num_vf_en = 0;
+		return err;
+	}
+
+	dev_info(&pdev->dev, "SRIOV enabled, number of VF available %d\n",
+		 vf_en);
+
+	nic->flags |= NIC_SRIOV_ENABLED;
+	return 0;
+}
+
+static int nic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	struct device *dev = &pdev->dev;
+	struct nicpf *nic;
+	u8     max_lmac;
+	int    err;
+
+	BUILD_BUG_ON(sizeof(union nic_mbx) > 16);
+
+	nic = devm_kzalloc(dev, sizeof(*nic), GFP_KERNEL);
+	if (!nic)
+		return -ENOMEM;
+
+	nic->hw = devm_kzalloc(dev, sizeof(struct hw_info), GFP_KERNEL);
+	if (!nic->hw)
+		return -ENOMEM;
+
+	pci_set_drvdata(pdev, nic);
+
+	nic->pdev = pdev;
+
+	err = pci_enable_device(pdev);
+	if (err) {
+		pci_set_drvdata(pdev, NULL);
+		return dev_err_probe(dev, err, "Failed to enable PCI device\n");
+	}
+
+	err = pci_request_regions(pdev, DRV_NAME);
+	if (err) {
+		dev_err(dev, "PCI request regions failed 0x%x\n", err);
+		goto err_disable_device;
+	}
+
+	err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(48));
+	if (err) {
+		dev_err(dev, "Unable to get usable DMA configuration\n");
+		goto err_release_regions;
+	}
+
+	/* MAP PF's configuration registers */
+	nic->reg_base = pcim_iomap(pdev, PCI_CFG_REG_BAR_NUM, 0);
+	if (!nic->reg_base) {
+		dev_err(dev, "Cannot map config register space, aborting\n");
+		err = -ENOMEM;
+		goto err_release_regions;
+	}
+
+	nic->node = nic_get_node_id(pdev);
+
+	/* Get HW capability info */
+	nic_get_hw_info(nic);
+
+	/* Allocate memory for LMAC tracking elements */
+	err = -ENOMEM;
+	max_lmac = nic->hw->bgx_cnt * MAX_LMAC_PER_BGX;
+
+	nic->vf_lmac_map = devm_kmalloc_array(dev, max_lmac, sizeof(u8),
+					      GFP_KERNEL);
+	if (!nic->vf_lmac_map)
+		goto err_release_regions;
+
+	/* Initialize hardware */
+	nic_init_hw(nic);
+
+	nic_set_lmac_vf_mapping(nic);
+
+	/* Register interrupts */
+	err = nic_register_interrupts(nic);
+	if (err)
+		goto err_release_regions;
+
+	/* Configure SRIOV */
+	err = nic_sriov_init(pdev, nic);
+	if (err)
+		goto err_unregister_interrupts;
+
+	return 0;
+
+err_unregister_interrupts:
+	nic_unregister_interrupts(nic);
+err_release_regions:
+	pci_release_regions(pdev);
+err_disable_device:
+	pci_disable_device(pdev);
+	pci_set_drvdata(pdev, NULL);
+	return err;
+}
+
+static void nic_remove(struct pci_dev *pdev)
+{
+	struct nicpf *nic = pci_get_drvdata(pdev);
+
+	if (!nic)
+		return;
+
+	if (nic->flags & NIC_SRIOV_ENABLED)
+		pci_disable_sriov(pdev);
+
+	nic_unregister_interrupts(nic);
+	pci_release_regions(pdev);
+
+	pci_disable_device(pdev);
+	pci_set_drvdata(pdev, NULL);
+}
+
+static struct pci_driver nic_driver = {
+	.name = DRV_NAME,
+	.id_table = nic_id_table,
+	.probe = nic_probe,
+	.remove = nic_remove,
+};
+
+static int __init nic_init_module(void)
+{
+	pr_info("%s, ver %s\n", DRV_NAME, DRV_VERSION);
+
+	return pci_register_driver(&nic_driver);
+}
+
+static void __exit nic_cleanup_module(void)
+{
+	pci_unregister_driver(&nic_driver);
+}
+
+module_init(nic_init_module);
+module_exit(nic_cleanup_module);
diff --git a/drivers/net/ethernet/cavium/thunder/nic_reg.h b/drivers/net/ethernet/cavium/thunder/nic_reg.h
new file mode 100644
index 000000000..b3bd24feb
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/nic_reg.h
@@ -0,0 +1,230 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (C) 2015 Cavium, Inc.
+ */
+
+#ifndef NIC_REG_H
+#define NIC_REG_H
+
+#define   NIC_PF_REG_COUNT			29573
+#define   NIC_VF_REG_COUNT			249
+
+/* Physical function register offsets */
+#define   NIC_PF_CFG				(0x0000)
+#define   NIC_PF_STATUS				(0x0010)
+#define   NIC_PF_INTR_TIMER_CFG			(0x0030)
+#define   NIC_PF_BIST_STATUS			(0x0040)
+#define   NIC_PF_SOFT_RESET			(0x0050)
+#define   NIC_PF_TCP_TIMER			(0x0060)
+#define   NIC_PF_BP_CFG				(0x0080)
+#define   NIC_PF_RRM_CFG			(0x0088)
+#define   NIC_PF_CQM_CFG			(0x00A0)
+#define   NIC_PF_CNM_CF				(0x00A8)
+#define   NIC_PF_CNM_STATUS			(0x00B0)
+#define   NIC_PF_CQ_AVG_CFG			(0x00C0)
+#define   NIC_PF_RRM_AVG_CFG			(0x00C8)
+#define   NIC_PF_INTF_0_1_SEND_CFG		(0x0200)
+#define   NIC_PF_INTF_0_1_BP_CFG		(0x0208)
+#define   NIC_PF_INTF_0_1_BP_DIS_0_1		(0x0210)
+#define   NIC_PF_INTF_0_1_BP_SW_0_1		(0x0220)
+#define   NIC_PF_RBDR_BP_STATE_0_3		(0x0240)
+#define   NIC_PF_MAILBOX_INT			(0x0410)
+#define   NIC_PF_MAILBOX_INT_W1S		(0x0430)
+#define   NIC_PF_MAILBOX_ENA_W1C		(0x0450)
+#define   NIC_PF_MAILBOX_ENA_W1S		(0x0470)
+#define   NIC_PF_RX_ETYPE_0_7			(0x0500)
+#define   NIC_PF_RX_GENEVE_DEF			(0x0580)
+#define    UDP_GENEVE_PORT_NUM				0x17C1ULL
+#define   NIC_PF_RX_GENEVE_PROT_DEF		(0x0588)
+#define    IPV6_PROT					0x86DDULL
+#define    IPV4_PROT					0x800ULL
+#define    ET_PROT					0x6558ULL
+#define   NIC_PF_RX_NVGRE_PROT_DEF		(0x0598)
+#define   NIC_PF_RX_VXLAN_DEF_0_1		(0x05A0)
+#define    UDP_VXLAN_PORT_NUM				0x12B5
+#define   NIC_PF_RX_VXLAN_PROT_DEF		(0x05B0)
+#define    IPV6_PROT_DEF				0x2ULL
+#define    IPV4_PROT_DEF				0x1ULL
+#define    ET_PROT_DEF					0x3ULL
+#define   NIC_PF_RX_CFG				(0x05D0)
+#define   NIC_PF_PKIND_0_15_CFG			(0x0600)
+#define   NIC_PF_ECC0_FLIP0			(0x1000)
+#define   NIC_PF_ECC1_FLIP0			(0x1008)
+#define   NIC_PF_ECC2_FLIP0			(0x1010)
+#define   NIC_PF_ECC3_FLIP0			(0x1018)
+#define   NIC_PF_ECC0_FLIP1			(0x1080)
+#define   NIC_PF_ECC1_FLIP1			(0x1088)
+#define   NIC_PF_ECC2_FLIP1			(0x1090)
+#define   NIC_PF_ECC3_FLIP1			(0x1098)
+#define   NIC_PF_ECC0_CDIS			(0x1100)
+#define   NIC_PF_ECC1_CDIS			(0x1108)
+#define   NIC_PF_ECC2_CDIS			(0x1110)
+#define   NIC_PF_ECC3_CDIS			(0x1118)
+#define   NIC_PF_BIST0_STATUS			(0x1280)
+#define   NIC_PF_BIST1_STATUS			(0x1288)
+#define   NIC_PF_BIST2_STATUS			(0x1290)
+#define   NIC_PF_BIST3_STATUS			(0x1298)
+#define   NIC_PF_ECC0_SBE_INT			(0x2000)
+#define   NIC_PF_ECC0_SBE_INT_W1S		(0x2008)
+#define   NIC_PF_ECC0_SBE_ENA_W1C		(0x2010)
+#define   NIC_PF_ECC0_SBE_ENA_W1S		(0x2018)
+#define   NIC_PF_ECC0_DBE_INT			(0x2100)
+#define   NIC_PF_ECC0_DBE_INT_W1S		(0x2108)
+#define   NIC_PF_ECC0_DBE_ENA_W1C		(0x2110)
+#define   NIC_PF_ECC0_DBE_ENA_W1S		(0x2118)
+#define   NIC_PF_ECC1_SBE_INT			(0x2200)
+#define   NIC_PF_ECC1_SBE_INT_W1S		(0x2208)
+#define   NIC_PF_ECC1_SBE_ENA_W1C		(0x2210)
+#define   NIC_PF_ECC1_SBE_ENA_W1S		(0x2218)
+#define   NIC_PF_ECC1_DBE_INT			(0x2300)
+#define   NIC_PF_ECC1_DBE_INT_W1S		(0x2308)
+#define   NIC_PF_ECC1_DBE_ENA_W1C		(0x2310)
+#define   NIC_PF_ECC1_DBE_ENA_W1S		(0x2318)
+#define   NIC_PF_ECC2_SBE_INT			(0x2400)
+#define   NIC_PF_ECC2_SBE_INT_W1S		(0x2408)
+#define   NIC_PF_ECC2_SBE_ENA_W1C		(0x2410)
+#define   NIC_PF_ECC2_SBE_ENA_W1S		(0x2418)
+#define   NIC_PF_ECC2_DBE_INT			(0x2500)
+#define   NIC_PF_ECC2_DBE_INT_W1S		(0x2508)
+#define   NIC_PF_ECC2_DBE_ENA_W1C		(0x2510)
+#define   NIC_PF_ECC2_DBE_ENA_W1S		(0x2518)
+#define   NIC_PF_ECC3_SBE_INT			(0x2600)
+#define   NIC_PF_ECC3_SBE_INT_W1S		(0x2608)
+#define   NIC_PF_ECC3_SBE_ENA_W1C		(0x2610)
+#define   NIC_PF_ECC3_SBE_ENA_W1S		(0x2618)
+#define   NIC_PF_ECC3_DBE_INT			(0x2700)
+#define   NIC_PF_ECC3_DBE_INT_W1S		(0x2708)
+#define   NIC_PF_ECC3_DBE_ENA_W1C		(0x2710)
+#define   NIC_PF_ECC3_DBE_ENA_W1S		(0x2718)
+#define   NIC_PF_INTFX_SEND_CFG			(0x4000)
+#define   NIC_PF_MCAM_0_191_ENA			(0x100000)
+#define   NIC_PF_MCAM_0_191_M_0_5_DATA		(0x110000)
+#define   NIC_PF_MCAM_CTRL			(0x120000)
+#define   NIC_PF_CPI_0_2047_CFG			(0x200000)
+#define   NIC_PF_MPI_0_2047_CFG			(0x210000)
+#define   NIC_PF_RSSI_0_4097_RQ			(0x220000)
+#define   NIC_PF_LMAC_0_7_CFG			(0x240000)
+#define   NIC_PF_LMAC_0_7_CFG2			(0x240100)
+#define   NIC_PF_LMAC_0_7_SW_XOFF		(0x242000)
+#define   NIC_PF_LMAC_0_7_CREDIT		(0x244000)
+#define   NIC_PF_CHAN_0_255_TX_CFG		(0x400000)
+#define   NIC_PF_CHAN_0_255_RX_CFG		(0x420000)
+#define   NIC_PF_CHAN_0_255_SW_XOFF		(0x440000)
+#define   NIC_PF_CHAN_0_255_CREDIT		(0x460000)
+#define   NIC_PF_CHAN_0_255_RX_BP_CFG		(0x480000)
+#define   NIC_PF_SW_SYNC_RX			(0x490000)
+#define   NIC_PF_SW_SYNC_RX_DONE		(0x490008)
+#define   NIC_PF_TL2_0_63_CFG			(0x500000)
+#define   NIC_PF_TL2_0_63_PRI			(0x520000)
+#define   NIC_PF_TL2_LMAC			(0x540000)
+#define   NIC_PF_TL2_0_63_SH_STATUS		(0x580000)
+#define   NIC_PF_TL3A_0_63_CFG			(0x5F0000)
+#define   NIC_PF_TL3_0_255_CFG			(0x600000)
+#define   NIC_PF_TL3_0_255_CHAN			(0x620000)
+#define   NIC_PF_TL3_0_255_PIR			(0x640000)
+#define   NIC_PF_TL3_0_255_SW_XOFF		(0x660000)
+#define   NIC_PF_TL3_0_255_CNM_RATE		(0x680000)
+#define   NIC_PF_TL3_0_255_SH_STATUS		(0x6A0000)
+#define   NIC_PF_TL4A_0_255_CFG			(0x6F0000)
+#define   NIC_PF_TL4_0_1023_CFG			(0x800000)
+#define   NIC_PF_TL4_0_1023_SW_XOFF		(0x820000)
+#define   NIC_PF_TL4_0_1023_SH_STATUS		(0x840000)
+#define   NIC_PF_TL4A_0_1023_CNM_RATE		(0x880000)
+#define   NIC_PF_TL4A_0_1023_CNM_STATUS		(0x8A0000)
+#define   NIC_PF_VF_0_127_MAILBOX_0_1		(0x20002030)
+#define   NIC_PF_VNIC_0_127_TX_STAT_0_4		(0x20004000)
+#define   NIC_PF_VNIC_0_127_RX_STAT_0_13	(0x20004100)
+#define   NIC_PF_QSET_0_127_LOCK_0_15		(0x20006000)
+#define   NIC_PF_QSET_0_127_CFG			(0x20010000)
+#define   NIC_PF_QSET_0_127_RQ_0_7_CFG		(0x20010400)
+#define   NIC_PF_QSET_0_127_RQ_0_7_DROP_CFG	(0x20010420)
+#define   NIC_PF_QSET_0_127_RQ_0_7_BP_CFG	(0x20010500)
+#define   NIC_PF_QSET_0_127_RQ_0_7_STAT_0_1	(0x20010600)
+#define   NIC_PF_QSET_0_127_SQ_0_7_CFG		(0x20010C00)
+#define   NIC_PF_QSET_0_127_SQ_0_7_CFG2		(0x20010C08)
+#define   NIC_PF_QSET_0_127_SQ_0_7_STAT_0_1	(0x20010D00)
+
+#define   NIC_PF_MSIX_VEC_0_18_ADDR		(0x000000)
+#define   NIC_PF_MSIX_VEC_0_CTL			(0x000008)
+#define   NIC_PF_MSIX_PBA_0			(0x0F0000)
+
+/* Virtual function register offsets */
+#define   NIC_VNIC_CFG				(0x000020)
+#define   NIC_VF_PF_MAILBOX_0_1			(0x000130)
+#define   NIC_VF_INT				(0x000200)
+#define   NIC_VF_INT_W1S			(0x000220)
+#define   NIC_VF_ENA_W1C			(0x000240)
+#define   NIC_VF_ENA_W1S			(0x000260)
+
+#define   NIC_VNIC_RSS_CFG			(0x0020E0)
+#define   NIC_VNIC_RSS_KEY_0_4			(0x002200)
+#define   NIC_VNIC_TX_STAT_0_4			(0x004000)
+#define   NIC_VNIC_RX_STAT_0_13			(0x004100)
+#define   NIC_QSET_RQ_GEN_CFG			(0x010010)
+
+#define   NIC_QSET_CQ_0_7_CFG			(0x010400)
+#define   NIC_QSET_CQ_0_7_CFG2			(0x010408)
+#define   NIC_QSET_CQ_0_7_THRESH		(0x010410)
+#define   NIC_QSET_CQ_0_7_BASE			(0x010420)
+#define   NIC_QSET_CQ_0_7_HEAD			(0x010428)
+#define   NIC_QSET_CQ_0_7_TAIL			(0x010430)
+#define   NIC_QSET_CQ_0_7_DOOR			(0x010438)
+#define   NIC_QSET_CQ_0_7_STATUS		(0x010440)
+#define   NIC_QSET_CQ_0_7_STATUS2		(0x010448)
+#define   NIC_QSET_CQ_0_7_DEBUG			(0x010450)
+
+#define   NIC_QSET_RQ_0_7_CFG			(0x010600)
+#define   NIC_QSET_RQ_0_7_STAT_0_1		(0x010700)
+
+#define   NIC_QSET_SQ_0_7_CFG			(0x010800)
+#define   NIC_QSET_SQ_0_7_THRESH		(0x010810)
+#define   NIC_QSET_SQ_0_7_BASE			(0x010820)
+#define   NIC_QSET_SQ_0_7_HEAD			(0x010828)
+#define   NIC_QSET_SQ_0_7_TAIL			(0x010830)
+#define   NIC_QSET_SQ_0_7_DOOR			(0x010838)
+#define   NIC_QSET_SQ_0_7_STATUS		(0x010840)
+#define   NIC_QSET_SQ_0_7_DEBUG			(0x010848)
+#define   NIC_QSET_SQ_0_7_STAT_0_1		(0x010900)
+
+#define   NIC_QSET_RBDR_0_1_CFG			(0x010C00)
+#define   NIC_QSET_RBDR_0_1_THRESH		(0x010C10)
+#define   NIC_QSET_RBDR_0_1_BASE		(0x010C20)
+#define   NIC_QSET_RBDR_0_1_HEAD		(0x010C28)
+#define   NIC_QSET_RBDR_0_1_TAIL		(0x010C30)
+#define   NIC_QSET_RBDR_0_1_DOOR		(0x010C38)
+#define   NIC_QSET_RBDR_0_1_STATUS0		(0x010C40)
+#define   NIC_QSET_RBDR_0_1_STATUS1		(0x010C48)
+#define   NIC_QSET_RBDR_0_1_PREFETCH_STATUS	(0x010C50)
+
+#define   NIC_VF_MSIX_VECTOR_0_19_ADDR		(0x000000)
+#define   NIC_VF_MSIX_VECTOR_0_19_CTL		(0x000008)
+#define   NIC_VF_MSIX_PBA			(0x0F0000)
+
+/* Offsets within registers */
+#define   NIC_MSIX_VEC_SHIFT			4
+#define   NIC_Q_NUM_SHIFT			18
+#define   NIC_QS_ID_SHIFT			21
+#define   NIC_VF_NUM_SHIFT			21
+
+/* Port kind configuration register */
+struct pkind_cfg {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64 reserved_42_63:22;
+	u64 hdr_sl:5;	/* Header skip length */
+	u64 rx_hdr:3;	/* TNS Receive header present */
+	u64 lenerr_en:1;/* L2 length error check enable */
+	u64 reserved_32_32:1;
+	u64 maxlen:16;	/* Max frame size */
+	u64 minlen:16;	/* Min frame size */
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64 minlen:16;
+	u64 maxlen:16;
+	u64 reserved_32_32:1;
+	u64 lenerr_en:1;
+	u64 rx_hdr:3;
+	u64 hdr_sl:5;
+	u64 reserved_42_63:22;
+#endif
+};
+
+#endif /* NIC_REG_H */
diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_ethtool.c b/drivers/net/ethernet/cavium/thunder/nicvf_ethtool.c
new file mode 100644
index 000000000..e5c71f907
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/nicvf_ethtool.c
@@ -0,0 +1,887 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2015 Cavium, Inc.
+ */
+
+/* ETHTOOL Support for VNIC_VF Device*/
+
+#include <linux/ethtool.h>
+#include <linux/pci.h>
+#include <linux/net_tstamp.h>
+
+#include "nic_reg.h"
+#include "nic.h"
+#include "nicvf_queues.h"
+#include "q_struct.h"
+#include "thunder_bgx.h"
+#include "../common/cavium_ptp.h"
+
+#define DRV_NAME	"nicvf"
+
+struct nicvf_stat {
+	char name[ETH_GSTRING_LEN];
+	unsigned int index;
+};
+
+#define NICVF_HW_STAT(stat) { \
+	.name = #stat, \
+	.index = offsetof(struct nicvf_hw_stats, stat) / sizeof(u64), \
+}
+
+#define NICVF_DRV_STAT(stat) { \
+	.name = #stat, \
+	.index = offsetof(struct nicvf_drv_stats, stat) / sizeof(u64), \
+}
+
+static const struct nicvf_stat nicvf_hw_stats[] = {
+	NICVF_HW_STAT(rx_bytes),
+	NICVF_HW_STAT(rx_frames),
+	NICVF_HW_STAT(rx_ucast_frames),
+	NICVF_HW_STAT(rx_bcast_frames),
+	NICVF_HW_STAT(rx_mcast_frames),
+	NICVF_HW_STAT(rx_drops),
+	NICVF_HW_STAT(rx_drop_red),
+	NICVF_HW_STAT(rx_drop_red_bytes),
+	NICVF_HW_STAT(rx_drop_overrun),
+	NICVF_HW_STAT(rx_drop_overrun_bytes),
+	NICVF_HW_STAT(rx_drop_bcast),
+	NICVF_HW_STAT(rx_drop_mcast),
+	NICVF_HW_STAT(rx_drop_l3_bcast),
+	NICVF_HW_STAT(rx_drop_l3_mcast),
+	NICVF_HW_STAT(rx_fcs_errors),
+	NICVF_HW_STAT(rx_l2_errors),
+	NICVF_HW_STAT(tx_bytes),
+	NICVF_HW_STAT(tx_frames),
+	NICVF_HW_STAT(tx_ucast_frames),
+	NICVF_HW_STAT(tx_bcast_frames),
+	NICVF_HW_STAT(tx_mcast_frames),
+	NICVF_HW_STAT(tx_drops),
+};
+
+static const struct nicvf_stat nicvf_drv_stats[] = {
+	NICVF_DRV_STAT(rx_bgx_truncated_pkts),
+	NICVF_DRV_STAT(rx_jabber_errs),
+	NICVF_DRV_STAT(rx_fcs_errs),
+	NICVF_DRV_STAT(rx_bgx_errs),
+	NICVF_DRV_STAT(rx_prel2_errs),
+	NICVF_DRV_STAT(rx_l2_hdr_malformed),
+	NICVF_DRV_STAT(rx_oversize),
+	NICVF_DRV_STAT(rx_undersize),
+	NICVF_DRV_STAT(rx_l2_len_mismatch),
+	NICVF_DRV_STAT(rx_l2_pclp),
+	NICVF_DRV_STAT(rx_ip_ver_errs),
+	NICVF_DRV_STAT(rx_ip_csum_errs),
+	NICVF_DRV_STAT(rx_ip_hdr_malformed),
+	NICVF_DRV_STAT(rx_ip_payload_malformed),
+	NICVF_DRV_STAT(rx_ip_ttl_errs),
+	NICVF_DRV_STAT(rx_l3_pclp),
+	NICVF_DRV_STAT(rx_l4_malformed),
+	NICVF_DRV_STAT(rx_l4_csum_errs),
+	NICVF_DRV_STAT(rx_udp_len_errs),
+	NICVF_DRV_STAT(rx_l4_port_errs),
+	NICVF_DRV_STAT(rx_tcp_flag_errs),
+	NICVF_DRV_STAT(rx_tcp_offset_errs),
+	NICVF_DRV_STAT(rx_l4_pclp),
+	NICVF_DRV_STAT(rx_truncated_pkts),
+
+	NICVF_DRV_STAT(tx_desc_fault),
+	NICVF_DRV_STAT(tx_hdr_cons_err),
+	NICVF_DRV_STAT(tx_subdesc_err),
+	NICVF_DRV_STAT(tx_max_size_exceeded),
+	NICVF_DRV_STAT(tx_imm_size_oflow),
+	NICVF_DRV_STAT(tx_data_seq_err),
+	NICVF_DRV_STAT(tx_mem_seq_err),
+	NICVF_DRV_STAT(tx_lock_viol),
+	NICVF_DRV_STAT(tx_data_fault),
+	NICVF_DRV_STAT(tx_tstmp_conflict),
+	NICVF_DRV_STAT(tx_tstmp_timeout),
+	NICVF_DRV_STAT(tx_mem_fault),
+	NICVF_DRV_STAT(tx_csum_overlap),
+	NICVF_DRV_STAT(tx_csum_overflow),
+
+	NICVF_DRV_STAT(tx_tso),
+	NICVF_DRV_STAT(tx_timeout),
+	NICVF_DRV_STAT(txq_stop),
+	NICVF_DRV_STAT(txq_wake),
+	NICVF_DRV_STAT(rcv_buffer_alloc_failures),
+	NICVF_DRV_STAT(page_alloc),
+};
+
+static const struct nicvf_stat nicvf_queue_stats[] = {
+	{ "bytes", 0 },
+	{ "frames", 1 },
+};
+
+static const unsigned int nicvf_n_hw_stats = ARRAY_SIZE(nicvf_hw_stats);
+static const unsigned int nicvf_n_drv_stats = ARRAY_SIZE(nicvf_drv_stats);
+static const unsigned int nicvf_n_queue_stats = ARRAY_SIZE(nicvf_queue_stats);
+
+static int nicvf_get_link_ksettings(struct net_device *netdev,
+				    struct ethtool_link_ksettings *cmd)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+	u32 supported, advertising;
+
+	supported = 0;
+	advertising = 0;
+
+	if (!nic->link_up) {
+		cmd->base.duplex = DUPLEX_UNKNOWN;
+		cmd->base.speed = SPEED_UNKNOWN;
+		return 0;
+	}
+
+	switch (nic->speed) {
+	case SPEED_1000:
+		cmd->base.port = PORT_MII | PORT_TP;
+		cmd->base.autoneg = AUTONEG_ENABLE;
+		supported |= SUPPORTED_MII | SUPPORTED_TP;
+		supported |= SUPPORTED_1000baseT_Full |
+				  SUPPORTED_1000baseT_Half |
+				  SUPPORTED_100baseT_Full  |
+				  SUPPORTED_100baseT_Half  |
+				  SUPPORTED_10baseT_Full   |
+				  SUPPORTED_10baseT_Half;
+		supported |= SUPPORTED_Autoneg;
+		advertising |= ADVERTISED_1000baseT_Full |
+				    ADVERTISED_1000baseT_Half |
+				    ADVERTISED_100baseT_Full  |
+				    ADVERTISED_100baseT_Half  |
+				    ADVERTISED_10baseT_Full   |
+				    ADVERTISED_10baseT_Half;
+		break;
+	case SPEED_10000:
+		if (nic->mac_type == BGX_MODE_RXAUI) {
+			cmd->base.port = PORT_TP;
+			supported |= SUPPORTED_TP;
+		} else {
+			cmd->base.port = PORT_FIBRE;
+			supported |= SUPPORTED_FIBRE;
+		}
+		cmd->base.autoneg = AUTONEG_DISABLE;
+		supported |= SUPPORTED_10000baseT_Full;
+		break;
+	case SPEED_40000:
+		cmd->base.port = PORT_FIBRE;
+		cmd->base.autoneg = AUTONEG_DISABLE;
+		supported |= SUPPORTED_FIBRE;
+		supported |= SUPPORTED_40000baseCR4_Full;
+		break;
+	}
+	cmd->base.duplex = nic->duplex;
+	cmd->base.speed = nic->speed;
+
+	ethtool_convert_legacy_u32_to_link_mode(cmd->link_modes.supported,
+						supported);
+	ethtool_convert_legacy_u32_to_link_mode(cmd->link_modes.advertising,
+						advertising);
+
+	return 0;
+}
+
+static u32 nicvf_get_link(struct net_device *netdev)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+
+	return nic->link_up;
+}
+
+static void nicvf_get_drvinfo(struct net_device *netdev,
+			      struct ethtool_drvinfo *info)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+
+	strscpy(info->driver, DRV_NAME, sizeof(info->driver));
+	strscpy(info->bus_info, pci_name(nic->pdev), sizeof(info->bus_info));
+}
+
+static u32 nicvf_get_msglevel(struct net_device *netdev)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+
+	return nic->msg_enable;
+}
+
+static void nicvf_set_msglevel(struct net_device *netdev, u32 lvl)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+
+	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;
+	int sqs;
+
+	if (sset != ETH_SS_STATS)
+		return;
+
+	for (stats = 0; stats < nicvf_n_hw_stats; stats++) {
+		memcpy(data, nicvf_hw_stats[stats].name, ETH_GSTRING_LEN);
+		data += ETH_GSTRING_LEN;
+	}
+
+	for (stats = 0; stats < nicvf_n_drv_stats; stats++) {
+		memcpy(data, nicvf_drv_stats[stats].name, ETH_GSTRING_LEN);
+		data += ETH_GSTRING_LEN;
+	}
+
+	nicvf_get_qset_strings(nic, &data, 0);
+
+	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++) {
+		sprintf(data, "bgx_rxstat%d: ", stats);
+		data += ETH_GSTRING_LEN;
+	}
+
+	for (stats = 0; stats < BGX_TX_STATS_COUNT; stats++) {
+		sprintf(data, "bgx_txstat%d: ", stats);
+		data += ETH_GSTRING_LEN;
+	}
+}
+
+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 +
+		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, tmp_stats;
+	int sqs, cpu;
+
+	nicvf_update_stats(nic);
+
+	/* Update LMAC stats */
+	nicvf_update_lmac_stats(nic);
+
+	for (stat = 0; stat < nicvf_n_hw_stats; stat++)
+		*(data++) = ((u64 *)&nic->hw_stats)
+				[nicvf_hw_stats[stat].index];
+	for (stat = 0; stat < nicvf_n_drv_stats; stat++) {
+		tmp_stats = 0;
+		for_each_possible_cpu(cpu)
+			tmp_stats += ((u64 *)per_cpu_ptr(nic->drv_stats, cpu))
+				     [nicvf_drv_stats[stat].index];
+		*(data++) = tmp_stats;
+	}
+
+	nicvf_get_qset_stats(nic, stats, &data);
+
+	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++)
+		*(data++) = nic->bgx_stats.rx_stats[stat];
+	for (stat = 0; stat < BGX_TX_STATS_COUNT; stat++)
+		*(data++) = nic->bgx_stats.tx_stats[stat];
+}
+
+static int nicvf_get_regs_len(struct net_device *dev)
+{
+	return sizeof(u64) * NIC_VF_REG_COUNT;
+}
+
+static void nicvf_get_regs(struct net_device *dev,
+			   struct ethtool_regs *regs, void *reg)
+{
+	struct nicvf *nic = netdev_priv(dev);
+	u64 *p = (u64 *)reg;
+	u64 reg_offset;
+	int mbox, key, stat, q;
+	int i = 0;
+
+	regs->version = 0;
+	memset(p, 0, NIC_VF_REG_COUNT);
+
+	p[i++] = nicvf_reg_read(nic, NIC_VNIC_CFG);
+	/* Mailbox registers */
+	for (mbox = 0; mbox < NIC_PF_VF_MAILBOX_SIZE; mbox++)
+		p[i++] = nicvf_reg_read(nic,
+					NIC_VF_PF_MAILBOX_0_1 | (mbox << 3));
+
+	p[i++] = nicvf_reg_read(nic, NIC_VF_INT);
+	p[i++] = nicvf_reg_read(nic, NIC_VF_INT_W1S);
+	p[i++] = nicvf_reg_read(nic, NIC_VF_ENA_W1C);
+	p[i++] = nicvf_reg_read(nic, NIC_VF_ENA_W1S);
+	p[i++] = nicvf_reg_read(nic, NIC_VNIC_RSS_CFG);
+
+	for (key = 0; key < RSS_HASH_KEY_SIZE; key++)
+		p[i++] = nicvf_reg_read(nic, NIC_VNIC_RSS_KEY_0_4 | (key << 3));
+
+	/* Tx/Rx statistics */
+	for (stat = 0; stat < TX_STATS_ENUM_LAST; stat++)
+		p[i++] = nicvf_reg_read(nic,
+					NIC_VNIC_TX_STAT_0_4 | (stat << 3));
+
+	for (i = 0; i < RX_STATS_ENUM_LAST; i++)
+		p[i++] = nicvf_reg_read(nic,
+					NIC_VNIC_RX_STAT_0_13 | (stat << 3));
+
+	p[i++] = nicvf_reg_read(nic, NIC_QSET_RQ_GEN_CFG);
+
+	/* All completion queue's registers */
+	for (q = 0; q < MAX_CMP_QUEUES_PER_QS; q++) {
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_CFG, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_CFG2, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_THRESH, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_BASE, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_HEAD, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_TAIL, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_DOOR, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_STATUS, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_STATUS2, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_DEBUG, q);
+	}
+
+	/* All receive queue's registers */
+	for (q = 0; q < MAX_RCV_QUEUES_PER_QS; q++) {
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_RQ_0_7_CFG, q);
+		p[i++] = nicvf_queue_reg_read(nic,
+						  NIC_QSET_RQ_0_7_STAT_0_1, q);
+		reg_offset = NIC_QSET_RQ_0_7_STAT_0_1 | (1 << 3);
+		p[i++] = nicvf_queue_reg_read(nic, reg_offset, q);
+	}
+
+	for (q = 0; q < MAX_SND_QUEUES_PER_QS; q++) {
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_CFG, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_THRESH, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_BASE, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_HEAD, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_TAIL, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_DOOR, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_STATUS, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_DEBUG, q);
+		/* Padding, was NIC_QSET_SQ_0_7_CNM_CHG, which
+		 * produces bus errors when read
+		 */
+		p[i++] = 0;
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_STAT_0_1, q);
+		reg_offset = NIC_QSET_SQ_0_7_STAT_0_1 | (1 << 3);
+		p[i++] = nicvf_queue_reg_read(nic, reg_offset, q);
+	}
+
+	for (q = 0; q < MAX_RCV_BUF_DESC_RINGS_PER_QS; q++) {
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_RBDR_0_1_CFG, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_RBDR_0_1_THRESH, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_RBDR_0_1_BASE, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_RBDR_0_1_HEAD, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_RBDR_0_1_TAIL, q);
+		p[i++] = nicvf_queue_reg_read(nic, NIC_QSET_RBDR_0_1_DOOR, q);
+		p[i++] = nicvf_queue_reg_read(nic,
+					      NIC_QSET_RBDR_0_1_STATUS0, q);
+		p[i++] = nicvf_queue_reg_read(nic,
+					      NIC_QSET_RBDR_0_1_STATUS1, q);
+		reg_offset = NIC_QSET_RBDR_0_1_PREFETCH_STATUS;
+		p[i++] = nicvf_queue_reg_read(nic, reg_offset, q);
+	}
+}
+
+static int nicvf_get_coalesce(struct net_device *netdev,
+			      struct ethtool_coalesce *cmd,
+			      struct kernel_ethtool_coalesce *kernel_coal,
+			      struct netlink_ext_ack *extack)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+
+	cmd->rx_coalesce_usecs = nic->cq_coalesce_usecs;
+	return 0;
+}
+
+static void nicvf_get_ringparam(struct net_device *netdev,
+				struct ethtool_ringparam *ring,
+				struct kernel_ethtool_ringparam *kernel_ring,
+				struct netlink_ext_ack *extack)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+	struct queue_set *qs = nic->qs;
+
+	ring->rx_max_pending = MAX_CMP_QUEUE_LEN;
+	ring->rx_pending = qs->cq_len;
+	ring->tx_max_pending = MAX_SND_QUEUE_LEN;
+	ring->tx_pending = qs->sq_len;
+}
+
+static int nicvf_set_ringparam(struct net_device *netdev,
+			       struct ethtool_ringparam *ring,
+			       struct kernel_ethtool_ringparam *kernel_ring,
+			       struct netlink_ext_ack *extack)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+	struct queue_set *qs = nic->qs;
+	u32 rx_count, tx_count;
+
+	/* Due to HW errata this is not supported on T88 pass 1.x silicon */
+	if (pass1_silicon(nic->pdev))
+		return -EINVAL;
+
+	if ((ring->rx_mini_pending) || (ring->rx_jumbo_pending))
+		return -EINVAL;
+
+	tx_count = clamp_t(u32, ring->tx_pending,
+			   MIN_SND_QUEUE_LEN, MAX_SND_QUEUE_LEN);
+	rx_count = clamp_t(u32, ring->rx_pending,
+			   MIN_CMP_QUEUE_LEN, MAX_CMP_QUEUE_LEN);
+
+	if ((tx_count == qs->sq_len) && (rx_count == qs->cq_len))
+		return 0;
+
+	/* Permitted lengths are 1K, 2K, 4K, 8K, 16K, 32K, 64K */
+	qs->sq_len = rounddown_pow_of_two(tx_count);
+	qs->cq_len = rounddown_pow_of_two(rx_count);
+
+	if (netif_running(netdev)) {
+		nicvf_stop(netdev);
+		nicvf_open(netdev);
+	}
+
+	return 0;
+}
+
+static int nicvf_get_rss_hash_opts(struct nicvf *nic,
+				   struct ethtool_rxnfc *info)
+{
+	info->data = 0;
+
+	switch (info->flow_type) {
+	case TCP_V4_FLOW:
+	case TCP_V6_FLOW:
+	case UDP_V4_FLOW:
+	case UDP_V6_FLOW:
+	case SCTP_V4_FLOW:
+	case SCTP_V6_FLOW:
+		info->data |= RXH_L4_B_0_1 | RXH_L4_B_2_3;
+		fallthrough;
+	case IPV4_FLOW:
+	case IPV6_FLOW:
+		info->data |= RXH_IP_SRC | RXH_IP_DST;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int nicvf_get_rxnfc(struct net_device *dev,
+			   struct ethtool_rxnfc *info, u32 *rules)
+{
+	struct nicvf *nic = netdev_priv(dev);
+	int ret = -EOPNOTSUPP;
+
+	switch (info->cmd) {
+	case ETHTOOL_GRXRINGS:
+		info->data = nic->rx_queues;
+		ret = 0;
+		break;
+	case ETHTOOL_GRXFH:
+		return nicvf_get_rss_hash_opts(nic, info);
+	default:
+		break;
+	}
+	return ret;
+}
+
+static int nicvf_set_rss_hash_opts(struct nicvf *nic,
+				   struct ethtool_rxnfc *info)
+{
+	struct nicvf_rss_info *rss = &nic->rss_info;
+	u64 rss_cfg = nicvf_reg_read(nic, NIC_VNIC_RSS_CFG);
+
+	if (!rss->enable)
+		netdev_err(nic->netdev,
+			   "RSS is disabled, hash cannot be set\n");
+
+	netdev_info(nic->netdev, "Set RSS flow type = %d, data = %lld\n",
+		    info->flow_type, info->data);
+
+	if (!(info->data & RXH_IP_SRC) || !(info->data & RXH_IP_DST))
+		return -EINVAL;
+
+	switch (info->flow_type) {
+	case TCP_V4_FLOW:
+	case TCP_V6_FLOW:
+		switch (info->data & (RXH_L4_B_0_1 | RXH_L4_B_2_3)) {
+		case 0:
+			rss_cfg &= ~(1ULL << RSS_HASH_TCP);
+			break;
+		case (RXH_L4_B_0_1 | RXH_L4_B_2_3):
+			rss_cfg |= (1ULL << RSS_HASH_TCP);
+			break;
+		default:
+			return -EINVAL;
+		}
+		break;
+	case UDP_V4_FLOW:
+	case UDP_V6_FLOW:
+		switch (info->data & (RXH_L4_B_0_1 | RXH_L4_B_2_3)) {
+		case 0:
+			rss_cfg &= ~(1ULL << RSS_HASH_UDP);
+			break;
+		case (RXH_L4_B_0_1 | RXH_L4_B_2_3):
+			rss_cfg |= (1ULL << RSS_HASH_UDP);
+			break;
+		default:
+			return -EINVAL;
+		}
+		break;
+	case SCTP_V4_FLOW:
+	case SCTP_V6_FLOW:
+		switch (info->data & (RXH_L4_B_0_1 | RXH_L4_B_2_3)) {
+		case 0:
+			rss_cfg &= ~(1ULL << RSS_HASH_L4ETC);
+			break;
+		case (RXH_L4_B_0_1 | RXH_L4_B_2_3):
+			rss_cfg |= (1ULL << RSS_HASH_L4ETC);
+			break;
+		default:
+			return -EINVAL;
+		}
+		break;
+	case IPV4_FLOW:
+	case IPV6_FLOW:
+		rss_cfg = RSS_HASH_IP;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	nicvf_reg_write(nic, NIC_VNIC_RSS_CFG, rss_cfg);
+	return 0;
+}
+
+static int nicvf_set_rxnfc(struct net_device *dev, struct ethtool_rxnfc *info)
+{
+	struct nicvf *nic = netdev_priv(dev);
+
+	switch (info->cmd) {
+	case ETHTOOL_SRXFH:
+		return nicvf_set_rss_hash_opts(nic, info);
+	default:
+		break;
+	}
+	return -EOPNOTSUPP;
+}
+
+static u32 nicvf_get_rxfh_key_size(struct net_device *netdev)
+{
+	return RSS_HASH_KEY_SIZE * sizeof(u64);
+}
+
+static u32 nicvf_get_rxfh_indir_size(struct net_device *dev)
+{
+	struct nicvf *nic = netdev_priv(dev);
+
+	return nic->rss_info.rss_size;
+}
+
+static int nicvf_get_rxfh(struct net_device *dev, u32 *indir, u8 *hkey,
+			  u8 *hfunc)
+{
+	struct nicvf *nic = netdev_priv(dev);
+	struct nicvf_rss_info *rss = &nic->rss_info;
+	int idx;
+
+	if (indir) {
+		for (idx = 0; idx < rss->rss_size; idx++)
+			indir[idx] = rss->ind_tbl[idx];
+	}
+
+	if (hkey)
+		memcpy(hkey, rss->key, RSS_HASH_KEY_SIZE * sizeof(u64));
+
+	if (hfunc)
+		*hfunc = ETH_RSS_HASH_TOP;
+
+	return 0;
+}
+
+static int nicvf_set_rxfh(struct net_device *dev, const u32 *indir,
+			  const u8 *hkey, const u8 hfunc)
+{
+	struct nicvf *nic = netdev_priv(dev);
+	struct nicvf_rss_info *rss = &nic->rss_info;
+	int idx;
+
+	if (hfunc != ETH_RSS_HASH_NO_CHANGE && hfunc != ETH_RSS_HASH_TOP)
+		return -EOPNOTSUPP;
+
+	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];
+	}
+
+	if (hkey) {
+		memcpy(rss->key, hkey, RSS_HASH_KEY_SIZE * sizeof(u64));
+		nicvf_set_rss_key(nic);
+	}
+
+	nicvf_config_rss(nic);
+	return 0;
+}
+
+/* Get no of queues device supports and current queue count */
+static void nicvf_get_channels(struct net_device *dev,
+			       struct ethtool_channels *channel)
+{
+	struct nicvf *nic = netdev_priv(dev);
+
+	memset(channel, 0, sizeof(*channel));
+
+	channel->max_rx = nic->max_queues;
+	channel->max_tx = nic->max_queues;
+
+	channel->rx_count = nic->rx_queues;
+	channel->tx_count = nic->tx_queues;
+}
+
+/* Set no of Tx, Rx queues to be used */
+static int nicvf_set_channels(struct net_device *dev,
+			      struct ethtool_channels *channel)
+{
+	struct nicvf *nic = netdev_priv(dev);
+	int err = 0;
+	bool if_up = netif_running(dev);
+	u8 cqcount, txq_count;
+
+	if (!channel->rx_count || !channel->tx_count)
+		return -EINVAL;
+	if (channel->rx_count > nic->max_queues)
+		return -EINVAL;
+	if (channel->tx_count > nic->max_queues)
+		return -EINVAL;
+
+	if (nic->xdp_prog &&
+	    ((channel->tx_count + channel->rx_count) > nic->max_queues)) {
+		netdev_err(nic->netdev,
+			   "XDP mode, RXQs + TXQs > Max %d\n",
+			   nic->max_queues);
+		return -EINVAL;
+	}
+
+	if (if_up)
+		nicvf_stop(dev);
+
+	nic->rx_queues = channel->rx_count;
+	nic->tx_queues = channel->tx_count;
+	if (!nic->xdp_prog)
+		nic->xdp_tx_queues = 0;
+	else
+		nic->xdp_tx_queues = channel->rx_count;
+
+	txq_count = nic->xdp_tx_queues + nic->tx_queues;
+	cqcount = max(nic->rx_queues, txq_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(u8, nic->rx_queues, MAX_RCV_QUEUES_PER_QS);
+	nic->qs->sq_cnt = min_t(u8, txq_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->tx_queues, nic->rx_queues);
+	if (err)
+		return err;
+
+	if (if_up)
+		nicvf_open(dev);
+
+	netdev_info(dev, "Setting num Tx rings to %d, Rx rings to %d success\n",
+		    nic->tx_queues, nic->rx_queues);
+
+	return err;
+}
+
+static void nicvf_get_pauseparam(struct net_device *dev,
+				 struct ethtool_pauseparam *pause)
+{
+	struct nicvf *nic = netdev_priv(dev);
+	union nic_mbx mbx = {};
+
+	/* Supported only for 10G/40G interfaces */
+	if ((nic->mac_type == BGX_MODE_SGMII) ||
+	    (nic->mac_type == BGX_MODE_QSGMII) ||
+	    (nic->mac_type == BGX_MODE_RGMII))
+		return;
+
+	mbx.pfc.msg = NIC_MBOX_MSG_PFC;
+	mbx.pfc.get = 1;
+	if (!nicvf_send_msg_to_pf(nic, &mbx)) {
+		pause->autoneg = nic->pfc.autoneg;
+		pause->rx_pause = nic->pfc.fc_rx;
+		pause->tx_pause = nic->pfc.fc_tx;
+	}
+}
+
+static int nicvf_set_pauseparam(struct net_device *dev,
+				struct ethtool_pauseparam *pause)
+{
+	struct nicvf *nic = netdev_priv(dev);
+	union nic_mbx mbx = {};
+
+	/* Supported only for 10G/40G interfaces */
+	if ((nic->mac_type == BGX_MODE_SGMII) ||
+	    (nic->mac_type == BGX_MODE_QSGMII) ||
+	    (nic->mac_type == BGX_MODE_RGMII))
+		return -EOPNOTSUPP;
+
+	if (pause->autoneg)
+		return -EOPNOTSUPP;
+
+	mbx.pfc.msg = NIC_MBOX_MSG_PFC;
+	mbx.pfc.get = 0;
+	mbx.pfc.fc_rx = pause->rx_pause;
+	mbx.pfc.fc_tx = pause->tx_pause;
+	if (nicvf_send_msg_to_pf(nic, &mbx))
+		return -EAGAIN;
+
+	nic->pfc.fc_rx = pause->rx_pause;
+	nic->pfc.fc_tx = pause->tx_pause;
+
+	return 0;
+}
+
+static int nicvf_get_ts_info(struct net_device *netdev,
+			     struct ethtool_ts_info *info)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+
+	if (!nic->ptp_clock)
+		return ethtool_op_get_ts_info(netdev, info);
+
+	info->so_timestamping = SOF_TIMESTAMPING_TX_SOFTWARE |
+				SOF_TIMESTAMPING_RX_SOFTWARE |
+				SOF_TIMESTAMPING_SOFTWARE |
+				SOF_TIMESTAMPING_TX_HARDWARE |
+				SOF_TIMESTAMPING_RX_HARDWARE |
+				SOF_TIMESTAMPING_RAW_HARDWARE;
+
+	info->phc_index = cavium_ptp_clock_index(nic->ptp_clock);
+
+	info->tx_types = (1 << HWTSTAMP_TX_OFF) | (1 << HWTSTAMP_TX_ON);
+
+	info->rx_filters = (1 << HWTSTAMP_FILTER_NONE) |
+			   (1 << HWTSTAMP_FILTER_ALL);
+
+	return 0;
+}
+
+static const struct ethtool_ops nicvf_ethtool_ops = {
+	.get_link		= nicvf_get_link,
+	.get_drvinfo		= nicvf_get_drvinfo,
+	.get_msglevel		= nicvf_get_msglevel,
+	.set_msglevel		= nicvf_set_msglevel,
+	.get_strings		= nicvf_get_strings,
+	.get_sset_count		= nicvf_get_sset_count,
+	.get_ethtool_stats	= nicvf_get_ethtool_stats,
+	.get_regs_len		= nicvf_get_regs_len,
+	.get_regs		= nicvf_get_regs,
+	.get_coalesce		= nicvf_get_coalesce,
+	.get_ringparam		= nicvf_get_ringparam,
+	.set_ringparam		= nicvf_set_ringparam,
+	.get_rxnfc		= nicvf_get_rxnfc,
+	.set_rxnfc		= nicvf_set_rxnfc,
+	.get_rxfh_key_size	= nicvf_get_rxfh_key_size,
+	.get_rxfh_indir_size	= nicvf_get_rxfh_indir_size,
+	.get_rxfh		= nicvf_get_rxfh,
+	.set_rxfh		= nicvf_set_rxfh,
+	.get_channels		= nicvf_get_channels,
+	.set_channels		= nicvf_set_channels,
+	.get_pauseparam         = nicvf_get_pauseparam,
+	.set_pauseparam         = nicvf_set_pauseparam,
+	.get_ts_info		= nicvf_get_ts_info,
+	.get_link_ksettings	= nicvf_get_link_ksettings,
+};
+
+void nicvf_set_ethtool_ops(struct net_device *netdev)
+{
+	netdev->ethtool_ops = &nicvf_ethtool_ops;
+}
diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_main.c b/drivers/net/ethernet/cavium/thunder/nicvf_main.c
new file mode 100644
index 000000000..f2f95493e
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/nicvf_main.c
@@ -0,0 +1,2323 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2015 Cavium, Inc.
+ */
+
+#include <linux/module.h>
+#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>
+#include <linux/prefetch.h>
+#include <linux/irq.h>
+#include <linux/iommu.h>
+#include <linux/bpf.h>
+#include <linux/bpf_trace.h>
+#include <linux/filter.h>
+#include <linux/net_tstamp.h>
+#include <linux/workqueue.h>
+
+#include "nic_reg.h"
+#include "nic.h"
+#include "nicvf_queues.h"
+#include "thunder_bgx.h"
+#include "../common/cavium_ptp.h"
+
+#define DRV_NAME	"nicvf"
+#define DRV_VERSION	"1.0"
+
+/* NOTE: Packets bigger than 1530 are split across multiple pages and XDP needs
+ * the buffer to be contiguous. Allow XDP to be set up only if we don't exceed
+ * this value, keeping headroom for the 14 byte Ethernet header and two
+ * VLAN tags (for QinQ)
+ */
+#define MAX_XDP_MTU	(1530 - ETH_HLEN - VLAN_HLEN * 2)
+
+/* Supported devices */
+static const struct pci_device_id nicvf_id_table[] = {
+	{ PCI_DEVICE_SUB(PCI_VENDOR_ID_CAVIUM,
+			 PCI_DEVICE_ID_THUNDER_NIC_VF,
+			 PCI_VENDOR_ID_CAVIUM,
+			 PCI_SUBSYS_DEVID_88XX_NIC_VF) },
+	{ PCI_DEVICE_SUB(PCI_VENDOR_ID_CAVIUM,
+			 PCI_DEVICE_ID_THUNDER_PASS1_NIC_VF,
+			 PCI_VENDOR_ID_CAVIUM,
+			 PCI_SUBSYS_DEVID_88XX_PASS1_NIC_VF) },
+	{ PCI_DEVICE_SUB(PCI_VENDOR_ID_CAVIUM,
+			 PCI_DEVICE_ID_THUNDER_NIC_VF,
+			 PCI_VENDOR_ID_CAVIUM,
+			 PCI_SUBSYS_DEVID_81XX_NIC_VF) },
+	{ PCI_DEVICE_SUB(PCI_VENDOR_ID_CAVIUM,
+			 PCI_DEVICE_ID_THUNDER_NIC_VF,
+			 PCI_VENDOR_ID_CAVIUM,
+			 PCI_SUBSYS_DEVID_83XX_NIC_VF) },
+	{ 0, }  /* end of table */
+};
+
+MODULE_AUTHOR("Sunil Goutham");
+MODULE_DESCRIPTION("Cavium Thunder NIC Virtual Function Driver");
+MODULE_LICENSE("GPL v2");
+MODULE_VERSION(DRV_VERSION);
+MODULE_DEVICE_TABLE(pci, nicvf_id_table);
+
+static int debug = 0x00;
+module_param(debug, int, 0644);
+MODULE_PARM_DESC(debug, "Debug message level bitmap");
+
+static int cpi_alg = CPI_ALG_NONE;
+module_param(cpi_alg, int, 0444);
+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;
+}
+
+/* The Cavium ThunderX network controller can *only* be found in SoCs
+ * containing the ThunderX ARM64 CPU implementation.  All accesses to the device
+ * registers on this platform are implicitly strongly ordered with respect
+ * to memory accesses. So writeq_relaxed() and readq_relaxed() are safe to use
+ * with no memory barriers in this driver.  The readq()/writeq() functions add
+ * explicit ordering operation which in this case are redundant, and only
+ * add overhead.
+ */
+
+/* Register read/write APIs */
+void nicvf_reg_write(struct nicvf *nic, u64 offset, u64 val)
+{
+	writeq_relaxed(val, nic->reg_base + offset);
+}
+
+u64 nicvf_reg_read(struct nicvf *nic, u64 offset)
+{
+	return readq_relaxed(nic->reg_base + offset);
+}
+
+void nicvf_queue_reg_write(struct nicvf *nic, u64 offset,
+			   u64 qidx, u64 val)
+{
+	void __iomem *addr = nic->reg_base + offset;
+
+	writeq_relaxed(val, addr + (qidx << NIC_Q_NUM_SHIFT));
+}
+
+u64 nicvf_queue_reg_read(struct nicvf *nic, u64 offset, u64 qidx)
+{
+	void __iomem *addr = nic->reg_base + offset;
+
+	return readq_relaxed(addr + (qidx << NIC_Q_NUM_SHIFT));
+}
+
+/* VF -> PF mailbox communication */
+static void nicvf_write_to_mbx(struct nicvf *nic, union nic_mbx *mbx)
+{
+	u64 *msg = (u64 *)mbx;
+
+	nicvf_reg_write(nic, NIC_VF_PF_MAILBOX_0_1 + 0, msg[0]);
+	nicvf_reg_write(nic, NIC_VF_PF_MAILBOX_0_1 + 8, msg[1]);
+}
+
+int nicvf_send_msg_to_pf(struct nicvf *nic, union nic_mbx *mbx)
+{
+	unsigned long timeout;
+	int ret = 0;
+
+	mutex_lock(&nic->rx_mode_mtx);
+
+	nic->pf_acked = false;
+	nic->pf_nacked = false;
+
+	nicvf_write_to_mbx(nic, mbx);
+
+	timeout = jiffies + msecs_to_jiffies(NIC_MBOX_MSG_TIMEOUT);
+	/* Wait for previous message to be acked, timeout 2sec */
+	while (!nic->pf_acked) {
+		if (nic->pf_nacked) {
+			netdev_err(nic->netdev,
+				   "PF NACK to mbox msg 0x%02x from VF%d\n",
+				   (mbx->msg.msg & 0xFF), nic->vf_id);
+			ret = -EINVAL;
+			break;
+		}
+		usleep_range(8000, 10000);
+		if (nic->pf_acked)
+			break;
+		if (time_after(jiffies, timeout)) {
+			netdev_err(nic->netdev,
+				   "PF didn't ACK to mbox msg 0x%02x from VF%d\n",
+				   (mbx->msg.msg & 0xFF), nic->vf_id);
+			ret = -EBUSY;
+			break;
+		}
+	}
+	mutex_unlock(&nic->rx_mode_mtx);
+	return ret;
+}
+
+/* Checks if VF is able to comminicate with PF
+* and also gets the VNIC number this VF is associated to.
+*/
+static int nicvf_check_pf_ready(struct nicvf *nic)
+{
+	union nic_mbx mbx = {};
+
+	mbx.msg.msg = NIC_MBOX_MSG_READY;
+	if (nicvf_send_msg_to_pf(nic, &mbx)) {
+		netdev_err(nic->netdev,
+			   "PF didn't respond to READY msg\n");
+		return 0;
+	}
+
+	return 1;
+}
+
+static void nicvf_send_cfg_done(struct nicvf *nic)
+{
+	union nic_mbx mbx = {};
+
+	mbx.msg.msg = NIC_MBOX_MSG_CFG_DONE;
+	if (nicvf_send_msg_to_pf(nic, &mbx)) {
+		netdev_err(nic->netdev,
+			   "PF didn't respond to CFG DONE msg\n");
+	}
+}
+
+static void nicvf_read_bgx_stats(struct nicvf *nic, struct bgx_stats_msg *bgx)
+{
+	if (bgx->rx)
+		nic->bgx_stats.rx_stats[bgx->idx] = bgx->stats;
+	else
+		nic->bgx_stats.tx_stats[bgx->idx] = bgx->stats;
+}
+
+static void  nicvf_handle_mbx_intr(struct nicvf *nic)
+{
+	union nic_mbx mbx = {};
+	u64 *mbx_data;
+	u64 mbx_addr;
+	int i;
+
+	mbx_addr = NIC_VF_PF_MAILBOX_0_1;
+	mbx_data = (u64 *)&mbx;
+
+	for (i = 0; i < NIC_PF_VF_MAILBOX_SIZE; i++) {
+		*mbx_data = nicvf_reg_read(nic, mbx_addr);
+		mbx_data++;
+		mbx_addr += sizeof(u64);
+	}
+
+	netdev_dbg(nic->netdev, "Mbox message: msg: 0x%x\n", mbx.msg.msg);
+	switch (mbx.msg.msg) {
+	case NIC_MBOX_MSG_READY:
+		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)
+			eth_hw_addr_set(nic->netdev, 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;
+		break;
+	case NIC_MBOX_MSG_ACK:
+		nic->pf_acked = true;
+		break;
+	case NIC_MBOX_MSG_NACK:
+		nic->pf_nacked = true;
+		break;
+	case NIC_MBOX_MSG_RSS_SIZE:
+		nic->rss_info.rss_size = mbx.rss_size.ind_tbl_size;
+		nic->pf_acked = true;
+		break;
+	case NIC_MBOX_MSG_BGX_STATS:
+		nicvf_read_bgx_stats(nic, &mbx.bgx_stats);
+		nic->pf_acked = true;
+		break;
+	case NIC_MBOX_MSG_BGX_LINK_CHANGE:
+		nic->pf_acked = true;
+		if (nic->link_up != mbx.link_status.link_up) {
+			nic->link_up = mbx.link_status.link_up;
+			nic->duplex = mbx.link_status.duplex;
+			nic->speed = mbx.link_status.speed;
+			nic->mac_type = mbx.link_status.mac_type;
+			if (nic->link_up) {
+				netdev_info(nic->netdev,
+					    "Link is Up %d Mbps %s duplex\n",
+					    nic->speed,
+					    nic->duplex == DUPLEX_FULL ?
+					    "Full" : "Half");
+				netif_carrier_on(nic->netdev);
+				netif_tx_start_all_queues(nic->netdev);
+			} else {
+				netdev_info(nic->netdev, "Link is Down\n");
+				netif_carrier_off(nic->netdev);
+				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;
+	case NIC_MBOX_MSG_PFC:
+		nic->pfc.autoneg = mbx.pfc.autoneg;
+		nic->pfc.fc_rx = mbx.pfc.fc_rx;
+		nic->pfc.fc_tx = mbx.pfc.fc_tx;
+		nic->pf_acked = true;
+		break;
+	default:
+		netdev_err(nic->netdev,
+			   "Invalid message from PF, msg 0x%x\n", mbx.msg.msg);
+		break;
+	}
+	nicvf_clear_intr(nic, NICVF_INTR_MBOX, 0);
+}
+
+static int nicvf_hw_set_mac_addr(struct nicvf *nic, struct net_device *netdev)
+{
+	union nic_mbx mbx = {};
+
+	mbx.mac.msg = NIC_MBOX_MSG_SET_MAC;
+	mbx.mac.vf_id = nic->vf_id;
+	ether_addr_copy(mbx.mac.mac_addr, netdev->dev_addr);
+
+	return nicvf_send_msg_to_pf(nic, &mbx);
+}
+
+static void nicvf_config_cpi(struct nicvf *nic)
+{
+	union nic_mbx mbx = {};
+
+	mbx.cpi_cfg.msg = NIC_MBOX_MSG_CPI_CFG;
+	mbx.cpi_cfg.vf_id = nic->vf_id;
+	mbx.cpi_cfg.cpi_alg = nic->cpi_alg;
+	mbx.cpi_cfg.rq_cnt = nic->qs->rq_cnt;
+
+	nicvf_send_msg_to_pf(nic, &mbx);
+}
+
+static void nicvf_get_rss_size(struct nicvf *nic)
+{
+	union nic_mbx mbx = {};
+
+	mbx.rss_size.msg = NIC_MBOX_MSG_RSS_SIZE;
+	mbx.rss_size.vf_id = nic->vf_id;
+	nicvf_send_msg_to_pf(nic, &mbx);
+}
+
+void nicvf_config_rss(struct nicvf *nic)
+{
+	union nic_mbx mbx = {};
+	struct nicvf_rss_info *rss = &nic->rss_info;
+	int ind_tbl_len = rss->rss_size;
+	int i, nextq = 0;
+
+	mbx.rss_cfg.vf_id = nic->vf_id;
+	mbx.rss_cfg.hash_bits = rss->hash_bits;
+	while (ind_tbl_len) {
+		mbx.rss_cfg.tbl_offset = nextq;
+		mbx.rss_cfg.tbl_len = min(ind_tbl_len,
+					       RSS_IND_TBL_LEN_PER_MBX_MSG);
+		mbx.rss_cfg.msg = mbx.rss_cfg.tbl_offset ?
+			  NIC_MBOX_MSG_RSS_CFG_CONT : NIC_MBOX_MSG_RSS_CFG;
+
+		for (i = 0; i < mbx.rss_cfg.tbl_len; i++)
+			mbx.rss_cfg.ind_tbl[i] = rss->ind_tbl[nextq++];
+
+		nicvf_send_msg_to_pf(nic, &mbx);
+
+		ind_tbl_len -= mbx.rss_cfg.tbl_len;
+	}
+}
+
+void nicvf_set_rss_key(struct nicvf *nic)
+{
+	struct nicvf_rss_info *rss = &nic->rss_info;
+	u64 key_addr = NIC_VNIC_RSS_KEY_0_4;
+	int idx;
+
+	for (idx = 0; idx < RSS_HASH_KEY_SIZE; idx++) {
+		nicvf_reg_write(nic, key_addr, rss->key[idx]);
+		key_addr += sizeof(u64);
+	}
+}
+
+static int nicvf_rss_init(struct nicvf *nic)
+{
+	struct nicvf_rss_info *rss = &nic->rss_info;
+	int idx;
+
+	nicvf_get_rss_size(nic);
+
+	if (cpi_alg != CPI_ALG_NONE) {
+		rss->enable = false;
+		rss->hash_bits = 0;
+		return 0;
+	}
+
+	rss->enable = true;
+
+	netdev_rss_key_fill(rss->key, RSS_HASH_KEY_SIZE * sizeof(u64));
+	nicvf_set_rss_key(nic);
+
+	rss->cfg = RSS_IP_HASH_ENA | RSS_TCP_HASH_ENA | RSS_UDP_HASH_ENA;
+	nicvf_reg_write(nic, NIC_VNIC_RSS_CFG, rss->cfg);
+
+	rss->hash_bits =  ilog2(rounddown_pow_of_two(rss->rss_size));
+
+	for (idx = 0; idx < rss->rss_size; idx++)
+		rss->ind_tbl[idx] = ethtool_rxfh_indir_default(idx,
+							       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;
+
+	tx_queues = nic->tx_queues + nic->xdp_tx_queues;
+	if (tx_queues > MAX_SND_QUEUES_PER_QS)
+		tx_queues = 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)
+{
+	int err = 0;
+
+	err = netif_set_real_num_tx_queues(netdev, tx_queues);
+	if (err) {
+		netdev_err(netdev,
+			   "Failed to set no of Tx queues: %d\n", tx_queues);
+		return err;
+	}
+
+	err = netif_set_real_num_rx_queues(netdev, rx_queues);
+	if (err)
+		netdev_err(netdev,
+			   "Failed to set no of Rx queues: %d\n", rx_queues);
+	return err;
+}
+
+static int nicvf_init_resources(struct nicvf *nic)
+{
+	int err;
+
+	/* Enable Qset */
+	nicvf_qset_config(nic, true);
+
+	/* Initialize queues and HW for data transfer */
+	err = nicvf_config_data_transfer(nic, true);
+	if (err) {
+		netdev_err(nic->netdev,
+			   "Failed to alloc/config VF's QSet resources\n");
+		return err;
+	}
+
+	return 0;
+}
+
+static inline bool nicvf_xdp_rx(struct nicvf *nic, struct bpf_prog *prog,
+				struct cqe_rx_t *cqe_rx, struct snd_queue *sq,
+				struct rcv_queue *rq, struct sk_buff **skb)
+{
+	unsigned char *hard_start, *data;
+	struct xdp_buff xdp;
+	struct page *page;
+	u32 action;
+	u16 len, offset = 0;
+	u64 dma_addr, cpu_addr;
+	void *orig_data;
+
+	/* Retrieve packet buffer's DMA address and length */
+	len = *((u16 *)((void *)cqe_rx + (3 * sizeof(u64))));
+	dma_addr = *((u64 *)((void *)cqe_rx + (7 * sizeof(u64))));
+
+	cpu_addr = nicvf_iova_to_phys(nic, dma_addr);
+	if (!cpu_addr)
+		return false;
+	cpu_addr = (u64)phys_to_virt(cpu_addr);
+	page = virt_to_page((void *)cpu_addr);
+
+	xdp_init_buff(&xdp, RCV_FRAG_LEN + XDP_PACKET_HEADROOM,
+		      &rq->xdp_rxq);
+	hard_start = page_address(page);
+	data = (unsigned char *)cpu_addr;
+	xdp_prepare_buff(&xdp, hard_start, data - hard_start, len, false);
+	orig_data = xdp.data;
+
+	action = bpf_prog_run_xdp(prog, &xdp);
+
+	len = xdp.data_end - xdp.data;
+	/* Check if XDP program has changed headers */
+	if (orig_data != xdp.data) {
+		offset = orig_data - xdp.data;
+		dma_addr -= offset;
+	}
+
+	switch (action) {
+	case XDP_PASS:
+		/* Check if it's a recycled page, if not
+		 * unmap the DMA mapping.
+		 *
+		 * Recycled page holds an extra reference.
+		 */
+		if (page_ref_count(page) == 1) {
+			dma_addr &= PAGE_MASK;
+			dma_unmap_page_attrs(&nic->pdev->dev, dma_addr,
+					     RCV_FRAG_LEN + XDP_PACKET_HEADROOM,
+					     DMA_FROM_DEVICE,
+					     DMA_ATTR_SKIP_CPU_SYNC);
+		}
+
+		/* Build SKB and pass on packet to network stack */
+		*skb = build_skb(xdp.data,
+				 RCV_FRAG_LEN - cqe_rx->align_pad + offset);
+		if (!*skb)
+			put_page(page);
+		else
+			skb_put(*skb, len);
+		return false;
+	case XDP_TX:
+		nicvf_xdp_sq_append_pkt(nic, sq, (u64)xdp.data, dma_addr, len);
+		return true;
+	default:
+		bpf_warn_invalid_xdp_action(nic->netdev, prog, action);
+		fallthrough;
+	case XDP_ABORTED:
+		trace_xdp_exception(nic->netdev, prog, action);
+		fallthrough;
+	case XDP_DROP:
+		/* Check if it's a recycled page, if not
+		 * unmap the DMA mapping.
+		 *
+		 * Recycled page holds an extra reference.
+		 */
+		if (page_ref_count(page) == 1) {
+			dma_addr &= PAGE_MASK;
+			dma_unmap_page_attrs(&nic->pdev->dev, dma_addr,
+					     RCV_FRAG_LEN + XDP_PACKET_HEADROOM,
+					     DMA_FROM_DEVICE,
+					     DMA_ATTR_SKIP_CPU_SYNC);
+		}
+		put_page(page);
+		return true;
+	}
+	return false;
+}
+
+static void nicvf_snd_ptp_handler(struct net_device *netdev,
+				  struct cqe_send_t *cqe_tx)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+	struct skb_shared_hwtstamps ts;
+	u64 ns;
+
+	nic = nic->pnicvf;
+
+	/* Sync for 'ptp_skb' */
+	smp_rmb();
+
+	/* New timestamp request can be queued now */
+	atomic_set(&nic->tx_ptp_skbs, 0);
+
+	/* Check for timestamp requested skb */
+	if (!nic->ptp_skb)
+		return;
+
+	/* Check if timestamping is timedout, which is set to 10us */
+	if (cqe_tx->send_status == CQ_TX_ERROP_TSTMP_TIMEOUT ||
+	    cqe_tx->send_status == CQ_TX_ERROP_TSTMP_CONFLICT)
+		goto no_tstamp;
+
+	/* Get the timestamp */
+	memset(&ts, 0, sizeof(ts));
+	ns = cavium_ptp_tstamp2time(nic->ptp_clock, cqe_tx->ptp_timestamp);
+	ts.hwtstamp = ns_to_ktime(ns);
+	skb_tstamp_tx(nic->ptp_skb, &ts);
+
+no_tstamp:
+	/* Free the original skb */
+	dev_kfree_skb_any(nic->ptp_skb);
+	nic->ptp_skb = NULL;
+	/* Sync 'ptp_skb' */
+	smp_wmb();
+}
+
+static void nicvf_snd_pkt_handler(struct net_device *netdev,
+				  struct cqe_send_t *cqe_tx,
+				  int budget, int *subdesc_cnt,
+				  unsigned int *tx_pkts, unsigned int *tx_bytes)
+{
+	struct sk_buff *skb = NULL;
+	struct page *page;
+	struct nicvf *nic = netdev_priv(netdev);
+	struct snd_queue *sq;
+	struct sq_hdr_subdesc *hdr;
+	struct sq_hdr_subdesc *tso_sqe;
+
+	sq = &nic->qs->sq[cqe_tx->sq_idx];
+
+	hdr = (struct sq_hdr_subdesc *)GET_SQ_DESC(sq, cqe_tx->sqe_ptr);
+	if (hdr->subdesc_type != SQ_DESC_TYPE_HEADER)
+		return;
+
+	/* Check for errors */
+	if (cqe_tx->send_status)
+		nicvf_check_cqe_tx_errs(nic->pnicvf, cqe_tx);
+
+	/* Is this a XDP designated Tx queue */
+	if (sq->is_xdp) {
+		page = (struct page *)sq->xdp_page[cqe_tx->sqe_ptr];
+		/* Check if it's recycled page or else unmap DMA mapping */
+		if (page && (page_ref_count(page) == 1))
+			nicvf_unmap_sndq_buffers(nic, sq, cqe_tx->sqe_ptr,
+						 hdr->subdesc_cnt);
+
+		/* Release page reference for recycling */
+		if (page)
+			put_page(page);
+		sq->xdp_page[cqe_tx->sqe_ptr] = (u64)NULL;
+		*subdesc_cnt += hdr->subdesc_cnt + 1;
+		return;
+	}
+
+	skb = (struct sk_buff *)sq->skbuff[cqe_tx->sqe_ptr];
+	if (skb) {
+		/* Check for dummy descriptor used for HW TSO offload on 88xx */
+		if (hdr->dont_send) {
+			/* Get actual TSO descriptors and free them */
+			tso_sqe =
+			 (struct sq_hdr_subdesc *)GET_SQ_DESC(sq, hdr->rsvd2);
+			nicvf_unmap_sndq_buffers(nic, sq, hdr->rsvd2,
+						 tso_sqe->subdesc_cnt);
+			*subdesc_cnt += tso_sqe->subdesc_cnt + 1;
+		} else {
+			nicvf_unmap_sndq_buffers(nic, sq, cqe_tx->sqe_ptr,
+						 hdr->subdesc_cnt);
+		}
+		*subdesc_cnt += hdr->subdesc_cnt + 1;
+		prefetch(skb);
+		(*tx_pkts)++;
+		*tx_bytes += skb->len;
+		/* If timestamp is requested for this skb, don't free it */
+		if (skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS &&
+		    !nic->pnicvf->ptp_skb)
+			nic->pnicvf->ptp_skb = skb;
+		else
+			napi_consume_skb(skb, budget);
+		sq->skbuff[cqe_tx->sqe_ptr] = (u64)NULL;
+	} else {
+		/* In case of SW TSO on 88xx, only last segment will have
+		 * a SKB attached, so just free SQEs here.
+		 */
+		if (!nic->hw_tso)
+			*subdesc_cnt += hdr->subdesc_cnt + 1;
+	}
+}
+
+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 inline void nicvf_set_rxtstamp(struct nicvf *nic, struct sk_buff *skb)
+{
+	u64 ns;
+
+	if (!nic->ptp_clock || !nic->hw_rx_tstamp)
+		return;
+
+	/* The first 8 bytes is the timestamp */
+	ns = cavium_ptp_tstamp2time(nic->ptp_clock,
+				    be64_to_cpu(*(__be64 *)skb->data));
+	skb_hwtstamps(skb)->hwtstamp = ns_to_ktime(ns);
+
+	__skb_pull(skb, 8);
+}
+
+static void nicvf_rcv_pkt_handler(struct net_device *netdev,
+				  struct napi_struct *napi,
+				  struct cqe_rx_t *cqe_rx,
+				  struct snd_queue *sq, struct rcv_queue *rq)
+{
+	struct sk_buff *skb = NULL;
+	struct nicvf *nic = netdev_priv(netdev);
+	struct nicvf *snic = nic;
+	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 */
+	if (cqe_rx->err_level || cqe_rx->err_opcode) {
+		err = nicvf_check_cqe_rx_errs(nic, cqe_rx);
+		if (err && !cqe_rx->rb_cnt)
+			return;
+	}
+
+	/* For XDP, ignore pkts spanning multiple pages */
+	if (nic->xdp_prog && (cqe_rx->rb_cnt == 1)) {
+		/* Packet consumed by XDP */
+		if (nicvf_xdp_rx(snic, nic->xdp_prog, cqe_rx, sq, rq, &skb))
+			return;
+	} else {
+		skb = nicvf_get_rcv_skb(snic, cqe_rx,
+					nic->xdp_prog ? true : false);
+	}
+
+	if (!skb)
+		return;
+
+	if (netif_msg_pktdata(nic)) {
+		netdev_info(nic->netdev, "skb 0x%p, len=%d\n", skb, skb->len);
+		print_hex_dump(KERN_INFO, "", DUMP_PREFIX_OFFSET, 16, 1,
+			       skb->data, skb->len, true);
+	}
+
+	/* If error packet, drop it here */
+	if (err) {
+		dev_kfree_skb_any(skb);
+		return;
+	}
+
+	nicvf_set_rxtstamp(nic, skb);
+	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;
+	} else {
+		skb_checksum_none_assert(skb);
+	}
+
+	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
+		netif_receive_skb(skb);
+}
+
+static int nicvf_cq_intr_handler(struct net_device *netdev, u8 cq_idx,
+				 struct napi_struct *napi, int budget)
+{
+	int processed_cqe, work_done = 0, tx_done = 0;
+	int cqe_count, cqe_head;
+	int subdesc_cnt = 0;
+	struct nicvf *nic = netdev_priv(netdev);
+	struct queue_set *qs = nic->qs;
+	struct cmp_queue *cq = &qs->cq[cq_idx];
+	struct cqe_rx_t *cq_desc;
+	struct netdev_queue *txq;
+	struct snd_queue *sq = &qs->sq[cq_idx];
+	struct rcv_queue *rq = &qs->rq[cq_idx];
+	unsigned int tx_pkts = 0, tx_bytes = 0, txq_idx;
+
+	spin_lock_bh(&cq->lock);
+loop:
+	processed_cqe = 0;
+	/* Get no of valid CQ entries to process */
+	cqe_count = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_STATUS, cq_idx);
+	cqe_count &= CQ_CQE_COUNT;
+	if (!cqe_count)
+		goto done;
+
+	/* Get head of the valid CQ entries */
+	cqe_head = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_HEAD, cq_idx) >> 9;
+	cqe_head &= 0xFFFF;
+
+	while (processed_cqe < cqe_count) {
+		/* Get the CQ descriptor */
+		cq_desc = (struct cqe_rx_t *)GET_CQ_DESC(cq, cqe_head);
+		cqe_head++;
+		cqe_head &= (cq->dmem.q_len - 1);
+		/* Initiate prefetch for next descriptor */
+		prefetch((struct cqe_rx_t *)GET_CQ_DESC(cq, cqe_head));
+
+		if ((work_done >= budget) && napi &&
+		    (cq_desc->cqe_type != CQE_TYPE_SEND)) {
+			break;
+		}
+
+		switch (cq_desc->cqe_type) {
+		case CQE_TYPE_RX:
+			nicvf_rcv_pkt_handler(netdev, napi, cq_desc, sq, rq);
+			work_done++;
+		break;
+		case CQE_TYPE_SEND:
+			nicvf_snd_pkt_handler(netdev, (void *)cq_desc,
+					      budget, &subdesc_cnt,
+					      &tx_pkts, &tx_bytes);
+			tx_done++;
+		break;
+		case CQE_TYPE_SEND_PTP:
+			nicvf_snd_ptp_handler(netdev, (void *)cq_desc);
+		break;
+		case CQE_TYPE_INVALID:
+		case CQE_TYPE_RX_SPLIT:
+		case CQE_TYPE_RX_TCP:
+			/* Ignore for now */
+		break;
+		}
+		processed_cqe++;
+	}
+
+	/* Ring doorbell to inform H/W to reuse processed CQEs */
+	nicvf_queue_reg_write(nic, NIC_QSET_CQ_0_7_DOOR,
+			      cq_idx, processed_cqe);
+
+	if ((work_done < budget) && napi)
+		goto loop;
+
+done:
+	/* Update SQ's descriptor free count */
+	if (subdesc_cnt)
+		nicvf_put_sq_desc(sq, subdesc_cnt);
+
+	txq_idx = nicvf_netdev_qidx(nic, cq_idx);
+	/* Handle XDP TX queues */
+	if (nic->pnicvf->xdp_prog) {
+		if (txq_idx < nic->pnicvf->xdp_tx_queues) {
+			nicvf_xdp_sq_doorbell(nic, sq, cq_idx);
+			goto out;
+		}
+		nic = nic->pnicvf;
+		txq_idx -= nic->pnicvf->xdp_tx_queues;
+	}
+
+	/* Wakeup TXQ if its stopped earlier due to SQ full */
+	if (tx_done ||
+	    (atomic_read(&sq->free_cnt) >= MIN_SQ_DESC_PER_PKT_XMIT)) {
+		netdev = nic->pnicvf->netdev;
+		txq = netdev_get_tx_queue(netdev, txq_idx);
+		if (tx_pkts)
+			netdev_tx_completed_queue(txq, tx_pkts, tx_bytes);
+
+		/* To read updated queue and carrier status */
+		smp_mb();
+		if (netif_tx_queue_stopped(txq) && netif_carrier_ok(netdev)) {
+			netif_tx_wake_queue(txq);
+			nic = nic->pnicvf;
+			this_cpu_inc(nic->drv_stats->txq_wake);
+			netif_warn(nic, tx_err, netdev,
+				   "Transmit queue wakeup SQ%d\n", txq_idx);
+		}
+	}
+
+out:
+	spin_unlock_bh(&cq->lock);
+	return work_done;
+}
+
+static int nicvf_poll(struct napi_struct *napi, int budget)
+{
+	u64  cq_head;
+	int  work_done = 0;
+	struct net_device *netdev = napi->dev;
+	struct nicvf *nic = netdev_priv(netdev);
+	struct nicvf_cq_poll *cq;
+
+	cq = container_of(napi, struct nicvf_cq_poll, napi);
+	work_done = nicvf_cq_intr_handler(netdev, cq->cq_idx, napi, budget);
+
+	if (work_done < budget) {
+		/* Slow packet rate, exit polling */
+		napi_complete_done(napi, work_done);
+		/* Re-enable interrupts */
+		cq_head = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_HEAD,
+					       cq->cq_idx);
+		nicvf_clear_intr(nic, NICVF_INTR_CQ, cq->cq_idx);
+		nicvf_queue_reg_write(nic, NIC_QSET_CQ_0_7_HEAD,
+				      cq->cq_idx, cq_head);
+		nicvf_enable_intr(nic, NICVF_INTR_CQ, cq->cq_idx);
+	}
+	return work_done;
+}
+
+/* Qset error interrupt handler
+ *
+ * As of now only CQ errors are handled
+ */
+static void nicvf_handle_qs_err(struct tasklet_struct *t)
+{
+	struct nicvf *nic = from_tasklet(nic, t, qs_err_task);
+	struct queue_set *qs = nic->qs;
+	int qidx;
+	u64 status;
+
+	netif_tx_disable(nic->netdev);
+
+	/* Check if it is CQ err */
+	for (qidx = 0; qidx < qs->cq_cnt; qidx++) {
+		status = nicvf_queue_reg_read(nic, NIC_QSET_CQ_0_7_STATUS,
+					      qidx);
+		if (!(status & CQ_ERR_MASK))
+			continue;
+		/* Process already queued CQEs and reconfig CQ */
+		nicvf_disable_intr(nic, NICVF_INTR_CQ, qidx);
+		nicvf_sq_disable(nic, qidx);
+		nicvf_cq_intr_handler(nic->netdev, qidx, NULL, 0);
+		nicvf_cmp_queue_config(nic, qs, qidx, true);
+		nicvf_sq_free_used_descs(nic->netdev, &qs->sq[qidx], qidx);
+		nicvf_sq_enable(nic, &qs->sq[qidx], qidx);
+
+		nicvf_enable_intr(nic, NICVF_INTR_CQ, qidx);
+	}
+
+	netif_tx_start_all_queues(nic->netdev);
+	/* Re-enable Qset error interrupt */
+	nicvf_enable_intr(nic, NICVF_INTR_QS_ERR, 0);
+}
+
+static void nicvf_dump_intr_status(struct nicvf *nic)
+{
+	netif_info(nic, intr, nic->netdev, "interrupt status 0x%llx\n",
+		   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))
+		return IRQ_HANDLED;
+
+	nicvf_handle_mbx_intr(nic);
+
+	return IRQ_HANDLED;
+}
+
+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_irqoff(&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)
+{
+	struct nicvf *nic = (struct nicvf *)nicvf_irq;
+	u8 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);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static irqreturn_t nicvf_qs_err_intr_handler(int irq, void *nicvf_irq)
+{
+	struct nicvf *nic = (struct nicvf *)nicvf_irq;
+
+	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);
+
+	return IRQ_HANDLED;
+}
+
+static void nicvf_set_irq_affinity(struct nicvf *nic)
+{
+	int vec, cpu;
+
+	for (vec = 0; vec < nic->num_vec; vec++) {
+		if (!nic->irq_allocated[vec])
+			continue;
+
+		if (!zalloc_cpumask_var(&nic->affinity_mask[vec], GFP_KERNEL))
+			return;
+		 /* CQ interrupts */
+		if (vec < NICVF_INTR_ID_SQ)
+			/* Leave CPU0 for RBDR and other interrupts */
+			cpu = nicvf_netdev_qidx(nic, vec) + 1;
+		else
+			cpu = 0;
+
+		cpumask_set_cpu(cpumask_local_spread(cpu, nic->node),
+				nic->affinity_mask[vec]);
+		irq_set_affinity_hint(pci_irq_vector(nic->pdev, vec),
+				      nic->affinity_mask[vec]);
+	}
+}
+
+static int nicvf_register_interrupts(struct nicvf *nic)
+{
+	int irq, ret = 0;
+
+	for_each_cq_irq(irq)
+		sprintf(nic->irq_name[irq], "%s-rxtx-%d",
+			nic->pnicvf->netdev->name,
+			nicvf_netdev_qidx(nic, irq));
+
+	for_each_sq_irq(irq)
+		sprintf(nic->irq_name[irq], "%s-sq-%d",
+			nic->pnicvf->netdev->name,
+			nicvf_netdev_qidx(nic, irq - NICVF_INTR_ID_SQ));
+
+	for_each_rbdr_irq(irq)
+		sprintf(nic->irq_name[irq], "%s-rbdr-%d",
+			nic->pnicvf->netdev->name,
+			nic->sqs_mode ? (nic->sqs_id + 1) : 0);
+
+	/* Register CQ interrupts */
+	for (irq = 0; irq < nic->qs->cq_cnt; irq++) {
+		ret = request_irq(pci_irq_vector(nic->pdev, irq),
+				  nicvf_intr_handler,
+				  0, nic->irq_name[irq], nic->napi[irq]);
+		if (ret)
+			goto err;
+		nic->irq_allocated[irq] = true;
+	}
+
+	/* Register RBDR interrupt */
+	for (irq = NICVF_INTR_ID_RBDR;
+	     irq < (NICVF_INTR_ID_RBDR + nic->qs->rbdr_cnt); irq++) {
+		ret = request_irq(pci_irq_vector(nic->pdev, irq),
+				  nicvf_rbdr_intr_handler,
+				  0, nic->irq_name[irq], nic);
+		if (ret)
+			goto err;
+		nic->irq_allocated[irq] = true;
+	}
+
+	/* Register QS error interrupt */
+	sprintf(nic->irq_name[NICVF_INTR_ID_QS_ERR], "%s-qset-err-%d",
+		nic->pnicvf->netdev->name,
+		nic->sqs_mode ? (nic->sqs_id + 1) : 0);
+	irq = NICVF_INTR_ID_QS_ERR;
+	ret = request_irq(pci_irq_vector(nic->pdev, irq),
+			  nicvf_qs_err_intr_handler,
+			  0, nic->irq_name[irq], nic);
+	if (ret)
+		goto err;
+
+	nic->irq_allocated[irq] = true;
+
+	/* Set IRQ affinities */
+	nicvf_set_irq_affinity(nic);
+
+err:
+	if (ret)
+		netdev_err(nic->netdev, "request_irq failed, vector %d\n", irq);
+
+	return ret;
+}
+
+static void nicvf_unregister_interrupts(struct nicvf *nic)
+{
+	struct pci_dev *pdev = nic->pdev;
+	int irq;
+
+	/* Free registered interrupts */
+	for (irq = 0; irq < nic->num_vec; irq++) {
+		if (!nic->irq_allocated[irq])
+			continue;
+
+		irq_set_affinity_hint(pci_irq_vector(pdev, irq), NULL);
+		free_cpumask_var(nic->affinity_mask[irq]);
+
+		if (irq < NICVF_INTR_ID_SQ)
+			free_irq(pci_irq_vector(pdev, irq), nic->napi[irq]);
+		else
+			free_irq(pci_irq_vector(pdev, irq), nic);
+
+		nic->irq_allocated[irq] = false;
+	}
+
+	/* Disable MSI-X */
+	pci_free_irq_vectors(pdev);
+	nic->num_vec = 0;
+}
+
+/* Initialize MSIX vectors and register MISC interrupt.
+ * Send READY message to PF to check if its alive
+ */
+static int nicvf_register_misc_interrupt(struct nicvf *nic)
+{
+	int ret = 0;
+	int irq = NICVF_INTR_ID_MISC;
+
+	/* Return if mailbox interrupt is already registered */
+	if (nic->pdev->msix_enabled)
+		return 0;
+
+	/* Enable MSI-X */
+	nic->num_vec = pci_msix_vec_count(nic->pdev);
+	ret = pci_alloc_irq_vectors(nic->pdev, nic->num_vec, nic->num_vec,
+				    PCI_IRQ_MSIX);
+	if (ret < 0) {
+		netdev_err(nic->netdev,
+			   "Req for #%d msix vectors failed\n", nic->num_vec);
+		return ret;
+	}
+
+	sprintf(nic->irq_name[irq], "%s Mbox", "NICVF");
+	/* Register Misc interrupt */
+	ret = request_irq(pci_irq_vector(nic->pdev, irq),
+			  nicvf_misc_intr_handler, 0, nic->irq_name[irq], nic);
+
+	if (ret)
+		return ret;
+	nic->irq_allocated[irq] = true;
+
+	/* Enable mailbox interrupt */
+	nicvf_enable_intr(nic, NICVF_INTR_MBOX, 0);
+
+	/* Check if VF is able to communicate with PF */
+	if (!nicvf_check_pf_ready(nic)) {
+		nicvf_disable_intr(nic, NICVF_INTR_MBOX, 0);
+		nicvf_unregister_interrupts(nic);
+		return -EIO;
+	}
+
+	return 0;
+}
+
+static netdev_tx_t nicvf_xmit(struct sk_buff *skb, struct net_device *netdev)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+	int qid = skb_get_queue_mapping(skb);
+	struct netdev_queue *txq = netdev_get_tx_queue(netdev, qid);
+	struct nicvf *snic;
+	struct snd_queue *sq;
+	int tmp;
+
+	/* Check for minimum packet length */
+	if (skb->len <= ETH_HLEN) {
+		dev_kfree_skb(skb);
+		return NETDEV_TX_OK;
+	}
+
+	/* In XDP case, initial HW tx queues are used for XDP,
+	 * but stack's queue mapping starts at '0', so skip the
+	 * Tx queues attached to Rx queues for XDP.
+	 */
+	if (nic->xdp_prog)
+		qid += nic->xdp_tx_queues;
+
+	snic = nic;
+	/* Get secondary Qset's SQ structure */
+	if (qid >= MAX_SND_QUEUES_PER_QS) {
+		tmp = qid / MAX_SND_QUEUES_PER_QS;
+		snic = (struct nicvf *)nic->snicvf[tmp - 1];
+		if (!snic) {
+			netdev_warn(nic->netdev,
+				    "Secondary Qset#%d's ptr not initialized\n",
+				    tmp - 1);
+			dev_kfree_skb(skb);
+			return NETDEV_TX_OK;
+		}
+		qid = qid % MAX_SND_QUEUES_PER_QS;
+	}
+
+	sq = &snic->qs->sq[qid];
+	if (!netif_tx_queue_stopped(txq) &&
+	    !nicvf_sq_append_skb(snic, sq, skb, qid)) {
+		netif_tx_stop_queue(txq);
+
+		/* Barrier, so that stop_queue visible to other cpus */
+		smp_mb();
+
+		/* Check again, incase another cpu freed descriptors */
+		if (atomic_read(&sq->free_cnt) > MIN_SQ_DESC_PER_PKT_XMIT) {
+			netif_tx_wake_queue(txq);
+		} else {
+			this_cpu_inc(nic->drv_stats->txq_stop);
+			netif_warn(nic, tx_err, netdev,
+				   "Transmit ring full, stopping SQ%d\n", 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;
+	struct nicvf *nic = netdev_priv(netdev);
+	struct queue_set *qs = nic->qs;
+	struct nicvf_cq_poll *cq_poll = NULL;
+	union nic_mbx mbx = {};
+
+	/* wait till all queued set_rx_mode tasks completes */
+	if (nic->nicvf_rx_mode_wq) {
+		cancel_delayed_work_sync(&nic->link_change_work);
+		drain_workqueue(nic->nicvf_rx_mode_wq);
+	}
+
+	mbx.msg.msg = NIC_MBOX_MSG_SHUTDOWN;
+	nicvf_send_msg_to_pf(nic, &mbx);
+
+	netif_carrier_off(netdev);
+	netif_tx_stop_all_queues(nic->netdev);
+	nic->link_up = false;
+
+	/* 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++) {
+		nicvf_disable_intr(nic, NICVF_INTR_RBDR, qidx);
+		nicvf_clear_intr(nic, NICVF_INTR_RBDR, qidx);
+	}
+	nicvf_disable_intr(nic, NICVF_INTR_QS_ERR, 0);
+	nicvf_clear_intr(nic, NICVF_INTR_QS_ERR, 0);
+
+	/* Wait for pending IRQ handlers to finish */
+	for (irq = 0; irq < nic->num_vec; irq++)
+		synchronize_irq(pci_irq_vector(nic->pdev, irq));
+
+	tasklet_kill(&nic->rbdr_task);
+	tasklet_kill(&nic->qs_err_task);
+	if (nic->rb_work_scheduled)
+		cancel_delayed_work_sync(&nic->rbdr_work);
+
+	for (qidx = 0; qidx < nic->qs->cq_cnt; qidx++) {
+		cq_poll = nic->napi[qidx];
+		if (!cq_poll)
+			continue;
+		napi_synchronize(&cq_poll->napi);
+		/* CQ intr is enabled while napi_complete,
+		 * so disable it now
+		 */
+		nicvf_disable_intr(nic, NICVF_INTR_CQ, qidx);
+		nicvf_clear_intr(nic, NICVF_INTR_CQ, qidx);
+		napi_disable(&cq_poll->napi);
+		netif_napi_del(&cq_poll->napi);
+	}
+
+	netif_tx_disable(netdev);
+
+	for (qidx = 0; qidx < netdev->num_tx_queues; qidx++)
+		netdev_tx_reset_queue(netdev_get_tx_queue(netdev, qidx));
+
+	/* Free resources */
+	nicvf_config_data_transfer(nic, false);
+
+	/* Disable HW Qset */
+	nicvf_qset_config(nic, false);
+
+	/* disable mailbox interrupt */
+	nicvf_disable_intr(nic, NICVF_INTR_MBOX, 0);
+
+	nicvf_unregister_interrupts(nic);
+
+	nicvf_free_cq_poll(nic);
+
+	/* Free any pending SKB saved to receive timestamp */
+	if (nic->ptp_skb) {
+		dev_kfree_skb_any(nic->ptp_skb);
+		nic->ptp_skb = NULL;
+	}
+
+	/* Clear multiqset info */
+	nic->pnicvf = nic;
+
+	return 0;
+}
+
+static int nicvf_config_hw_rx_tstamp(struct nicvf *nic, bool enable)
+{
+	union nic_mbx mbx = {};
+
+	mbx.ptp.msg = NIC_MBOX_MSG_PTP_CFG;
+	mbx.ptp.enable = enable;
+
+	return nicvf_send_msg_to_pf(nic, &mbx);
+}
+
+static int nicvf_update_hw_max_frs(struct nicvf *nic, int mtu)
+{
+	union nic_mbx mbx = {};
+
+	mbx.frs.msg = NIC_MBOX_MSG_SET_MAX_FRS;
+	mbx.frs.max_frs = mtu;
+	mbx.frs.vf_id = nic->vf_id;
+
+	return nicvf_send_msg_to_pf(nic, &mbx);
+}
+
+static void nicvf_link_status_check_task(struct work_struct *work_arg)
+{
+	struct nicvf *nic = container_of(work_arg,
+					 struct nicvf,
+					 link_change_work.work);
+	union nic_mbx mbx = {};
+	mbx.msg.msg = NIC_MBOX_MSG_BGX_LINK_CHANGE;
+	nicvf_send_msg_to_pf(nic, &mbx);
+	queue_delayed_work(nic->nicvf_rx_mode_wq,
+			   &nic->link_change_work, 2 * HZ);
+}
+
+int nicvf_open(struct net_device *netdev)
+{
+	int cpu, err, qidx;
+	struct nicvf *nic = netdev_priv(netdev);
+	struct queue_set *qs = nic->qs;
+	struct nicvf_cq_poll *cq_poll = NULL;
+
+	/* wait till all queued set_rx_mode tasks completes if any */
+	if (nic->nicvf_rx_mode_wq)
+		drain_workqueue(nic->nicvf_rx_mode_wq);
+
+	netif_carrier_off(netdev);
+
+	err = nicvf_register_misc_interrupt(nic);
+	if (err)
+		return err;
+
+	/* Register NAPI handler for processing CQEs */
+	for (qidx = 0; qidx < qs->cq_cnt; qidx++) {
+		cq_poll = kzalloc(sizeof(*cq_poll), GFP_KERNEL);
+		if (!cq_poll) {
+			err = -ENOMEM;
+			goto napi_del;
+		}
+		cq_poll->cq_idx = qidx;
+		cq_poll->nicvf = nic;
+		netif_napi_add(netdev, &cq_poll->napi, nicvf_poll);
+		napi_enable(&cq_poll->napi);
+		nic->napi[qidx] = cq_poll;
+	}
+
+	/* Check if we got MAC address from PF or else generate a radom MAC */
+	if (!nic->sqs_mode && is_zero_ether_addr(netdev->dev_addr)) {
+		eth_hw_addr_random(netdev);
+		nicvf_hw_set_mac_addr(nic, netdev);
+	}
+
+	if (nic->set_mac_pending) {
+		nic->set_mac_pending = false;
+		nicvf_hw_set_mac_addr(nic, netdev);
+	}
+
+	/* Init tasklet for handling Qset err interrupt */
+	tasklet_setup(&nic->qs_err_task, nicvf_handle_qs_err);
+
+	/* Init RBDR tasklet which will refill RBDR */
+	tasklet_setup(&nic->rbdr_task, nicvf_rbdr_task);
+	INIT_DELAYED_WORK(&nic->rbdr_work, nicvf_rbdr_work);
+
+	/* Configure CPI alorithm */
+	nic->cpi_alg = cpi_alg;
+	if (!nic->sqs_mode)
+		nicvf_config_cpi(nic);
+
+	nicvf_request_sqs(nic);
+	if (nic->sqs_mode)
+		nicvf_get_primary_vf_struct(nic);
+
+	/* Configure PTP timestamp */
+	if (nic->ptp_clock)
+		nicvf_config_hw_rx_tstamp(nic, nic->hw_rx_tstamp);
+	atomic_set(&nic->tx_ptp_skbs, 0);
+	nic->ptp_skb = NULL;
+
+	/* Configure receive side scaling and MTU */
+	if (!nic->sqs_mode) {
+		nicvf_rss_init(nic);
+		err = nicvf_update_hw_max_frs(nic, netdev->mtu);
+		if (err)
+			goto cleanup;
+
+		/* Clear percpu stats */
+		for_each_possible_cpu(cpu)
+			memset(per_cpu_ptr(nic->drv_stats, cpu), 0,
+			       sizeof(struct nicvf_drv_stats));
+	}
+
+	err = nicvf_register_interrupts(nic);
+	if (err)
+		goto cleanup;
+
+	/* Initialize the queues */
+	err = nicvf_init_resources(nic);
+	if (err)
+		goto cleanup;
+
+	/* Make sure queue initialization is written */
+	wmb();
+
+	nicvf_reg_write(nic, NIC_VF_INT, -1);
+	/* Enable Qset err interrupt */
+	nicvf_enable_intr(nic, NICVF_INTR_QS_ERR, 0);
+
+	/* Enable completion queue interrupt */
+	for (qidx = 0; qidx < qs->cq_cnt; qidx++)
+		nicvf_enable_intr(nic, NICVF_INTR_CQ, qidx);
+
+	/* Enable RBDR threshold interrupt */
+	for (qidx = 0; qidx < qs->rbdr_cnt; qidx++)
+		nicvf_enable_intr(nic, NICVF_INTR_RBDR, qidx);
+
+	/* Send VF config done msg to PF */
+	nicvf_send_cfg_done(nic);
+
+	if (nic->nicvf_rx_mode_wq) {
+		INIT_DELAYED_WORK(&nic->link_change_work,
+				  nicvf_link_status_check_task);
+		queue_delayed_work(nic->nicvf_rx_mode_wq,
+				   &nic->link_change_work, 0);
+	}
+
+	return 0;
+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];
+		if (!cq_poll)
+			continue;
+		napi_disable(&cq_poll->napi);
+		netif_napi_del(&cq_poll->napi);
+	}
+	nicvf_free_cq_poll(nic);
+	return err;
+}
+
+static int nicvf_change_mtu(struct net_device *netdev, int new_mtu)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+	int orig_mtu = netdev->mtu;
+
+	/* For now just support only the usual MTU sized frames,
+	 * plus some headroom for VLAN, QinQ.
+	 */
+	if (nic->xdp_prog && new_mtu > MAX_XDP_MTU) {
+		netdev_warn(netdev, "Jumbo frames not yet supported with XDP, current MTU %d.\n",
+			    netdev->mtu);
+		return -EINVAL;
+	}
+
+	netdev->mtu = new_mtu;
+
+	if (!netif_running(netdev))
+		return 0;
+
+	if (nicvf_update_hw_max_frs(nic, new_mtu)) {
+		netdev->mtu = orig_mtu;
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int nicvf_set_mac_address(struct net_device *netdev, void *p)
+{
+	struct sockaddr *addr = p;
+	struct nicvf *nic = netdev_priv(netdev);
+
+	if (!is_valid_ether_addr(addr->sa_data))
+		return -EADDRNOTAVAIL;
+
+	eth_hw_addr_set(netdev, addr->sa_data);
+
+	if (nic->pdev->msix_enabled) {
+		if (nicvf_hw_set_mac_addr(nic, netdev))
+			return -EBUSY;
+	} else {
+		nic->set_mac_pending = true;
+	}
+
+	return 0;
+}
+
+void nicvf_update_lmac_stats(struct nicvf *nic)
+{
+	int stat = 0;
+	union nic_mbx mbx = {};
+
+	if (!netif_running(nic->netdev))
+		return;
+
+	mbx.bgx_stats.msg = NIC_MBOX_MSG_BGX_STATS;
+	mbx.bgx_stats.vf_id = nic->vf_id;
+	/* Rx stats */
+	mbx.bgx_stats.rx = 1;
+	while (stat < BGX_RX_STATS_COUNT) {
+		mbx.bgx_stats.idx = stat;
+		if (nicvf_send_msg_to_pf(nic, &mbx))
+			return;
+		stat++;
+	}
+
+	stat = 0;
+
+	/* Tx stats */
+	mbx.bgx_stats.rx = 0;
+	while (stat < BGX_TX_STATS_COUNT) {
+		mbx.bgx_stats.idx = stat;
+		if (nicvf_send_msg_to_pf(nic, &mbx))
+			return;
+		stat++;
+	}
+}
+
+void nicvf_update_stats(struct nicvf *nic)
+{
+	int qidx, cpu;
+	u64 tmp_stats = 0;
+	struct nicvf_hw_stats *stats = &nic->hw_stats;
+	struct nicvf_drv_stats *drv_stats;
+	struct queue_set *qs = nic->qs;
+
+#define GET_RX_STATS(reg) \
+	nicvf_reg_read(nic, NIC_VNIC_RX_STAT_0_13 | (reg << 3))
+#define GET_TX_STATS(reg) \
+	nicvf_reg_read(nic, NIC_VNIC_TX_STAT_0_4 | (reg << 3))
+
+	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);
+	stats->rx_drop_l3_mcast = GET_RX_STATS(RX_DRP_L3MCAST);
+
+	stats->tx_bytes = GET_TX_STATS(TX_OCTS);
+	stats->tx_ucast_frames = GET_TX_STATS(TX_UCAST);
+	stats->tx_bcast_frames = GET_TX_STATS(TX_BCAST);
+	stats->tx_mcast_frames = GET_TX_STATS(TX_MCAST);
+	stats->tx_drops = GET_TX_STATS(TX_DROP);
+
+	/* On T88 pass 2.0, the dummy SQE added for TSO notification
+	 * via CQE has 'dont_send' set. Hence HW drops the pkt pointed
+	 * pointed by dummy SQE and results in tx_drops counter being
+	 * incremented. Subtracting it from tx_tso counter will give
+	 * exact tx_drops counter.
+	 */
+	if (nic->t88 && nic->hw_tso) {
+		for_each_possible_cpu(cpu) {
+			drv_stats = per_cpu_ptr(nic->drv_stats, cpu);
+			tmp_stats += drv_stats->tx_tso;
+		}
+		stats->tx_drops = tmp_stats - stats->tx_drops;
+	}
+	stats->tx_frames = stats->tx_ucast_frames +
+			   stats->tx_bcast_frames +
+			   stats->tx_mcast_frames;
+	stats->rx_frames = stats->rx_ucast_frames +
+			   stats->rx_bcast_frames +
+			   stats->rx_mcast_frames;
+	stats->rx_drops = stats->rx_drop_red +
+			  stats->rx_drop_overrun;
+
+	/* Update RQ and SQ stats */
+	for (qidx = 0; qidx < qs->rq_cnt; qidx++)
+		nicvf_update_rq_stats(nic, qidx);
+	for (qidx = 0; qidx < qs->sq_cnt; qidx++)
+		nicvf_update_sq_stats(nic, qidx);
+}
+
+static void 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->hw_stats;
+
+	nicvf_update_stats(nic);
+
+	stats->rx_bytes = hw_stats->rx_bytes;
+	stats->rx_packets = hw_stats->rx_frames;
+	stats->rx_dropped = hw_stats->rx_drops;
+	stats->multicast = hw_stats->rx_mcast_frames;
+
+	stats->tx_bytes = hw_stats->tx_bytes;
+	stats->tx_packets = hw_stats->tx_frames;
+	stats->tx_dropped = hw_stats->tx_drops;
+
+}
+
+static void nicvf_tx_timeout(struct net_device *dev, unsigned int txqueue)
+{
+	struct nicvf *nic = netdev_priv(dev);
+
+	netif_warn(nic, tx_err, dev, "Transmit timed out, resetting\n");
+
+	this_cpu_inc(nic->drv_stats->tx_timeout);
+	schedule_work(&nic->reset_task);
+}
+
+static void nicvf_reset_task(struct work_struct *work)
+{
+	struct nicvf *nic;
+
+	nic = container_of(work, struct nicvf, reset_task);
+
+	if (!netif_running(nic->netdev))
+		return;
+
+	nicvf_stop(nic->netdev);
+	nicvf_open(nic->netdev);
+	netif_trans_update(nic->netdev);
+}
+
+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 void nicvf_set_xdp_queues(struct nicvf *nic, bool bpf_attached)
+{
+	u8 cq_count, txq_count;
+
+	/* Set XDP Tx queue count same as Rx queue count */
+	if (!bpf_attached)
+		nic->xdp_tx_queues = 0;
+	else
+		nic->xdp_tx_queues = nic->rx_queues;
+
+	/* If queue count > MAX_CMP_QUEUES_PER_QS, then additional qsets
+	 * needs to be allocated, check how many.
+	 */
+	txq_count = nic->xdp_tx_queues + nic->tx_queues;
+	cq_count = max(nic->rx_queues, txq_count);
+	if (cq_count > MAX_CMP_QUEUES_PER_QS) {
+		nic->sqs_count = roundup(cq_count, MAX_CMP_QUEUES_PER_QS);
+		nic->sqs_count = (nic->sqs_count / MAX_CMP_QUEUES_PER_QS) - 1;
+	} else {
+		nic->sqs_count = 0;
+	}
+
+	/* Set primary Qset's resources */
+	nic->qs->rq_cnt = min_t(u8, nic->rx_queues, MAX_RCV_QUEUES_PER_QS);
+	nic->qs->sq_cnt = min_t(u8, txq_count, MAX_SND_QUEUES_PER_QS);
+	nic->qs->cq_cnt = max_t(u8, nic->qs->rq_cnt, nic->qs->sq_cnt);
+
+	/* Update stack */
+	nicvf_set_real_num_queues(nic->netdev, nic->tx_queues, nic->rx_queues);
+}
+
+static int nicvf_xdp_setup(struct nicvf *nic, struct bpf_prog *prog)
+{
+	struct net_device *dev = nic->netdev;
+	bool if_up = netif_running(nic->netdev);
+	struct bpf_prog *old_prog;
+	bool bpf_attached = false;
+	int ret = 0;
+
+	/* For now just support only the usual MTU sized frames,
+	 * plus some headroom for VLAN, QinQ.
+	 */
+	if (prog && dev->mtu > MAX_XDP_MTU) {
+		netdev_warn(dev, "Jumbo frames not yet supported with XDP, current MTU %d.\n",
+			    dev->mtu);
+		return -EOPNOTSUPP;
+	}
+
+	/* ALL SQs attached to CQs i.e same as RQs, are treated as
+	 * XDP Tx queues and more Tx queues are allocated for
+	 * network stack to send pkts out.
+	 *
+	 * No of Tx queues are either same as Rx queues or whatever
+	 * is left in max no of queues possible.
+	 */
+	if ((nic->rx_queues + nic->tx_queues) > nic->max_queues) {
+		netdev_warn(dev,
+			    "Failed to attach BPF prog, RXQs + TXQs > Max %d\n",
+			    nic->max_queues);
+		return -ENOMEM;
+	}
+
+	if (if_up)
+		nicvf_stop(nic->netdev);
+
+	old_prog = xchg(&nic->xdp_prog, prog);
+	/* Detach old prog, if any */
+	if (old_prog)
+		bpf_prog_put(old_prog);
+
+	if (nic->xdp_prog) {
+		/* Attach BPF program */
+		bpf_prog_add(nic->xdp_prog, nic->rx_queues - 1);
+		bpf_attached = true;
+	}
+
+	/* Calculate Tx queues needed for XDP and network stack */
+	nicvf_set_xdp_queues(nic, bpf_attached);
+
+	if (if_up) {
+		/* Reinitialize interface, clean slate */
+		nicvf_open(nic->netdev);
+		netif_trans_update(nic->netdev);
+	}
+
+	return ret;
+}
+
+static int nicvf_xdp(struct net_device *netdev, struct netdev_bpf *xdp)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+
+	/* To avoid checks while retrieving buffer address from CQE_RX,
+	 * do not support XDP for T88 pass1.x silicons which are anyway
+	 * not in use widely.
+	 */
+	if (pass1_silicon(nic->pdev))
+		return -EOPNOTSUPP;
+
+	switch (xdp->command) {
+	case XDP_SETUP_PROG:
+		return nicvf_xdp_setup(nic, xdp->prog);
+	default:
+		return -EINVAL;
+	}
+}
+
+static int nicvf_config_hwtstamp(struct net_device *netdev, struct ifreq *ifr)
+{
+	struct hwtstamp_config config;
+	struct nicvf *nic = netdev_priv(netdev);
+
+	if (!nic->ptp_clock)
+		return -ENODEV;
+
+	if (copy_from_user(&config, ifr->ifr_data, sizeof(config)))
+		return -EFAULT;
+
+	switch (config.tx_type) {
+	case HWTSTAMP_TX_OFF:
+	case HWTSTAMP_TX_ON:
+		break;
+	default:
+		return -ERANGE;
+	}
+
+	switch (config.rx_filter) {
+	case HWTSTAMP_FILTER_NONE:
+		nic->hw_rx_tstamp = false;
+		break;
+	case HWTSTAMP_FILTER_ALL:
+	case HWTSTAMP_FILTER_SOME:
+	case HWTSTAMP_FILTER_PTP_V1_L4_EVENT:
+	case HWTSTAMP_FILTER_PTP_V1_L4_SYNC:
+	case HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_L4_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_L4_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_L2_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_L2_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_L2_DELAY_REQ:
+	case HWTSTAMP_FILTER_PTP_V2_EVENT:
+	case HWTSTAMP_FILTER_PTP_V2_SYNC:
+	case HWTSTAMP_FILTER_PTP_V2_DELAY_REQ:
+		nic->hw_rx_tstamp = true;
+		config.rx_filter = HWTSTAMP_FILTER_ALL;
+		break;
+	default:
+		return -ERANGE;
+	}
+
+	if (netif_running(netdev))
+		nicvf_config_hw_rx_tstamp(nic, nic->hw_rx_tstamp);
+
+	if (copy_to_user(ifr->ifr_data, &config, sizeof(config)))
+		return -EFAULT;
+
+	return 0;
+}
+
+static int nicvf_ioctl(struct net_device *netdev, struct ifreq *req, int cmd)
+{
+	switch (cmd) {
+	case SIOCSHWTSTAMP:
+		return nicvf_config_hwtstamp(netdev, req);
+	default:
+		return -EOPNOTSUPP;
+	}
+}
+
+static void __nicvf_set_rx_mode_task(u8 mode, struct xcast_addr_list *mc_addrs,
+				     struct nicvf *nic)
+{
+	union nic_mbx mbx = {};
+	int idx;
+
+	/* From the inside of VM code flow we have only 128 bits memory
+	 * available to send message to host's PF, so send all mc addrs
+	 * one by one, starting from flush command in case if kernel
+	 * requests to configure specific MAC filtering
+	 */
+
+	/* flush DMAC filters and reset RX mode */
+	mbx.xcast.msg = NIC_MBOX_MSG_RESET_XCAST;
+	if (nicvf_send_msg_to_pf(nic, &mbx) < 0)
+		goto free_mc;
+
+	if (mode & BGX_XCAST_MCAST_FILTER) {
+		/* once enabling filtering, we need to signal to PF to add
+		 * its' own LMAC to the filter to accept packets for it.
+		 */
+		mbx.xcast.msg = NIC_MBOX_MSG_ADD_MCAST;
+		mbx.xcast.mac = 0;
+		if (nicvf_send_msg_to_pf(nic, &mbx) < 0)
+			goto free_mc;
+	}
+
+	/* check if we have any specific MACs to be added to PF DMAC filter */
+	if (mc_addrs) {
+		/* now go through kernel list of MACs and add them one by one */
+		for (idx = 0; idx < mc_addrs->count; idx++) {
+			mbx.xcast.msg = NIC_MBOX_MSG_ADD_MCAST;
+			mbx.xcast.mac = mc_addrs->mc[idx];
+			if (nicvf_send_msg_to_pf(nic, &mbx) < 0)
+				goto free_mc;
+		}
+	}
+
+	/* and finally set rx mode for PF accordingly */
+	mbx.xcast.msg = NIC_MBOX_MSG_SET_XCAST;
+	mbx.xcast.mode = mode;
+
+	nicvf_send_msg_to_pf(nic, &mbx);
+free_mc:
+	kfree(mc_addrs);
+}
+
+static void nicvf_set_rx_mode_task(struct work_struct *work_arg)
+{
+	struct nicvf_work *vf_work = container_of(work_arg, struct nicvf_work,
+						  work);
+	struct nicvf *nic = container_of(vf_work, struct nicvf, rx_mode_work);
+	u8 mode;
+	struct xcast_addr_list *mc;
+
+	/* Save message data locally to prevent them from
+	 * being overwritten by next ndo_set_rx_mode call().
+	 */
+	spin_lock_bh(&nic->rx_mode_wq_lock);
+	mode = vf_work->mode;
+	mc = vf_work->mc;
+	vf_work->mc = NULL;
+	spin_unlock_bh(&nic->rx_mode_wq_lock);
+
+	__nicvf_set_rx_mode_task(mode, mc, nic);
+}
+
+static void nicvf_set_rx_mode(struct net_device *netdev)
+{
+	struct nicvf *nic = netdev_priv(netdev);
+	struct netdev_hw_addr *ha;
+	struct xcast_addr_list *mc_list = NULL;
+	u8 mode = 0;
+
+	if (netdev->flags & IFF_PROMISC) {
+		mode = BGX_XCAST_BCAST_ACCEPT | BGX_XCAST_MCAST_ACCEPT;
+	} else {
+		if (netdev->flags & IFF_BROADCAST)
+			mode |= BGX_XCAST_BCAST_ACCEPT;
+
+		if (netdev->flags & IFF_ALLMULTI) {
+			mode |= BGX_XCAST_MCAST_ACCEPT;
+		} else if (netdev->flags & IFF_MULTICAST) {
+			mode |= BGX_XCAST_MCAST_FILTER;
+			/* here we need to copy mc addrs */
+			if (netdev_mc_count(netdev)) {
+				mc_list = kmalloc(struct_size(mc_list, mc,
+							      netdev_mc_count(netdev)),
+						  GFP_ATOMIC);
+				if (unlikely(!mc_list))
+					return;
+				mc_list->count = 0;
+				netdev_hw_addr_list_for_each(ha, &netdev->mc) {
+					mc_list->mc[mc_list->count] =
+						ether_addr_to_u64(ha->addr);
+					mc_list->count++;
+				}
+			}
+		}
+	}
+	spin_lock(&nic->rx_mode_wq_lock);
+	kfree(nic->rx_mode_work.mc);
+	nic->rx_mode_work.mc = mc_list;
+	nic->rx_mode_work.mode = mode;
+	queue_work(nic->nicvf_rx_mode_wq, &nic->rx_mode_work.work);
+	spin_unlock(&nic->rx_mode_wq_lock);
+}
+
+static const struct net_device_ops nicvf_netdev_ops = {
+	.ndo_open		= nicvf_open,
+	.ndo_stop		= nicvf_stop,
+	.ndo_start_xmit		= nicvf_xmit,
+	.ndo_change_mtu		= nicvf_change_mtu,
+	.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,
+	.ndo_bpf		= nicvf_xdp,
+	.ndo_eth_ioctl           = nicvf_ioctl,
+	.ndo_set_rx_mode        = nicvf_set_rx_mode,
+};
+
+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;
+	int    err, qcount;
+	u16    sdevid;
+	struct cavium_ptp *ptp_clock;
+
+	ptp_clock = cavium_ptp_get();
+	if (IS_ERR(ptp_clock)) {
+		if (PTR_ERR(ptp_clock) == -ENODEV)
+			/* In virtualized environment we proceed without ptp */
+			ptp_clock = NULL;
+		else
+			return PTR_ERR(ptp_clock);
+	}
+
+	err = pci_enable_device(pdev);
+	if (err)
+		return dev_err_probe(dev, err, "Failed to enable PCI device\n");
+
+	err = pci_request_regions(pdev, DRV_NAME);
+	if (err) {
+		dev_err(dev, "PCI request regions failed 0x%x\n", err);
+		goto err_disable_device;
+	}
+
+	err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(48));
+	if (err) {
+		dev_err(dev, "Unable to get usable DMA configuration\n");
+		goto err_release_regions;
+	}
+
+	qcount = netif_get_num_default_rss_queues();
+
+	/* Restrict multiqset support only for host bound VFs */
+	if (pdev->is_virtfn) {
+		/* Set max number of queues per VF */
+		qcount = min_t(int, num_online_cpus(),
+			       (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;
+	}
+
+	pci_set_drvdata(pdev, netdev);
+
+	SET_NETDEV_DEV(netdev, &pdev->dev);
+
+	nic = netdev_priv(netdev);
+	nic->netdev = netdev;
+	nic->pdev = pdev;
+	nic->pnicvf = nic;
+	nic->max_queues = qcount;
+	/* If no of CPUs are too low, there won't be any queues left
+	 * for XDP_TX, hence double it.
+	 */
+	if (!nic->t88)
+		nic->max_queues *= 2;
+	nic->ptp_clock = ptp_clock;
+
+	/* Initialize mutex that serializes usage of VF's mailbox */
+	mutex_init(&nic->rx_mode_mtx);
+
+	/* MAP VF's configuration registers */
+	nic->reg_base = pcim_iomap(pdev, PCI_CFG_REG_BAR_NUM, 0);
+	if (!nic->reg_base) {
+		dev_err(dev, "Cannot map config register space, aborting\n");
+		err = -ENOMEM;
+		goto err_free_netdev;
+	}
+
+	nic->drv_stats = netdev_alloc_pcpu_stats(struct nicvf_drv_stats);
+	if (!nic->drv_stats) {
+		err = -ENOMEM;
+		goto err_free_netdev;
+	}
+
+	err = nicvf_set_qset_resources(nic);
+	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;
+
+	nicvf_send_vf_struct(nic);
+
+	if (!pass1_silicon(nic->pdev))
+		nic->hw_tso = true;
+
+	/* Get iommu domain for iova to physical addr conversion */
+	nic->iommu_domain = iommu_get_domain_for_dev(dev);
+
+	pci_read_config_word(nic->pdev, PCI_SUBSYSTEM_ID, &sdevid);
+	if (sdevid == 0xA134)
+		nic->t88 = true;
+
+	/* 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_SG |
+			       NETIF_F_TSO | NETIF_F_GRO | NETIF_F_TSO6 |
+			       NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
+			       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_IPV6_CSUM | NETIF_F_TSO | NETIF_F_TSO6;
+
+	netdev->netdev_ops = &nicvf_netdev_ops;
+	netdev->watchdog_timeo = NICVF_TX_TIMEOUT;
+
+	/* MTU range: 64 - 9200 */
+	netdev->min_mtu = NIC_HW_MIN_FRS;
+	netdev->max_mtu = NIC_HW_MAX_FRS;
+
+	INIT_WORK(&nic->reset_task, nicvf_reset_task);
+
+	nic->nicvf_rx_mode_wq = alloc_ordered_workqueue("nicvf_rx_mode_wq_VF%d",
+							WQ_MEM_RECLAIM,
+							nic->vf_id);
+	if (!nic->nicvf_rx_mode_wq) {
+		err = -ENOMEM;
+		dev_err(dev, "Failed to allocate work queue\n");
+		goto err_unregister_interrupts;
+	}
+
+	INIT_WORK(&nic->rx_mode_work.work, nicvf_set_rx_mode_task);
+	spin_lock_init(&nic->rx_mode_wq_lock);
+
+	err = register_netdev(netdev);
+	if (err) {
+		dev_err(dev, "Failed to register netdevice\n");
+		goto err_destroy_workqueue;
+	}
+
+	nic->msg_enable = debug;
+
+	nicvf_set_ethtool_ops(netdev);
+
+	return 0;
+
+err_destroy_workqueue:
+	destroy_workqueue(nic->nicvf_rx_mode_wq);
+err_unregister_interrupts:
+	nicvf_unregister_interrupts(nic);
+err_free_netdev:
+	pci_set_drvdata(pdev, NULL);
+	if (nic->drv_stats)
+		free_percpu(nic->drv_stats);
+	free_netdev(netdev);
+err_release_regions:
+	pci_release_regions(pdev);
+err_disable_device:
+	pci_disable_device(pdev);
+	return err;
+}
+
+static void nicvf_remove(struct pci_dev *pdev)
+{
+	struct net_device *netdev = pci_get_drvdata(pdev);
+	struct nicvf *nic;
+	struct net_device *pnetdev;
+
+	if (!netdev)
+		return;
+
+	nic = netdev_priv(netdev);
+	pnetdev = nic->pnicvf->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);
+	if (nic->nicvf_rx_mode_wq) {
+		destroy_workqueue(nic->nicvf_rx_mode_wq);
+		nic->nicvf_rx_mode_wq = NULL;
+	}
+	nicvf_unregister_interrupts(nic);
+	pci_set_drvdata(pdev, NULL);
+	if (nic->drv_stats)
+		free_percpu(nic->drv_stats);
+	cavium_ptp_put(nic->ptp_clock);
+	free_netdev(netdev);
+	pci_release_regions(pdev);
+	pci_disable_device(pdev);
+}
+
+static void nicvf_shutdown(struct pci_dev *pdev)
+{
+	nicvf_remove(pdev);
+}
+
+static struct pci_driver nicvf_driver = {
+	.name = DRV_NAME,
+	.id_table = nicvf_id_table,
+	.probe = nicvf_probe,
+	.remove = nicvf_remove,
+	.shutdown = nicvf_shutdown,
+};
+
+static int __init nicvf_init_module(void)
+{
+	pr_info("%s, ver %s\n", DRV_NAME, DRV_VERSION);
+	return pci_register_driver(&nicvf_driver);
+}
+
+static void __exit nicvf_cleanup_module(void)
+{
+	pci_unregister_driver(&nicvf_driver);
+}
+
+module_init(nicvf_init_module);
+module_exit(nicvf_cleanup_module);
diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_queues.c b/drivers/net/ethernet/cavium/thunder/nicvf_queues.c
new file mode 100644
index 000000000..06397cc8b
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/nicvf_queues.c
@@ -0,0 +1,1972 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2015 Cavium, Inc.
+ */
+
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include <linux/ip.h>
+#include <linux/etherdevice.h>
+#include <linux/iommu.h>
+#include <net/ip.h>
+#include <net/tso.h>
+#include <uapi/linux/bpf.h>
+
+#include "nic_reg.h"
+#include "nic.h"
+#include "q_struct.h"
+#include "nicvf_queues.h"
+
+static inline void nicvf_sq_add_gather_subdesc(struct snd_queue *sq, int qentry,
+					       int size, u64 data);
+static void nicvf_get_page(struct nicvf *nic)
+{
+	if (!nic->rb_pageref || !nic->rb_page)
+		return;
+
+	page_ref_add(nic->rb_page, nic->rb_pageref);
+	nic->rb_pageref = 0;
+}
+
+/* Poll a register for a specific value */
+static int nicvf_poll_reg(struct nicvf *nic, int qidx,
+			  u64 reg, int bit_pos, int bits, int val)
+{
+	u64 bit_mask;
+	u64 reg_val;
+	int timeout = 10;
+
+	bit_mask = (1ULL << bits) - 1;
+	bit_mask = (bit_mask << bit_pos);
+
+	while (timeout) {
+		reg_val = nicvf_queue_reg_read(nic, reg, qidx);
+		if (((reg_val & bit_mask) >> bit_pos) == val)
+			return 0;
+		usleep_range(1000, 2000);
+		timeout--;
+	}
+	netdev_err(nic->netdev, "Poll on reg 0x%llx failed\n", reg);
+	return 1;
+}
+
+/* Allocate memory for a queue's descriptors */
+static int nicvf_alloc_q_desc_mem(struct nicvf *nic, struct q_desc_mem *dmem,
+				  int q_len, int desc_size, int align_bytes)
+{
+	dmem->q_len = q_len;
+	dmem->size = (desc_size * q_len) + align_bytes;
+	/* Save address, need it while freeing */
+	dmem->unalign_base = dma_alloc_coherent(&nic->pdev->dev, dmem->size,
+						&dmem->dma, GFP_KERNEL);
+	if (!dmem->unalign_base)
+		return -ENOMEM;
+
+	/* Align memory address for 'align_bytes' */
+	dmem->phys_base = NICVF_ALIGNED_ADDR((u64)dmem->dma, align_bytes);
+	dmem->base = dmem->unalign_base + (dmem->phys_base - dmem->dma);
+	return 0;
+}
+
+/* Free queue's descriptor memory */
+static void nicvf_free_q_desc_mem(struct nicvf *nic, struct q_desc_mem *dmem)
+{
+	if (!dmem)
+		return;
+
+	dma_free_coherent(&nic->pdev->dev, dmem->size,
+			  dmem->unalign_base, dmem->dma);
+	dmem->unalign_base = NULL;
+	dmem->base = NULL;
+}
+
+#define XDP_PAGE_REFCNT_REFILL 256
+
+/* Allocate a new page or recycle one if possible
+ *
+ * We cannot optimize dma mapping here, since
+ * 1. It's only one RBDR ring for 8 Rx queues.
+ * 2. CQE_RX gives address of the buffer where pkt has been DMA'ed
+ *    and not idx into RBDR ring, so can't refer to saved info.
+ * 3. There are multiple receive buffers per page
+ */
+static inline struct pgcache *nicvf_alloc_page(struct nicvf *nic,
+					       struct rbdr *rbdr, gfp_t gfp)
+{
+	int ref_count;
+	struct page *page = NULL;
+	struct pgcache *pgcache, *next;
+
+	/* Check if page is already allocated */
+	pgcache = &rbdr->pgcache[rbdr->pgidx];
+	page = pgcache->page;
+	/* Check if page can be recycled */
+	if (page) {
+		ref_count = page_ref_count(page);
+		/* This page can be recycled if internal ref_count and page's
+		 * ref_count are equal, indicating that the page has been used
+		 * once for packet transmission. For non-XDP mode, internal
+		 * ref_count is always '1'.
+		 */
+		if (rbdr->is_xdp) {
+			if (ref_count == pgcache->ref_count)
+				pgcache->ref_count--;
+			else
+				page = NULL;
+		} else if (ref_count != 1) {
+			page = NULL;
+		}
+	}
+
+	if (!page) {
+		page = alloc_pages(gfp | __GFP_COMP | __GFP_NOWARN, 0);
+		if (!page)
+			return NULL;
+
+		this_cpu_inc(nic->pnicvf->drv_stats->page_alloc);
+
+		/* Check for space */
+		if (rbdr->pgalloc >= rbdr->pgcnt) {
+			/* Page can still be used */
+			nic->rb_page = page;
+			return NULL;
+		}
+
+		/* Save the page in page cache */
+		pgcache->page = page;
+		pgcache->dma_addr = 0;
+		pgcache->ref_count = 0;
+		rbdr->pgalloc++;
+	}
+
+	/* Take additional page references for recycling */
+	if (rbdr->is_xdp) {
+		/* Since there is single RBDR (i.e single core doing
+		 * page recycling) per 8 Rx queues, in XDP mode adjusting
+		 * page references atomically is the biggest bottleneck, so
+		 * take bunch of references at a time.
+		 *
+		 * So here, below reference counts defer by '1'.
+		 */
+		if (!pgcache->ref_count) {
+			pgcache->ref_count = XDP_PAGE_REFCNT_REFILL;
+			page_ref_add(page, XDP_PAGE_REFCNT_REFILL);
+		}
+	} else {
+		/* In non-XDP case, single 64K page is divided across multiple
+		 * receive buffers, so cost of recycling is less anyway.
+		 * So we can do with just one extra reference.
+		 */
+		page_ref_add(page, 1);
+	}
+
+	rbdr->pgidx++;
+	rbdr->pgidx &= (rbdr->pgcnt - 1);
+
+	/* Prefetch refcount of next page in page cache */
+	next = &rbdr->pgcache[rbdr->pgidx];
+	page = next->page;
+	if (page)
+		prefetch(&page->_refcount);
+
+	return pgcache;
+}
+
+/* Allocate buffer for packet reception */
+static inline int nicvf_alloc_rcv_buffer(struct nicvf *nic, struct rbdr *rbdr,
+					 gfp_t gfp, u32 buf_len, u64 *rbuf)
+{
+	struct pgcache *pgcache = NULL;
+
+	/* Check if request can be accomodated in previous allocated page.
+	 * But in XDP mode only one buffer per page is permitted.
+	 */
+	if (!rbdr->is_xdp && nic->rb_page &&
+	    ((nic->rb_page_offset + buf_len) <= PAGE_SIZE)) {
+		nic->rb_pageref++;
+		goto ret;
+	}
+
+	nicvf_get_page(nic);
+	nic->rb_page = NULL;
+
+	/* Get new page, either recycled or new one */
+	pgcache = nicvf_alloc_page(nic, rbdr, gfp);
+	if (!pgcache && !nic->rb_page) {
+		this_cpu_inc(nic->pnicvf->drv_stats->rcv_buffer_alloc_failures);
+		return -ENOMEM;
+	}
+
+	nic->rb_page_offset = 0;
+
+	/* Reserve space for header modifications by BPF program */
+	if (rbdr->is_xdp)
+		buf_len += XDP_PACKET_HEADROOM;
+
+	/* Check if it's recycled */
+	if (pgcache)
+		nic->rb_page = pgcache->page;
+ret:
+	if (rbdr->is_xdp && pgcache && pgcache->dma_addr) {
+		*rbuf = pgcache->dma_addr;
+	} else {
+		/* HW will ensure data coherency, CPU sync not required */
+		*rbuf = (u64)dma_map_page_attrs(&nic->pdev->dev, nic->rb_page,
+						nic->rb_page_offset, buf_len,
+						DMA_FROM_DEVICE,
+						DMA_ATTR_SKIP_CPU_SYNC);
+		if (dma_mapping_error(&nic->pdev->dev, (dma_addr_t)*rbuf)) {
+			if (!nic->rb_page_offset)
+				__free_pages(nic->rb_page, 0);
+			nic->rb_page = NULL;
+			return -ENOMEM;
+		}
+		if (pgcache)
+			pgcache->dma_addr = *rbuf + XDP_PACKET_HEADROOM;
+		nic->rb_page_offset += buf_len;
+	}
+
+	return 0;
+}
+
+/* Build skb around receive buffer */
+static struct sk_buff *nicvf_rb_ptr_to_skb(struct nicvf *nic,
+					   u64 rb_ptr, int len)
+{
+	void *data;
+	struct sk_buff *skb;
+
+	data = phys_to_virt(rb_ptr);
+
+	/* Now build an skb to give to stack */
+	skb = build_skb(data, RCV_FRAG_LEN);
+	if (!skb) {
+		put_page(virt_to_page(data));
+		return NULL;
+	}
+
+	prefetch(skb->data);
+	return skb;
+}
+
+/* Allocate RBDR ring and populate receive buffers */
+static int  nicvf_init_rbdr(struct nicvf *nic, struct rbdr *rbdr,
+			    int ring_len, int buf_size)
+{
+	int idx;
+	u64 rbuf;
+	struct rbdr_entry_t *desc;
+	int err;
+
+	err = nicvf_alloc_q_desc_mem(nic, &rbdr->dmem, ring_len,
+				     sizeof(struct rbdr_entry_t),
+				     NICVF_RCV_BUF_ALIGN_BYTES);
+	if (err)
+		return err;
+
+	rbdr->desc = rbdr->dmem.base;
+	/* Buffer size has to be in multiples of 128 bytes */
+	rbdr->dma_size = buf_size;
+	rbdr->enable = true;
+	rbdr->thresh = RBDR_THRESH;
+	rbdr->head = 0;
+	rbdr->tail = 0;
+
+	/* Initialize page recycling stuff.
+	 *
+	 * Can't use single buffer per page especially with 64K pages.
+	 * On embedded platforms i.e 81xx/83xx available memory itself
+	 * is low and minimum ring size of RBDR is 8K, that takes away
+	 * lots of memory.
+	 *
+	 * But for XDP it has to be a single buffer per page.
+	 */
+	if (!nic->pnicvf->xdp_prog) {
+		rbdr->pgcnt = ring_len / (PAGE_SIZE / buf_size);
+		rbdr->is_xdp = false;
+	} else {
+		rbdr->pgcnt = ring_len;
+		rbdr->is_xdp = true;
+	}
+	rbdr->pgcnt = roundup_pow_of_two(rbdr->pgcnt);
+	rbdr->pgcache = kcalloc(rbdr->pgcnt, sizeof(*rbdr->pgcache),
+				GFP_KERNEL);
+	if (!rbdr->pgcache)
+		return -ENOMEM;
+	rbdr->pgidx = 0;
+	rbdr->pgalloc = 0;
+
+	nic->rb_page = NULL;
+	for (idx = 0; idx < ring_len; idx++) {
+		err = nicvf_alloc_rcv_buffer(nic, rbdr, GFP_KERNEL,
+					     RCV_FRAG_LEN, &rbuf);
+		if (err) {
+			/* To free already allocated and mapped ones */
+			rbdr->tail = idx - 1;
+			return err;
+		}
+
+		desc = GET_RBDR_DESC(rbdr, idx);
+		desc->buf_addr = rbuf & ~(NICVF_RCV_BUF_ALIGN_BYTES - 1);
+	}
+
+	nicvf_get_page(nic);
+
+	return 0;
+}
+
+/* Free RBDR ring and its receive buffers */
+static void nicvf_free_rbdr(struct nicvf *nic, struct rbdr *rbdr)
+{
+	int head, tail;
+	u64 buf_addr, phys_addr;
+	struct pgcache *pgcache;
+	struct rbdr_entry_t *desc;
+
+	if (!rbdr)
+		return;
+
+	rbdr->enable = false;
+	if (!rbdr->dmem.base)
+		return;
+
+	head = rbdr->head;
+	tail = rbdr->tail;
+
+	/* Release page references */
+	while (head != tail) {
+		desc = GET_RBDR_DESC(rbdr, head);
+		buf_addr = desc->buf_addr;
+		phys_addr = nicvf_iova_to_phys(nic, buf_addr);
+		dma_unmap_page_attrs(&nic->pdev->dev, buf_addr, RCV_FRAG_LEN,
+				     DMA_FROM_DEVICE, DMA_ATTR_SKIP_CPU_SYNC);
+		if (phys_addr)
+			put_page(virt_to_page(phys_to_virt(phys_addr)));
+		head++;
+		head &= (rbdr->dmem.q_len - 1);
+	}
+	/* Release buffer of tail desc */
+	desc = GET_RBDR_DESC(rbdr, tail);
+	buf_addr = desc->buf_addr;
+	phys_addr = nicvf_iova_to_phys(nic, buf_addr);
+	dma_unmap_page_attrs(&nic->pdev->dev, buf_addr, RCV_FRAG_LEN,
+			     DMA_FROM_DEVICE, DMA_ATTR_SKIP_CPU_SYNC);
+	if (phys_addr)
+		put_page(virt_to_page(phys_to_virt(phys_addr)));
+
+	/* Sync page cache info */
+	smp_rmb();
+
+	/* Release additional page references held for recycling */
+	head = 0;
+	while (head < rbdr->pgcnt) {
+		pgcache = &rbdr->pgcache[head];
+		if (pgcache->page && page_ref_count(pgcache->page) != 0) {
+			if (rbdr->is_xdp) {
+				page_ref_sub(pgcache->page,
+					     pgcache->ref_count - 1);
+			}
+			put_page(pgcache->page);
+		}
+		head++;
+	}
+
+	/* Free RBDR ring */
+	nicvf_free_q_desc_mem(nic, &rbdr->dmem);
+}
+
+/* Refill receive buffer descriptors with new buffers.
+ */
+static void nicvf_refill_rbdr(struct nicvf *nic, gfp_t gfp)
+{
+	struct queue_set *qs = nic->qs;
+	int rbdr_idx = qs->rbdr_cnt;
+	int tail, qcount;
+	int refill_rb_cnt;
+	struct rbdr *rbdr;
+	struct rbdr_entry_t *desc;
+	u64 rbuf;
+	int new_rb = 0;
+
+refill:
+	if (!rbdr_idx)
+		return;
+	rbdr_idx--;
+	rbdr = &qs->rbdr[rbdr_idx];
+	/* Check if it's enabled */
+	if (!rbdr->enable)
+		goto next_rbdr;
+
+	/* Get no of desc's to be refilled */
+	qcount = nicvf_queue_reg_read(nic, NIC_QSET_RBDR_0_1_STATUS0, rbdr_idx);
+	qcount &= 0x7FFFF;
+	/* Doorbell can be ringed with a max of ring size minus 1 */
+	if (qcount >= (qs->rbdr_len - 1))
+		goto next_rbdr;
+	else
+		refill_rb_cnt = qs->rbdr_len - qcount - 1;
+
+	/* Sync page cache info */
+	smp_rmb();
+
+	/* Start filling descs from tail */
+	tail = nicvf_queue_reg_read(nic, NIC_QSET_RBDR_0_1_TAIL, rbdr_idx) >> 3;
+	while (refill_rb_cnt) {
+		tail++;
+		tail &= (rbdr->dmem.q_len - 1);
+
+		if (nicvf_alloc_rcv_buffer(nic, rbdr, gfp, RCV_FRAG_LEN, &rbuf))
+			break;
+
+		desc = GET_RBDR_DESC(rbdr, tail);
+		desc->buf_addr = rbuf & ~(NICVF_RCV_BUF_ALIGN_BYTES - 1);
+		refill_rb_cnt--;
+		new_rb++;
+	}
+
+	nicvf_get_page(nic);
+
+	/* make sure all memory stores are done before ringing doorbell */
+	smp_wmb();
+
+	/* Check if buffer allocation failed */
+	if (refill_rb_cnt)
+		nic->rb_alloc_fail = true;
+	else
+		nic->rb_alloc_fail = false;
+
+	/* Notify HW */
+	nicvf_queue_reg_write(nic, NIC_QSET_RBDR_0_1_DOOR,
+			      rbdr_idx, new_rb);
+next_rbdr:
+	/* Re-enable RBDR interrupts only if buffer allocation is success */
+	if (!nic->rb_alloc_fail && rbdr->enable &&
+	    netif_running(nic->pnicvf->netdev))
+		nicvf_enable_intr(nic, NICVF_INTR_RBDR, rbdr_idx);
+
+	if (rbdr_idx)
+		goto refill;
+}
+
+/* Alloc rcv buffers in non-atomic mode for better success */
+void nicvf_rbdr_work(struct work_struct *work)
+{
+	struct nicvf *nic = container_of(work, struct nicvf, rbdr_work.work);
+
+	nicvf_refill_rbdr(nic, GFP_KERNEL);
+	if (nic->rb_alloc_fail)
+		schedule_delayed_work(&nic->rbdr_work, msecs_to_jiffies(10));
+	else
+		nic->rb_work_scheduled = false;
+}
+
+/* In Softirq context, alloc rcv buffers in atomic mode */
+void nicvf_rbdr_task(struct tasklet_struct *t)
+{
+	struct nicvf *nic = from_tasklet(nic, t, rbdr_task);
+
+	nicvf_refill_rbdr(nic, GFP_ATOMIC);
+	if (nic->rb_alloc_fail) {
+		nic->rb_work_scheduled = true;
+		schedule_delayed_work(&nic->rbdr_work, msecs_to_jiffies(10));
+	}
+}
+
+/* Initialize completion queue */
+static int nicvf_init_cmp_queue(struct nicvf *nic,
+				struct cmp_queue *cq, int q_len)
+{
+	int err;
+
+	err = nicvf_alloc_q_desc_mem(nic, &cq->dmem, q_len, CMP_QUEUE_DESC_SIZE,
+				     NICVF_CQ_BASE_ALIGN_BYTES);
+	if (err)
+		return err;
+
+	cq->desc = cq->dmem.base;
+	cq->thresh = pass1_silicon(nic->pdev) ? 0 : CMP_QUEUE_CQE_THRESH;
+	nic->cq_coalesce_usecs = (CMP_QUEUE_TIMER_THRESH * 0.05) - 1;
+
+	return 0;
+}
+
+static void nicvf_free_cmp_queue(struct nicvf *nic, struct cmp_queue *cq)
+{
+	if (!cq)
+		return;
+	if (!cq->dmem.base)
+		return;
+
+	nicvf_free_q_desc_mem(nic, &cq->dmem);
+}
+
+/* Initialize transmit queue */
+static int nicvf_init_snd_queue(struct nicvf *nic,
+				struct snd_queue *sq, int q_len, int qidx)
+{
+	int err;
+
+	err = nicvf_alloc_q_desc_mem(nic, &sq->dmem, q_len, SND_QUEUE_DESC_SIZE,
+				     NICVF_SQ_BASE_ALIGN_BYTES);
+	if (err)
+		return err;
+
+	sq->desc = sq->dmem.base;
+	sq->skbuff = kcalloc(q_len, sizeof(u64), GFP_KERNEL);
+	if (!sq->skbuff)
+		return -ENOMEM;
+
+	sq->head = 0;
+	sq->tail = 0;
+	sq->thresh = SND_QUEUE_THRESH;
+
+	/* Check if this SQ is a XDP TX queue */
+	if (nic->sqs_mode)
+		qidx += ((nic->sqs_id + 1) * MAX_SND_QUEUES_PER_QS);
+	if (qidx < nic->pnicvf->xdp_tx_queues) {
+		/* Alloc memory to save page pointers for XDP_TX */
+		sq->xdp_page = kcalloc(q_len, sizeof(u64), GFP_KERNEL);
+		if (!sq->xdp_page)
+			return -ENOMEM;
+		sq->xdp_desc_cnt = 0;
+		sq->xdp_free_cnt = q_len - 1;
+		sq->is_xdp = true;
+	} else {
+		sq->xdp_page = NULL;
+		sq->xdp_desc_cnt = 0;
+		sq->xdp_free_cnt = 0;
+		sq->is_xdp = false;
+
+		atomic_set(&sq->free_cnt, q_len - 1);
+
+		/* Preallocate memory for TSO segment's header */
+		sq->tso_hdrs = dma_alloc_coherent(&nic->pdev->dev,
+						  q_len * TSO_HEADER_SIZE,
+						  &sq->tso_hdrs_phys,
+						  GFP_KERNEL);
+		if (!sq->tso_hdrs)
+			return -ENOMEM;
+	}
+
+	return 0;
+}
+
+void nicvf_unmap_sndq_buffers(struct nicvf *nic, struct snd_queue *sq,
+			      int hdr_sqe, u8 subdesc_cnt)
+{
+	u8 idx;
+	struct sq_gather_subdesc *gather;
+
+	/* Unmap DMA mapped skb data buffers */
+	for (idx = 0; idx < subdesc_cnt; idx++) {
+		hdr_sqe++;
+		hdr_sqe &= (sq->dmem.q_len - 1);
+		gather = (struct sq_gather_subdesc *)GET_SQ_DESC(sq, hdr_sqe);
+		/* HW will ensure data coherency, CPU sync not required */
+		dma_unmap_page_attrs(&nic->pdev->dev, gather->addr,
+				     gather->size, DMA_TO_DEVICE,
+				     DMA_ATTR_SKIP_CPU_SYNC);
+	}
+}
+
+static void nicvf_free_snd_queue(struct nicvf *nic, struct snd_queue *sq)
+{
+	struct sk_buff *skb;
+	struct page *page;
+	struct sq_hdr_subdesc *hdr;
+	struct sq_hdr_subdesc *tso_sqe;
+
+	if (!sq)
+		return;
+	if (!sq->dmem.base)
+		return;
+
+	if (sq->tso_hdrs) {
+		dma_free_coherent(&nic->pdev->dev,
+				  sq->dmem.q_len * TSO_HEADER_SIZE,
+				  sq->tso_hdrs, sq->tso_hdrs_phys);
+		sq->tso_hdrs = NULL;
+	}
+
+	/* Free pending skbs in the queue */
+	smp_rmb();
+	while (sq->head != sq->tail) {
+		skb = (struct sk_buff *)sq->skbuff[sq->head];
+		if (!skb || !sq->xdp_page)
+			goto next;
+
+		page = (struct page *)sq->xdp_page[sq->head];
+		if (!page)
+			goto next;
+		else
+			put_page(page);
+
+		hdr = (struct sq_hdr_subdesc *)GET_SQ_DESC(sq, sq->head);
+		/* Check for dummy descriptor used for HW TSO offload on 88xx */
+		if (hdr->dont_send) {
+			/* Get actual TSO descriptors and unmap them */
+			tso_sqe =
+			 (struct sq_hdr_subdesc *)GET_SQ_DESC(sq, hdr->rsvd2);
+			nicvf_unmap_sndq_buffers(nic, sq, hdr->rsvd2,
+						 tso_sqe->subdesc_cnt);
+		} else {
+			nicvf_unmap_sndq_buffers(nic, sq, sq->head,
+						 hdr->subdesc_cnt);
+		}
+		if (skb)
+			dev_kfree_skb_any(skb);
+next:
+		sq->head++;
+		sq->head &= (sq->dmem.q_len - 1);
+	}
+	kfree(sq->skbuff);
+	kfree(sq->xdp_page);
+	nicvf_free_q_desc_mem(nic, &sq->dmem);
+}
+
+static void nicvf_reclaim_snd_queue(struct nicvf *nic,
+				    struct queue_set *qs, int qidx)
+{
+	/* Disable send queue */
+	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_CFG, qidx, 0);
+	/* Check if SQ is stopped */
+	if (nicvf_poll_reg(nic, qidx, NIC_QSET_SQ_0_7_STATUS, 21, 1, 0x01))
+		return;
+	/* Reset send queue */
+	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_CFG, qidx, NICVF_SQ_RESET);
+}
+
+static void nicvf_reclaim_rcv_queue(struct nicvf *nic,
+				    struct queue_set *qs, int qidx)
+{
+	union nic_mbx mbx = {};
+
+	/* Make sure all packets in the pipeline are written back into mem */
+	mbx.msg.msg = NIC_MBOX_MSG_RQ_SW_SYNC;
+	nicvf_send_msg_to_pf(nic, &mbx);
+}
+
+static void nicvf_reclaim_cmp_queue(struct nicvf *nic,
+				    struct queue_set *qs, int qidx)
+{
+	/* Disable timer threshold (doesn't get reset upon CQ reset */
+	nicvf_queue_reg_write(nic, NIC_QSET_CQ_0_7_CFG2, qidx, 0);
+	/* Disable completion queue */
+	nicvf_queue_reg_write(nic, NIC_QSET_CQ_0_7_CFG, qidx, 0);
+	/* Reset completion queue */
+	nicvf_queue_reg_write(nic, NIC_QSET_CQ_0_7_CFG, qidx, NICVF_CQ_RESET);
+}
+
+static void nicvf_reclaim_rbdr(struct nicvf *nic,
+			       struct rbdr *rbdr, int qidx)
+{
+	u64 tmp, fifo_state;
+	int timeout = 10;
+
+	/* Save head and tail pointers for feeing up buffers */
+	rbdr->head = nicvf_queue_reg_read(nic,
+					  NIC_QSET_RBDR_0_1_HEAD,
+					  qidx) >> 3;
+	rbdr->tail = nicvf_queue_reg_read(nic,
+					  NIC_QSET_RBDR_0_1_TAIL,
+					  qidx) >> 3;
+
+	/* If RBDR FIFO is in 'FAIL' state then do a reset first
+	 * before relaiming.
+	 */
+	fifo_state = nicvf_queue_reg_read(nic, NIC_QSET_RBDR_0_1_STATUS0, qidx);
+	if (((fifo_state >> 62) & 0x03) == 0x3)
+		nicvf_queue_reg_write(nic, NIC_QSET_RBDR_0_1_CFG,
+				      qidx, NICVF_RBDR_RESET);
+
+	/* Disable RBDR */
+	nicvf_queue_reg_write(nic, NIC_QSET_RBDR_0_1_CFG, qidx, 0);
+	if (nicvf_poll_reg(nic, qidx, NIC_QSET_RBDR_0_1_STATUS0, 62, 2, 0x00))
+		return;
+	while (1) {
+		tmp = nicvf_queue_reg_read(nic,
+					   NIC_QSET_RBDR_0_1_PREFETCH_STATUS,
+					   qidx);
+		if ((tmp & 0xFFFFFFFF) == ((tmp >> 32) & 0xFFFFFFFF))
+			break;
+		usleep_range(1000, 2000);
+		timeout--;
+		if (!timeout) {
+			netdev_err(nic->netdev,
+				   "Failed polling on prefetch status\n");
+			return;
+		}
+	}
+	nicvf_queue_reg_write(nic, NIC_QSET_RBDR_0_1_CFG,
+			      qidx, NICVF_RBDR_RESET);
+
+	if (nicvf_poll_reg(nic, qidx, NIC_QSET_RBDR_0_1_STATUS0, 62, 2, 0x02))
+		return;
+	nicvf_queue_reg_write(nic, NIC_QSET_RBDR_0_1_CFG, qidx, 0x00);
+	if (nicvf_poll_reg(nic, qidx, NIC_QSET_RBDR_0_1_STATUS0, 62, 2, 0x00))
+		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);
+}
+
+static void nicvf_reset_rcv_queue_stats(struct nicvf *nic)
+{
+	union nic_mbx mbx = {};
+
+	/* Reset all RQ/SQ and VF stats */
+	mbx.reset_stat.msg = NIC_MBOX_MSG_RESET_STAT_COUNTER;
+	mbx.reset_stat.rx_stat_mask = 0x3FFF;
+	mbx.reset_stat.tx_stat_mask = 0x1F;
+	mbx.reset_stat.rq_stat_mask = 0xFFFF;
+	mbx.reset_stat.sq_stat_mask = 0xFFFF;
+	nicvf_send_msg_to_pf(nic, &mbx);
+}
+
+/* Configures receive queue */
+static void nicvf_rcv_queue_config(struct nicvf *nic, struct queue_set *qs,
+				   int qidx, bool enable)
+{
+	union nic_mbx mbx = {};
+	struct rcv_queue *rq;
+	struct rq_cfg rq_cfg;
+
+	rq = &qs->rq[qidx];
+	rq->enable = enable;
+
+	/* Disable receive queue */
+	nicvf_queue_reg_write(nic, NIC_QSET_RQ_0_7_CFG, qidx, 0);
+
+	if (!rq->enable) {
+		nicvf_reclaim_rcv_queue(nic, qs, qidx);
+		xdp_rxq_info_unreg(&rq->xdp_rxq);
+		return;
+	}
+
+	rq->cq_qs = qs->vnic_id;
+	rq->cq_idx = qidx;
+	rq->start_rbdr_qs = qs->vnic_id;
+	rq->start_qs_rbdr_idx = qs->rbdr_cnt - 1;
+	rq->cont_rbdr_qs = qs->vnic_id;
+	rq->cont_qs_rbdr_idx = qs->rbdr_cnt - 1;
+	/* all writes of RBDR data to be loaded into L2 Cache as well*/
+	rq->caching = 1;
+
+	/* Driver have no proper error path for failed XDP RX-queue info reg */
+	WARN_ON(xdp_rxq_info_reg(&rq->xdp_rxq, nic->netdev, qidx, 0) < 0);
+
+	/* Send a mailbox msg to PF to config RQ */
+	mbx.rq.msg = NIC_MBOX_MSG_RQ_CFG;
+	mbx.rq.qs_num = qs->vnic_id;
+	mbx.rq.rq_num = qidx;
+	mbx.rq.cfg = ((u64)rq->caching << 26) | (rq->cq_qs << 19) |
+			  (rq->cq_idx << 16) | (rq->cont_rbdr_qs << 9) |
+			  (rq->cont_qs_rbdr_idx << 8) |
+			  (rq->start_rbdr_qs << 1) | (rq->start_qs_rbdr_idx);
+	nicvf_send_msg_to_pf(nic, &mbx);
+
+	mbx.rq.msg = NIC_MBOX_MSG_RQ_BP_CFG;
+	mbx.rq.cfg = BIT_ULL(63) | BIT_ULL(62) |
+		     (RQ_PASS_RBDR_LVL << 16) | (RQ_PASS_CQ_LVL << 8) |
+		     (qs->vnic_id << 0);
+	nicvf_send_msg_to_pf(nic, &mbx);
+
+	/* RQ drop config
+	 * Enable CQ drop to reserve sufficient CQEs for all tx packets
+	 */
+	mbx.rq.msg = NIC_MBOX_MSG_RQ_DROP_CFG;
+	mbx.rq.cfg = BIT_ULL(63) | BIT_ULL(62) |
+		     (RQ_PASS_RBDR_LVL << 40) | (RQ_DROP_RBDR_LVL << 32) |
+		     (RQ_PASS_CQ_LVL << 16) | (RQ_DROP_CQ_LVL << 8);
+	nicvf_send_msg_to_pf(nic, &mbx);
+
+	if (!nic->sqs_mode && (qidx == 0)) {
+		/* Enable checking L3/L4 length and TCP/UDP checksums
+		 * Also allow IPv6 pkts with zero UDP checksum.
+		 */
+		nicvf_queue_reg_write(nic, NIC_QSET_RQ_GEN_CFG, 0,
+				      (BIT(24) | BIT(23) | BIT(21) | BIT(20)));
+		nicvf_config_vlan_stripping(nic, nic->netdev->features);
+	}
+
+	/* Enable Receive queue */
+	memset(&rq_cfg, 0, sizeof(struct rq_cfg));
+	rq_cfg.ena = 1;
+	rq_cfg.tcp_ena = 0;
+	nicvf_queue_reg_write(nic, NIC_QSET_RQ_0_7_CFG, qidx, *(u64 *)&rq_cfg);
+}
+
+/* Configures completion queue */
+void nicvf_cmp_queue_config(struct nicvf *nic, struct queue_set *qs,
+			    int qidx, bool enable)
+{
+	struct cmp_queue *cq;
+	struct cq_cfg cq_cfg;
+
+	cq = &qs->cq[qidx];
+	cq->enable = enable;
+
+	if (!cq->enable) {
+		nicvf_reclaim_cmp_queue(nic, qs, qidx);
+		return;
+	}
+
+	/* Reset completion queue */
+	nicvf_queue_reg_write(nic, NIC_QSET_CQ_0_7_CFG, qidx, NICVF_CQ_RESET);
+
+	if (!cq->enable)
+		return;
+
+	spin_lock_init(&cq->lock);
+	/* Set completion queue base address */
+	nicvf_queue_reg_write(nic, NIC_QSET_CQ_0_7_BASE,
+			      qidx, (u64)(cq->dmem.phys_base));
+
+	/* Enable Completion queue */
+	memset(&cq_cfg, 0, sizeof(struct cq_cfg));
+	cq_cfg.ena = 1;
+	cq_cfg.reset = 0;
+	cq_cfg.caching = 0;
+	cq_cfg.qsize = ilog2(qs->cq_len >> 10);
+	cq_cfg.avg_con = 0;
+	nicvf_queue_reg_write(nic, NIC_QSET_CQ_0_7_CFG, qidx, *(u64 *)&cq_cfg);
+
+	/* Set threshold value for interrupt generation */
+	nicvf_queue_reg_write(nic, NIC_QSET_CQ_0_7_THRESH, qidx, cq->thresh);
+	nicvf_queue_reg_write(nic, NIC_QSET_CQ_0_7_CFG2,
+			      qidx, CMP_QUEUE_TIMER_THRESH);
+}
+
+/* Configures transmit queue */
+static void nicvf_snd_queue_config(struct nicvf *nic, struct queue_set *qs,
+				   int qidx, bool enable)
+{
+	union nic_mbx mbx = {};
+	struct snd_queue *sq;
+	struct sq_cfg sq_cfg;
+
+	sq = &qs->sq[qidx];
+	sq->enable = enable;
+
+	if (!sq->enable) {
+		nicvf_reclaim_snd_queue(nic, qs, qidx);
+		return;
+	}
+
+	/* Reset send queue */
+	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_CFG, qidx, NICVF_SQ_RESET);
+
+	sq->cq_qs = qs->vnic_id;
+	sq->cq_idx = qidx;
+
+	/* Send a mailbox msg to PF to config SQ */
+	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);
+
+	/* Set queue base address */
+	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_BASE,
+			      qidx, (u64)(sq->dmem.phys_base));
+
+	/* Enable send queue  & set queue size */
+	memset(&sq_cfg, 0, sizeof(struct sq_cfg));
+	sq_cfg.ena = 1;
+	sq_cfg.reset = 0;
+	sq_cfg.ldwb = 0;
+	sq_cfg.qsize = ilog2(qs->sq_len >> 10);
+	sq_cfg.tstmp_bgx_intf = 0;
+	/* CQ's level at which HW will stop processing SQEs to avoid
+	 * transmitting a pkt with no space in CQ to post CQE_TX.
+	 */
+	sq_cfg.cq_limit = (CMP_QUEUE_PIPELINE_RSVD * 256) / qs->cq_len;
+	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_CFG, qidx, *(u64 *)&sq_cfg);
+
+	/* Set threshold value for interrupt generation */
+	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_THRESH, qidx, sq->thresh);
+
+	/* Set queue:cpu affinity for better load distribution */
+	if (cpu_online(qidx)) {
+		cpumask_set_cpu(qidx, &sq->affinity_mask);
+		netif_set_xps_queue(nic->netdev,
+				    &sq->affinity_mask, qidx);
+	}
+}
+
+/* Configures receive buffer descriptor ring */
+static void nicvf_rbdr_config(struct nicvf *nic, struct queue_set *qs,
+			      int qidx, bool enable)
+{
+	struct rbdr *rbdr;
+	struct rbdr_cfg rbdr_cfg;
+
+	rbdr = &qs->rbdr[qidx];
+	nicvf_reclaim_rbdr(nic, rbdr, qidx);
+	if (!enable)
+		return;
+
+	/* Set descriptor base address */
+	nicvf_queue_reg_write(nic, NIC_QSET_RBDR_0_1_BASE,
+			      qidx, (u64)(rbdr->dmem.phys_base));
+
+	/* Enable RBDR  & set queue size */
+	/* Buffer size should be in multiples of 128 bytes */
+	memset(&rbdr_cfg, 0, sizeof(struct rbdr_cfg));
+	rbdr_cfg.ena = 1;
+	rbdr_cfg.reset = 0;
+	rbdr_cfg.ldwb = 0;
+	rbdr_cfg.qsize = RBDR_SIZE;
+	rbdr_cfg.avg_con = 0;
+	rbdr_cfg.lines = rbdr->dma_size / 128;
+	nicvf_queue_reg_write(nic, NIC_QSET_RBDR_0_1_CFG,
+			      qidx, *(u64 *)&rbdr_cfg);
+
+	/* Notify HW */
+	nicvf_queue_reg_write(nic, NIC_QSET_RBDR_0_1_DOOR,
+			      qidx, qs->rbdr_len - 1);
+
+	/* Set threshold value for interrupt generation */
+	nicvf_queue_reg_write(nic, NIC_QSET_RBDR_0_1_THRESH,
+			      qidx, rbdr->thresh - 1);
+}
+
+/* Requests PF to assign and enable Qset */
+void nicvf_qset_config(struct nicvf *nic, bool enable)
+{
+	union nic_mbx mbx = {};
+	struct queue_set *qs = nic->qs;
+	struct qs_cfg *qs_cfg;
+
+	if (!qs) {
+		netdev_warn(nic->netdev,
+			    "Qset is still not allocated, don't init queues\n");
+		return;
+	}
+
+	qs->enable = enable;
+	qs->vnic_id = nic->vf_id;
+
+	/* 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;
+	if (qs->enable) {
+		qs_cfg->ena = 1;
+#ifdef __BIG_ENDIAN
+		qs_cfg->be = 1;
+#endif
+		qs_cfg->vnic = qs->vnic_id;
+		/* Enable Tx timestamping capability */
+		if (nic->ptp_clock)
+			qs_cfg->send_tstmp_ena = 1;
+	}
+	nicvf_send_msg_to_pf(nic, &mbx);
+}
+
+static void nicvf_free_resources(struct nicvf *nic)
+{
+	int qidx;
+	struct queue_set *qs = nic->qs;
+
+	/* Free receive buffer descriptor ring */
+	for (qidx = 0; qidx < qs->rbdr_cnt; qidx++)
+		nicvf_free_rbdr(nic, &qs->rbdr[qidx]);
+
+	/* Free completion queue */
+	for (qidx = 0; qidx < qs->cq_cnt; qidx++)
+		nicvf_free_cmp_queue(nic, &qs->cq[qidx]);
+
+	/* Free send queue */
+	for (qidx = 0; qidx < qs->sq_cnt; qidx++)
+		nicvf_free_snd_queue(nic, &qs->sq[qidx]);
+}
+
+static int nicvf_alloc_resources(struct nicvf *nic)
+{
+	int qidx;
+	struct queue_set *qs = nic->qs;
+
+	/* Alloc receive buffer descriptor ring */
+	for (qidx = 0; qidx < qs->rbdr_cnt; qidx++) {
+		if (nicvf_init_rbdr(nic, &qs->rbdr[qidx], qs->rbdr_len,
+				    DMA_BUFFER_LEN))
+			goto alloc_fail;
+	}
+
+	/* Alloc send queue */
+	for (qidx = 0; qidx < qs->sq_cnt; qidx++) {
+		if (nicvf_init_snd_queue(nic, &qs->sq[qidx], qs->sq_len, qidx))
+			goto alloc_fail;
+	}
+
+	/* Alloc completion queue */
+	for (qidx = 0; qidx < qs->cq_cnt; qidx++) {
+		if (nicvf_init_cmp_queue(nic, &qs->cq[qidx], qs->cq_len))
+			goto alloc_fail;
+	}
+
+	return 0;
+alloc_fail:
+	nicvf_free_resources(nic);
+	return -ENOMEM;
+}
+
+int nicvf_set_qset_resources(struct nicvf *nic)
+{
+	struct queue_set *qs;
+
+	qs = devm_kzalloc(&nic->pdev->dev, sizeof(*qs), GFP_KERNEL);
+	if (!qs)
+		return -ENOMEM;
+	nic->qs = qs;
+
+	/* Set count of each queue */
+	qs->rbdr_cnt = DEFAULT_RBDR_CNT;
+	qs->rq_cnt = min_t(u8, MAX_RCV_QUEUES_PER_QS, num_online_cpus());
+	qs->sq_cnt = min_t(u8, MAX_SND_QUEUES_PER_QS, num_online_cpus());
+	qs->cq_cnt = max_t(u8, qs->rq_cnt, qs->sq_cnt);
+
+	/* Set queue lengths */
+	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;
+	nic->xdp_tx_queues = 0;
+
+	return 0;
+}
+
+int nicvf_config_data_transfer(struct nicvf *nic, bool enable)
+{
+	bool disable = false;
+	struct queue_set *qs = nic->qs;
+	struct queue_set *pqs = nic->pnicvf->qs;
+	int qidx;
+
+	if (!qs)
+		return 0;
+
+	/* Take primary VF's queue lengths.
+	 * This is needed to take queue lengths set from ethtool
+	 * into consideration.
+	 */
+	if (nic->sqs_mode && pqs) {
+		qs->cq_len = pqs->cq_len;
+		qs->sq_len = pqs->sq_len;
+	}
+
+	if (enable) {
+		if (nicvf_alloc_resources(nic))
+			return -ENOMEM;
+
+		for (qidx = 0; qidx < qs->sq_cnt; qidx++)
+			nicvf_snd_queue_config(nic, qs, qidx, enable);
+		for (qidx = 0; qidx < qs->cq_cnt; qidx++)
+			nicvf_cmp_queue_config(nic, qs, qidx, enable);
+		for (qidx = 0; qidx < qs->rbdr_cnt; qidx++)
+			nicvf_rbdr_config(nic, qs, qidx, enable);
+		for (qidx = 0; qidx < qs->rq_cnt; qidx++)
+			nicvf_rcv_queue_config(nic, qs, qidx, enable);
+	} else {
+		for (qidx = 0; qidx < qs->rq_cnt; qidx++)
+			nicvf_rcv_queue_config(nic, qs, qidx, disable);
+		for (qidx = 0; qidx < qs->rbdr_cnt; qidx++)
+			nicvf_rbdr_config(nic, qs, qidx, disable);
+		for (qidx = 0; qidx < qs->sq_cnt; qidx++)
+			nicvf_snd_queue_config(nic, qs, qidx, disable);
+		for (qidx = 0; qidx < qs->cq_cnt; qidx++)
+			nicvf_cmp_queue_config(nic, qs, qidx, disable);
+
+		nicvf_free_resources(nic);
+	}
+
+	/* Reset RXQ's stats.
+	 * SQ's stats will get reset automatically once SQ is reset.
+	 */
+	nicvf_reset_rcv_queue_stats(nic);
+
+	return 0;
+}
+
+/* Get a free desc from SQ
+ * returns descriptor ponter & descriptor number
+ */
+static inline int nicvf_get_sq_desc(struct snd_queue *sq, int desc_cnt)
+{
+	int qentry;
+
+	qentry = sq->tail;
+	if (!sq->is_xdp)
+		atomic_sub(desc_cnt, &sq->free_cnt);
+	else
+		sq->xdp_free_cnt -= desc_cnt;
+	sq->tail += desc_cnt;
+	sq->tail &= (sq->dmem.q_len - 1);
+
+	return qentry;
+}
+
+/* Rollback to previous tail pointer when descriptors not used */
+static inline void nicvf_rollback_sq_desc(struct snd_queue *sq,
+					  int qentry, int desc_cnt)
+{
+	sq->tail = qentry;
+	atomic_add(desc_cnt, &sq->free_cnt);
+}
+
+/* Free descriptor back to SQ for future use */
+void nicvf_put_sq_desc(struct snd_queue *sq, int desc_cnt)
+{
+	if (!sq->is_xdp)
+		atomic_add(desc_cnt, &sq->free_cnt);
+	else
+		sq->xdp_free_cnt += desc_cnt;
+	sq->head += desc_cnt;
+	sq->head &= (sq->dmem.q_len - 1);
+}
+
+static inline int nicvf_get_nxt_sqentry(struct snd_queue *sq, int qentry)
+{
+	qentry++;
+	qentry &= (sq->dmem.q_len - 1);
+	return qentry;
+}
+
+void nicvf_sq_enable(struct nicvf *nic, struct snd_queue *sq, int qidx)
+{
+	u64 sq_cfg;
+
+	sq_cfg = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_CFG, qidx);
+	sq_cfg |= NICVF_SQ_EN;
+	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_CFG, qidx, sq_cfg);
+	/* Ring doorbell so that H/W restarts processing SQEs */
+	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_DOOR, qidx, 0);
+}
+
+void nicvf_sq_disable(struct nicvf *nic, int qidx)
+{
+	u64 sq_cfg;
+
+	sq_cfg = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_CFG, qidx);
+	sq_cfg &= ~NICVF_SQ_EN;
+	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_CFG, qidx, sq_cfg);
+}
+
+void nicvf_sq_free_used_descs(struct net_device *netdev, struct snd_queue *sq,
+			      int qidx)
+{
+	u64 head;
+	struct sk_buff *skb;
+	struct nicvf *nic = netdev_priv(netdev);
+	struct sq_hdr_subdesc *hdr;
+
+	head = nicvf_queue_reg_read(nic, NIC_QSET_SQ_0_7_HEAD, qidx) >> 4;
+	while (sq->head != head) {
+		hdr = (struct sq_hdr_subdesc *)GET_SQ_DESC(sq, sq->head);
+		if (hdr->subdesc_type != SQ_DESC_TYPE_HEADER) {
+			nicvf_put_sq_desc(sq, 1);
+			continue;
+		}
+		skb = (struct sk_buff *)sq->skbuff[sq->head];
+		if (skb)
+			dev_kfree_skb_any(skb);
+		atomic64_add(1, (atomic64_t *)&netdev->stats.tx_packets);
+		atomic64_add(hdr->tot_len,
+			     (atomic64_t *)&netdev->stats.tx_bytes);
+		nicvf_put_sq_desc(sq, hdr->subdesc_cnt + 1);
+	}
+}
+
+/* XDP Transmit APIs */
+void nicvf_xdp_sq_doorbell(struct nicvf *nic,
+			   struct snd_queue *sq, int sq_num)
+{
+	if (!sq->xdp_desc_cnt)
+		return;
+
+	/* make sure all memory stores are done before ringing doorbell */
+	wmb();
+
+	/* Inform HW to xmit all TSO segments */
+	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_DOOR,
+			      sq_num, sq->xdp_desc_cnt);
+	sq->xdp_desc_cnt = 0;
+}
+
+static inline void
+nicvf_xdp_sq_add_hdr_subdesc(struct snd_queue *sq, int qentry,
+			     int subdesc_cnt, u64 data, int len)
+{
+	struct sq_hdr_subdesc *hdr;
+
+	hdr = (struct sq_hdr_subdesc *)GET_SQ_DESC(sq, qentry);
+	memset(hdr, 0, SND_QUEUE_DESC_SIZE);
+	hdr->subdesc_type = SQ_DESC_TYPE_HEADER;
+	hdr->subdesc_cnt = subdesc_cnt;
+	hdr->tot_len = len;
+	hdr->post_cqe = 1;
+	sq->xdp_page[qentry] = (u64)virt_to_page((void *)data);
+}
+
+int nicvf_xdp_sq_append_pkt(struct nicvf *nic, struct snd_queue *sq,
+			    u64 bufaddr, u64 dma_addr, u16 len)
+{
+	int subdesc_cnt = MIN_SQ_DESC_PER_PKT_XMIT;
+	int qentry;
+
+	if (subdesc_cnt > sq->xdp_free_cnt)
+		return 0;
+
+	qentry = nicvf_get_sq_desc(sq, subdesc_cnt);
+
+	nicvf_xdp_sq_add_hdr_subdesc(sq, qentry, subdesc_cnt - 1, bufaddr, len);
+
+	qentry = nicvf_get_nxt_sqentry(sq, qentry);
+	nicvf_sq_add_gather_subdesc(sq, qentry, len, dma_addr);
+
+	sq->xdp_desc_cnt += subdesc_cnt;
+
+	return 1;
+}
+
+/* Calculate no of SQ subdescriptors needed to transmit all
+ * segments of this TSO packet.
+ * Taken from 'Tilera network driver' with a minor modification.
+ */
+static int nicvf_tso_count_subdescs(struct sk_buff *skb)
+{
+	struct skb_shared_info *sh = skb_shinfo(skb);
+	unsigned int sh_len = skb_tcp_all_headers(skb);
+	unsigned int data_len = skb->len - sh_len;
+	unsigned int p_len = sh->gso_size;
+	long f_id = -1;    /* id of the current fragment */
+	long f_size = skb_headlen(skb) - sh_len;  /* current fragment size */
+	long f_used = 0;  /* bytes used from the current fragment */
+	long n;            /* size of the current piece of payload */
+	int num_edescs = 0;
+	int segment;
+
+	for (segment = 0; segment < sh->gso_segs; segment++) {
+		unsigned int p_used = 0;
+
+		/* One edesc for header and for each piece of the payload. */
+		for (num_edescs++; p_used < p_len; num_edescs++) {
+			/* Advance as needed. */
+			while (f_used >= f_size) {
+				f_id++;
+				f_size = skb_frag_size(&sh->frags[f_id]);
+				f_used = 0;
+			}
+
+			/* Use bytes from the current fragment. */
+			n = p_len - p_used;
+			if (n > f_size - f_used)
+				n = f_size - f_used;
+			f_used += n;
+			p_used += n;
+		}
+
+		/* The last segment may be less than gso_size. */
+		data_len -= p_len;
+		if (data_len < p_len)
+			p_len = data_len;
+	}
+
+	/* '+ gso_segs' for SQ_HDR_SUDESCs for each segment */
+	return num_edescs + sh->gso_segs;
+}
+
+#define POST_CQE_DESC_COUNT 2
+
+/* Get the number of SQ descriptors needed to xmit this skb */
+static int nicvf_sq_subdesc_required(struct nicvf *nic, struct sk_buff *skb)
+{
+	int subdesc_cnt = MIN_SQ_DESC_PER_PKT_XMIT;
+
+	if (skb_shinfo(skb)->gso_size && !nic->hw_tso) {
+		subdesc_cnt = nicvf_tso_count_subdescs(skb);
+		return subdesc_cnt;
+	}
+
+	/* Dummy descriptors to get TSO pkt completion notification */
+	if (nic->t88 && nic->hw_tso && skb_shinfo(skb)->gso_size)
+		subdesc_cnt += POST_CQE_DESC_COUNT;
+
+	if (skb_shinfo(skb)->nr_frags)
+		subdesc_cnt += skb_shinfo(skb)->nr_frags;
+
+	return subdesc_cnt;
+}
+
+/* Add SQ HEADER subdescriptor.
+ * First subdescriptor for every send descriptor.
+ */
+static inline void
+nicvf_sq_add_hdr_subdesc(struct nicvf *nic, struct snd_queue *sq, int qentry,
+			 int subdesc_cnt, struct sk_buff *skb, int len)
+{
+	int proto;
+	struct sq_hdr_subdesc *hdr;
+	union {
+		struct iphdr *v4;
+		struct ipv6hdr *v6;
+		unsigned char *hdr;
+	} ip;
+
+	ip.hdr = skb_network_header(skb);
+	hdr = (struct sq_hdr_subdesc *)GET_SQ_DESC(sq, qentry);
+	memset(hdr, 0, SND_QUEUE_DESC_SIZE);
+	hdr->subdesc_type = SQ_DESC_TYPE_HEADER;
+
+	if (nic->t88 && nic->hw_tso && skb_shinfo(skb)->gso_size) {
+		/* post_cqe = 0, to avoid HW posting a CQE for every TSO
+		 * segment transmitted on 88xx.
+		 */
+		hdr->subdesc_cnt = subdesc_cnt - POST_CQE_DESC_COUNT;
+	} else {
+		sq->skbuff[qentry] = (u64)skb;
+		/* Enable notification via CQE after processing SQE */
+		hdr->post_cqe = 1;
+		/* No of subdescriptors following this */
+		hdr->subdesc_cnt = subdesc_cnt;
+	}
+	hdr->tot_len = len;
+
+	/* Offload checksum calculation to HW */
+	if (skb->ip_summed == CHECKSUM_PARTIAL) {
+		if (ip.v4->version == 4)
+			hdr->csum_l3 = 1; /* Enable IP csum calculation */
+		hdr->l3_offset = skb_network_offset(skb);
+		hdr->l4_offset = skb_transport_offset(skb);
+
+		proto = (ip.v4->version == 4) ? ip.v4->protocol :
+			ip.v6->nexthdr;
+
+		switch (proto) {
+		case IPPROTO_TCP:
+			hdr->csum_l4 = SEND_L4_CSUM_TCP;
+			break;
+		case IPPROTO_UDP:
+			hdr->csum_l4 = SEND_L4_CSUM_UDP;
+			break;
+		case IPPROTO_SCTP:
+			hdr->csum_l4 = SEND_L4_CSUM_SCTP;
+			break;
+		}
+	}
+
+	if (nic->hw_tso && skb_shinfo(skb)->gso_size) {
+		hdr->tso = 1;
+		hdr->tso_start = skb_tcp_all_headers(skb);
+		hdr->tso_max_paysize = skb_shinfo(skb)->gso_size;
+		/* For non-tunneled pkts, point this to L2 ethertype */
+		hdr->inner_l3_offset = skb_network_offset(skb) - 2;
+		this_cpu_inc(nic->pnicvf->drv_stats->tx_tso);
+	}
+
+	/* Check if timestamp is requested */
+	if (!(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)) {
+		skb_tx_timestamp(skb);
+		return;
+	}
+
+	/* Tx timestamping not supported along with TSO, so ignore request */
+	if (skb_shinfo(skb)->gso_size)
+		return;
+
+	/* HW supports only a single outstanding packet to timestamp */
+	if (!atomic_add_unless(&nic->pnicvf->tx_ptp_skbs, 1, 1))
+		return;
+
+	/* Mark the SKB for later reference */
+	skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
+
+	/* Finally enable timestamp generation
+	 * Since 'post_cqe' is also set, two CQEs will be posted
+	 * for this packet i.e CQE_TYPE_SEND and CQE_TYPE_SEND_PTP.
+	 */
+	hdr->tstmp = 1;
+}
+
+/* SQ GATHER subdescriptor
+ * Must follow HDR descriptor
+ */
+static inline void nicvf_sq_add_gather_subdesc(struct snd_queue *sq, int qentry,
+					       int size, u64 data)
+{
+	struct sq_gather_subdesc *gather;
+
+	qentry &= (sq->dmem.q_len - 1);
+	gather = (struct sq_gather_subdesc *)GET_SQ_DESC(sq, qentry);
+
+	memset(gather, 0, SND_QUEUE_DESC_SIZE);
+	gather->subdesc_type = SQ_DESC_TYPE_GATHER;
+	gather->ld_type = NIC_SEND_LD_TYPE_E_LDD;
+	gather->size = size;
+	gather->addr = data;
+}
+
+/* Add HDR + IMMEDIATE subdescriptors right after descriptors of a TSO
+ * packet so that a CQE is posted as a notifation for transmission of
+ * TSO packet.
+ */
+static inline void nicvf_sq_add_cqe_subdesc(struct snd_queue *sq, int qentry,
+					    int tso_sqe, struct sk_buff *skb)
+{
+	struct sq_imm_subdesc *imm;
+	struct sq_hdr_subdesc *hdr;
+
+	sq->skbuff[qentry] = (u64)skb;
+
+	hdr = (struct sq_hdr_subdesc *)GET_SQ_DESC(sq, qentry);
+	memset(hdr, 0, SND_QUEUE_DESC_SIZE);
+	hdr->subdesc_type = SQ_DESC_TYPE_HEADER;
+	/* Enable notification via CQE after processing SQE */
+	hdr->post_cqe = 1;
+	/* There is no packet to transmit here */
+	hdr->dont_send = 1;
+	hdr->subdesc_cnt = POST_CQE_DESC_COUNT - 1;
+	hdr->tot_len = 1;
+	/* Actual TSO header SQE index, needed for cleanup */
+	hdr->rsvd2 = tso_sqe;
+
+	qentry = nicvf_get_nxt_sqentry(sq, qentry);
+	imm = (struct sq_imm_subdesc *)GET_SQ_DESC(sq, qentry);
+	memset(imm, 0, SND_QUEUE_DESC_SIZE);
+	imm->subdesc_type = SQ_DESC_TYPE_IMMEDIATE;
+	imm->len = 1;
+}
+
+static inline void nicvf_sq_doorbell(struct nicvf *nic, struct sk_buff *skb,
+				     int sq_num, int desc_cnt)
+{
+	struct netdev_queue *txq;
+
+	txq = netdev_get_tx_queue(nic->pnicvf->netdev,
+				  skb_get_queue_mapping(skb));
+
+	netdev_tx_sent_queue(txq, skb->len);
+
+	/* make sure all memory stores are done before ringing doorbell */
+	smp_wmb();
+
+	/* Inform HW to xmit all TSO segments */
+	nicvf_queue_reg_write(nic, NIC_QSET_SQ_0_7_DOOR,
+			      sq_num, desc_cnt);
+}
+
+/* Segment a TSO packet into 'gso_size' segments and append
+ * them to SQ for transfer
+ */
+static int nicvf_sq_append_tso(struct nicvf *nic, struct snd_queue *sq,
+			       int sq_num, int qentry, struct sk_buff *skb)
+{
+	struct tso_t tso;
+	int seg_subdescs = 0, desc_cnt = 0;
+	int seg_len, total_len, data_left;
+	int hdr_qentry = qentry;
+	int hdr_len;
+
+	hdr_len = tso_start(skb, &tso);
+
+	total_len = skb->len - hdr_len;
+	while (total_len > 0) {
+		char *hdr;
+
+		/* Save Qentry for adding HDR_SUBDESC at the end */
+		hdr_qentry = qentry;
+
+		data_left = min_t(int, skb_shinfo(skb)->gso_size, total_len);
+		total_len -= data_left;
+
+		/* Add segment's header */
+		qentry = nicvf_get_nxt_sqentry(sq, qentry);
+		hdr = sq->tso_hdrs + qentry * TSO_HEADER_SIZE;
+		tso_build_hdr(skb, hdr, &tso, data_left, total_len == 0);
+		nicvf_sq_add_gather_subdesc(sq, qentry, hdr_len,
+					    sq->tso_hdrs_phys +
+					    qentry * TSO_HEADER_SIZE);
+		/* HDR_SUDESC + GATHER */
+		seg_subdescs = 2;
+		seg_len = hdr_len;
+
+		/* Add segment's payload fragments */
+		while (data_left > 0) {
+			int size;
+
+			size = min_t(int, tso.size, data_left);
+
+			qentry = nicvf_get_nxt_sqentry(sq, qentry);
+			nicvf_sq_add_gather_subdesc(sq, qentry, size,
+						    virt_to_phys(tso.data));
+			seg_subdescs++;
+			seg_len += size;
+
+			data_left -= size;
+			tso_build_data(skb, &tso, size);
+		}
+		nicvf_sq_add_hdr_subdesc(nic, sq, hdr_qentry,
+					 seg_subdescs - 1, skb, seg_len);
+		sq->skbuff[hdr_qentry] = (u64)NULL;
+		qentry = nicvf_get_nxt_sqentry(sq, qentry);
+
+		desc_cnt += seg_subdescs;
+	}
+	/* Save SKB in the last segment for freeing */
+	sq->skbuff[hdr_qentry] = (u64)skb;
+
+	nicvf_sq_doorbell(nic, skb, sq_num, desc_cnt);
+
+	this_cpu_inc(nic->pnicvf->drv_stats->tx_tso);
+	return 1;
+}
+
+/* Append an skb to a SQ for packet transfer. */
+int nicvf_sq_append_skb(struct nicvf *nic, struct snd_queue *sq,
+			struct sk_buff *skb, u8 sq_num)
+{
+	int i, size;
+	int subdesc_cnt, hdr_sqe = 0;
+	int qentry;
+	u64 dma_addr;
+
+	subdesc_cnt = nicvf_sq_subdesc_required(nic, skb);
+	if (subdesc_cnt > atomic_read(&sq->free_cnt))
+		goto append_fail;
+
+	qentry = nicvf_get_sq_desc(sq, subdesc_cnt);
+
+	/* Check if its a TSO packet */
+	if (skb_shinfo(skb)->gso_size && !nic->hw_tso)
+		return nicvf_sq_append_tso(nic, sq, sq_num, qentry, skb);
+
+	/* Add SQ header subdesc */
+	nicvf_sq_add_hdr_subdesc(nic, sq, qentry, subdesc_cnt - 1,
+				 skb, skb->len);
+	hdr_sqe = qentry;
+
+	/* Add SQ gather subdescs */
+	qentry = nicvf_get_nxt_sqentry(sq, qentry);
+	size = skb_is_nonlinear(skb) ? skb_headlen(skb) : skb->len;
+	/* HW will ensure data coherency, CPU sync not required */
+	dma_addr = dma_map_page_attrs(&nic->pdev->dev, virt_to_page(skb->data),
+				      offset_in_page(skb->data), size,
+				      DMA_TO_DEVICE, DMA_ATTR_SKIP_CPU_SYNC);
+	if (dma_mapping_error(&nic->pdev->dev, dma_addr)) {
+		nicvf_rollback_sq_desc(sq, qentry, subdesc_cnt);
+		return 0;
+	}
+
+	nicvf_sq_add_gather_subdesc(sq, qentry, size, dma_addr);
+
+	/* Check for scattered buffer */
+	if (!skb_is_nonlinear(skb))
+		goto doorbell;
+
+	for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
+		const skb_frag_t *frag = &skb_shinfo(skb)->frags[i];
+
+		qentry = nicvf_get_nxt_sqentry(sq, qentry);
+		size = skb_frag_size(frag);
+		dma_addr = dma_map_page_attrs(&nic->pdev->dev,
+					      skb_frag_page(frag),
+					      skb_frag_off(frag), size,
+					      DMA_TO_DEVICE,
+					      DMA_ATTR_SKIP_CPU_SYNC);
+		if (dma_mapping_error(&nic->pdev->dev, dma_addr)) {
+			/* Free entire chain of mapped buffers
+			 * here 'i' = frags mapped + above mapped skb->data
+			 */
+			nicvf_unmap_sndq_buffers(nic, sq, hdr_sqe, i);
+			nicvf_rollback_sq_desc(sq, qentry, subdesc_cnt);
+			return 0;
+		}
+		nicvf_sq_add_gather_subdesc(sq, qentry, size, dma_addr);
+	}
+
+doorbell:
+	if (nic->t88 && skb_shinfo(skb)->gso_size) {
+		qentry = nicvf_get_nxt_sqentry(sq, qentry);
+		nicvf_sq_add_cqe_subdesc(sq, qentry, hdr_sqe, skb);
+	}
+
+	nicvf_sq_doorbell(nic, skb, sq_num, subdesc_cnt);
+
+	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;
+}
+
+static inline unsigned frag_num(unsigned i)
+{
+#ifdef __BIG_ENDIAN
+	return (i & ~3) + 3 - (i & 3);
+#else
+	return i;
+#endif
+}
+
+static void nicvf_unmap_rcv_buffer(struct nicvf *nic, u64 dma_addr,
+				   u64 buf_addr, bool xdp)
+{
+	struct page *page = NULL;
+	int len = RCV_FRAG_LEN;
+
+	if (xdp) {
+		page = virt_to_page(phys_to_virt(buf_addr));
+		/* Check if it's a recycled page, if not
+		 * unmap the DMA mapping.
+		 *
+		 * Recycled page holds an extra reference.
+		 */
+		if (page_ref_count(page) != 1)
+			return;
+
+		len += XDP_PACKET_HEADROOM;
+		/* Receive buffers in XDP mode are mapped from page start */
+		dma_addr &= PAGE_MASK;
+	}
+	dma_unmap_page_attrs(&nic->pdev->dev, dma_addr, len,
+			     DMA_FROM_DEVICE, DMA_ATTR_SKIP_CPU_SYNC);
+}
+
+/* Returns SKB for a received packet */
+struct sk_buff *nicvf_get_rcv_skb(struct nicvf *nic,
+				  struct cqe_rx_t *cqe_rx, bool xdp)
+{
+	int frag;
+	int payload_len = 0;
+	struct sk_buff *skb = NULL;
+	struct page *page;
+	int offset;
+	u16 *rb_lens = NULL;
+	u64 *rb_ptrs = NULL;
+	u64 phys_addr;
+
+	rb_lens = (void *)cqe_rx + (3 * sizeof(u64));
+	/* Except 88xx pass1 on all other chips CQE_RX2_S is added to
+	 * CQE_RX at word6, hence buffer pointers move by word
+	 *
+	 * Use existing 'hw_tso' flag which will be set for all chips
+	 * except 88xx pass1 instead of a additional cache line
+	 * access (or miss) by using pci dev's revision.
+	 */
+	if (!nic->hw_tso)
+		rb_ptrs = (void *)cqe_rx + (6 * sizeof(u64));
+	else
+		rb_ptrs = (void *)cqe_rx + (7 * sizeof(u64));
+
+	for (frag = 0; frag < cqe_rx->rb_cnt; frag++) {
+		payload_len = rb_lens[frag_num(frag)];
+		phys_addr = nicvf_iova_to_phys(nic, *rb_ptrs);
+		if (!phys_addr) {
+			if (skb)
+				dev_kfree_skb_any(skb);
+			return NULL;
+		}
+
+		if (!frag) {
+			/* First fragment */
+			nicvf_unmap_rcv_buffer(nic,
+					       *rb_ptrs - cqe_rx->align_pad,
+					       phys_addr, xdp);
+			skb = nicvf_rb_ptr_to_skb(nic,
+						  phys_addr - cqe_rx->align_pad,
+						  payload_len);
+			if (!skb)
+				return NULL;
+			skb_reserve(skb, cqe_rx->align_pad);
+			skb_put(skb, payload_len);
+		} else {
+			/* Add fragments */
+			nicvf_unmap_rcv_buffer(nic, *rb_ptrs, phys_addr, xdp);
+			page = virt_to_page(phys_to_virt(phys_addr));
+			offset = phys_to_virt(phys_addr) - page_address(page);
+			skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, page,
+					offset, payload_len, RCV_FRAG_LEN);
+		}
+		/* Next buffer pointer */
+		rb_ptrs++;
+	}
+	return skb;
+}
+
+static u64 nicvf_int_type_to_mask(int int_type, int q_idx)
+{
+	u64 reg_val;
+
+	switch (int_type) {
+	case NICVF_INTR_CQ:
+		reg_val = ((1ULL << q_idx) << NICVF_INTR_CQ_SHIFT);
+		break;
+	case NICVF_INTR_SQ:
+		reg_val = ((1ULL << q_idx) << NICVF_INTR_SQ_SHIFT);
+		break;
+	case NICVF_INTR_RBDR:
+		reg_val = ((1ULL << q_idx) << NICVF_INTR_RBDR_SHIFT);
+		break;
+	case NICVF_INTR_PKT_DROP:
+		reg_val = (1ULL << NICVF_INTR_PKT_DROP_SHIFT);
+		break;
+	case NICVF_INTR_TCP_TIMER:
+		reg_val = (1ULL << NICVF_INTR_TCP_TIMER_SHIFT);
+		break;
+	case NICVF_INTR_MBOX:
+		reg_val = (1ULL << NICVF_INTR_MBOX_SHIFT);
+		break;
+	case NICVF_INTR_QS_ERR:
+		reg_val = (1ULL << NICVF_INTR_QS_ERR_SHIFT);
+		break;
+	default:
+		reg_val = 0;
+	}
+
+	return reg_val;
+}
+
+/* Enable interrupt */
+void nicvf_enable_intr(struct nicvf *nic, int int_type, int q_idx)
+{
+	u64 mask = nicvf_int_type_to_mask(int_type, q_idx);
+
+	if (!mask) {
+		netdev_dbg(nic->netdev,
+			   "Failed to enable interrupt: unknown type\n");
+		return;
+	}
+	nicvf_reg_write(nic, NIC_VF_ENA_W1S,
+			nicvf_reg_read(nic, NIC_VF_ENA_W1S) | mask);
+}
+
+/* Disable interrupt */
+void nicvf_disable_intr(struct nicvf *nic, int int_type, int q_idx)
+{
+	u64 mask = nicvf_int_type_to_mask(int_type, q_idx);
+
+	if (!mask) {
+		netdev_dbg(nic->netdev,
+			   "Failed to disable interrupt: unknown type\n");
+		return;
+	}
+
+	nicvf_reg_write(nic, NIC_VF_ENA_W1C, mask);
+}
+
+/* Clear interrupt */
+void nicvf_clear_intr(struct nicvf *nic, int int_type, int q_idx)
+{
+	u64 mask = nicvf_int_type_to_mask(int_type, q_idx);
+
+	if (!mask) {
+		netdev_dbg(nic->netdev,
+			   "Failed to clear interrupt: unknown type\n");
+		return;
+	}
+
+	nicvf_reg_write(nic, NIC_VF_INT, mask);
+}
+
+/* Check if interrupt is enabled */
+int nicvf_is_intr_enabled(struct nicvf *nic, int int_type, int q_idx)
+{
+	u64 mask = nicvf_int_type_to_mask(int_type, q_idx);
+	/* If interrupt type is unknown, we treat it disabled. */
+	if (!mask) {
+		netdev_dbg(nic->netdev,
+			   "Failed to check interrupt enable: unknown type\n");
+		return 0;
+	}
+
+	return mask & nicvf_reg_read(nic, NIC_VF_ENA_W1S);
+}
+
+void nicvf_update_rq_stats(struct nicvf *nic, int rq_idx)
+{
+	struct rcv_queue *rq;
+
+#define GET_RQ_STATS(reg) \
+	nicvf_reg_read(nic, NIC_QSET_RQ_0_7_STAT_0_1 |\
+			    (rq_idx << NIC_Q_NUM_SHIFT) | (reg << 3))
+
+	rq = &nic->qs->rq[rq_idx];
+	rq->stats.bytes = GET_RQ_STATS(RQ_SQ_STATS_OCTS);
+	rq->stats.pkts = GET_RQ_STATS(RQ_SQ_STATS_PKTS);
+}
+
+void nicvf_update_sq_stats(struct nicvf *nic, int sq_idx)
+{
+	struct snd_queue *sq;
+
+#define GET_SQ_STATS(reg) \
+	nicvf_reg_read(nic, NIC_QSET_SQ_0_7_STAT_0_1 |\
+			    (sq_idx << NIC_Q_NUM_SHIFT) | (reg << 3))
+
+	sq = &nic->qs->sq[sq_idx];
+	sq->stats.bytes = GET_SQ_STATS(RQ_SQ_STATS_OCTS);
+	sq->stats.pkts = GET_SQ_STATS(RQ_SQ_STATS_PKTS);
+}
+
+/* Check for errors in the receive cmp.queue entry */
+int nicvf_check_cqe_rx_errs(struct nicvf *nic, struct cqe_rx_t *cqe_rx)
+{
+	netif_err(nic, rx_err, nic->netdev,
+		  "RX error CQE err_level 0x%x err_opcode 0x%x\n",
+		  cqe_rx->err_level, cqe_rx->err_opcode);
+
+	switch (cqe_rx->err_opcode) {
+	case CQ_RX_ERROP_RE_PARTIAL:
+		this_cpu_inc(nic->drv_stats->rx_bgx_truncated_pkts);
+		break;
+	case CQ_RX_ERROP_RE_JABBER:
+		this_cpu_inc(nic->drv_stats->rx_jabber_errs);
+		break;
+	case CQ_RX_ERROP_RE_FCS:
+		this_cpu_inc(nic->drv_stats->rx_fcs_errs);
+		break;
+	case CQ_RX_ERROP_RE_RX_CTL:
+		this_cpu_inc(nic->drv_stats->rx_bgx_errs);
+		break;
+	case CQ_RX_ERROP_PREL2_ERR:
+		this_cpu_inc(nic->drv_stats->rx_prel2_errs);
+		break;
+	case CQ_RX_ERROP_L2_MAL:
+		this_cpu_inc(nic->drv_stats->rx_l2_hdr_malformed);
+		break;
+	case CQ_RX_ERROP_L2_OVERSIZE:
+		this_cpu_inc(nic->drv_stats->rx_oversize);
+		break;
+	case CQ_RX_ERROP_L2_UNDERSIZE:
+		this_cpu_inc(nic->drv_stats->rx_undersize);
+		break;
+	case CQ_RX_ERROP_L2_LENMISM:
+		this_cpu_inc(nic->drv_stats->rx_l2_len_mismatch);
+		break;
+	case CQ_RX_ERROP_L2_PCLP:
+		this_cpu_inc(nic->drv_stats->rx_l2_pclp);
+		break;
+	case CQ_RX_ERROP_IP_NOT:
+		this_cpu_inc(nic->drv_stats->rx_ip_ver_errs);
+		break;
+	case CQ_RX_ERROP_IP_CSUM_ERR:
+		this_cpu_inc(nic->drv_stats->rx_ip_csum_errs);
+		break;
+	case CQ_RX_ERROP_IP_MAL:
+		this_cpu_inc(nic->drv_stats->rx_ip_hdr_malformed);
+		break;
+	case CQ_RX_ERROP_IP_MALD:
+		this_cpu_inc(nic->drv_stats->rx_ip_payload_malformed);
+		break;
+	case CQ_RX_ERROP_IP_HOP:
+		this_cpu_inc(nic->drv_stats->rx_ip_ttl_errs);
+		break;
+	case CQ_RX_ERROP_L3_PCLP:
+		this_cpu_inc(nic->drv_stats->rx_l3_pclp);
+		break;
+	case CQ_RX_ERROP_L4_MAL:
+		this_cpu_inc(nic->drv_stats->rx_l4_malformed);
+		break;
+	case CQ_RX_ERROP_L4_CHK:
+		this_cpu_inc(nic->drv_stats->rx_l4_csum_errs);
+		break;
+	case CQ_RX_ERROP_UDP_LEN:
+		this_cpu_inc(nic->drv_stats->rx_udp_len_errs);
+		break;
+	case CQ_RX_ERROP_L4_PORT:
+		this_cpu_inc(nic->drv_stats->rx_l4_port_errs);
+		break;
+	case CQ_RX_ERROP_TCP_FLAG:
+		this_cpu_inc(nic->drv_stats->rx_tcp_flag_errs);
+		break;
+	case CQ_RX_ERROP_TCP_OFFSET:
+		this_cpu_inc(nic->drv_stats->rx_tcp_offset_errs);
+		break;
+	case CQ_RX_ERROP_L4_PCLP:
+		this_cpu_inc(nic->drv_stats->rx_l4_pclp);
+		break;
+	case CQ_RX_ERROP_RBDR_TRUNC:
+		this_cpu_inc(nic->drv_stats->rx_truncated_pkts);
+		break;
+	}
+
+	return 1;
+}
+
+/* Check for errors in the send cmp.queue entry */
+int nicvf_check_cqe_tx_errs(struct nicvf *nic, struct cqe_send_t *cqe_tx)
+{
+	switch (cqe_tx->send_status) {
+	case CQ_TX_ERROP_DESC_FAULT:
+		this_cpu_inc(nic->drv_stats->tx_desc_fault);
+		break;
+	case CQ_TX_ERROP_HDR_CONS_ERR:
+		this_cpu_inc(nic->drv_stats->tx_hdr_cons_err);
+		break;
+	case CQ_TX_ERROP_SUBDC_ERR:
+		this_cpu_inc(nic->drv_stats->tx_subdesc_err);
+		break;
+	case CQ_TX_ERROP_MAX_SIZE_VIOL:
+		this_cpu_inc(nic->drv_stats->tx_max_size_exceeded);
+		break;
+	case CQ_TX_ERROP_IMM_SIZE_OFLOW:
+		this_cpu_inc(nic->drv_stats->tx_imm_size_oflow);
+		break;
+	case CQ_TX_ERROP_DATA_SEQUENCE_ERR:
+		this_cpu_inc(nic->drv_stats->tx_data_seq_err);
+		break;
+	case CQ_TX_ERROP_MEM_SEQUENCE_ERR:
+		this_cpu_inc(nic->drv_stats->tx_mem_seq_err);
+		break;
+	case CQ_TX_ERROP_LOCK_VIOL:
+		this_cpu_inc(nic->drv_stats->tx_lock_viol);
+		break;
+	case CQ_TX_ERROP_DATA_FAULT:
+		this_cpu_inc(nic->drv_stats->tx_data_fault);
+		break;
+	case CQ_TX_ERROP_TSTMP_CONFLICT:
+		this_cpu_inc(nic->drv_stats->tx_tstmp_conflict);
+		break;
+	case CQ_TX_ERROP_TSTMP_TIMEOUT:
+		this_cpu_inc(nic->drv_stats->tx_tstmp_timeout);
+		break;
+	case CQ_TX_ERROP_MEM_FAULT:
+		this_cpu_inc(nic->drv_stats->tx_mem_fault);
+		break;
+	case CQ_TX_ERROP_CK_OVERLAP:
+		this_cpu_inc(nic->drv_stats->tx_csum_overlap);
+		break;
+	case CQ_TX_ERROP_CK_OFLOW:
+		this_cpu_inc(nic->drv_stats->tx_csum_overflow);
+		break;
+	}
+
+	return 1;
+}
diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_queues.h b/drivers/net/ethernet/cavium/thunder/nicvf_queues.h
new file mode 100644
index 000000000..8453defc2
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/nicvf_queues.h
@@ -0,0 +1,374 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (C) 2015 Cavium, Inc.
+ */
+
+#ifndef NICVF_QUEUES_H
+#define NICVF_QUEUES_H
+
+#include <linux/netdevice.h>
+#include <linux/iommu.h>
+#include <net/xdp.h>
+#include "q_struct.h"
+
+#define MAX_QUEUE_SET			128
+#define MAX_RCV_QUEUES_PER_QS		8
+#define MAX_RCV_BUF_DESC_RINGS_PER_QS	2
+#define MAX_SND_QUEUES_PER_QS		8
+#define MAX_CMP_QUEUES_PER_QS		8
+
+/* VF's queue interrupt ranges */
+#define	NICVF_INTR_ID_CQ		0
+#define	NICVF_INTR_ID_SQ		8
+#define	NICVF_INTR_ID_RBDR		16
+#define	NICVF_INTR_ID_MISC		18
+#define	NICVF_INTR_ID_QS_ERR		19
+
+#define	for_each_cq_irq(irq)	\
+	for (irq = NICVF_INTR_ID_CQ; irq < NICVF_INTR_ID_SQ; irq++)
+#define	for_each_sq_irq(irq)	\
+	for (irq = NICVF_INTR_ID_SQ; irq < NICVF_INTR_ID_RBDR; irq++)
+#define	for_each_rbdr_irq(irq)	\
+	for (irq = NICVF_INTR_ID_RBDR; irq < NICVF_INTR_ID_MISC; irq++)
+
+#define RBDR_SIZE0		0ULL /* 8K entries */
+#define RBDR_SIZE1		1ULL /* 16K entries */
+#define RBDR_SIZE2		2ULL /* 32K entries */
+#define RBDR_SIZE3		3ULL /* 64K entries */
+#define RBDR_SIZE4		4ULL /* 126K entries */
+#define RBDR_SIZE5		5ULL /* 256K entries */
+#define RBDR_SIZE6		6ULL /* 512K entries */
+
+#define SND_QUEUE_SIZE0		0ULL /* 1K entries */
+#define SND_QUEUE_SIZE1		1ULL /* 2K entries */
+#define SND_QUEUE_SIZE2		2ULL /* 4K entries */
+#define SND_QUEUE_SIZE3		3ULL /* 8K entries */
+#define SND_QUEUE_SIZE4		4ULL /* 16K entries */
+#define SND_QUEUE_SIZE5		5ULL /* 32K entries */
+#define SND_QUEUE_SIZE6		6ULL /* 64K entries */
+
+#define CMP_QUEUE_SIZE0		0ULL /* 1K entries */
+#define CMP_QUEUE_SIZE1		1ULL /* 2K entries */
+#define CMP_QUEUE_SIZE2		2ULL /* 4K entries */
+#define CMP_QUEUE_SIZE3		3ULL /* 8K entries */
+#define CMP_QUEUE_SIZE4		4ULL /* 16K entries */
+#define CMP_QUEUE_SIZE5		5ULL /* 32K entries */
+#define CMP_QUEUE_SIZE6		6ULL /* 64K entries */
+
+/* Default queue count per QS, its lengths and threshold values */
+#define DEFAULT_RBDR_CNT	1
+
+#define SND_QSIZE		SND_QUEUE_SIZE0
+#define SND_QUEUE_LEN		(1ULL << (SND_QSIZE + 10))
+#define MIN_SND_QUEUE_LEN	(1ULL << (SND_QUEUE_SIZE0 + 10))
+#define MAX_SND_QUEUE_LEN	(1ULL << (SND_QUEUE_SIZE6 + 10))
+#define SND_QUEUE_THRESH	2ULL
+#define MIN_SQ_DESC_PER_PKT_XMIT	2
+/* Since timestamp not enabled, otherwise 2 */
+#define MAX_CQE_PER_PKT_XMIT		1
+
+/* Keep CQ and SQ sizes same, if timestamping
+ * is enabled this equation will change.
+ */
+#define CMP_QSIZE		CMP_QUEUE_SIZE0
+#define CMP_QUEUE_LEN		(1ULL << (CMP_QSIZE + 10))
+#define MIN_CMP_QUEUE_LEN	(1ULL << (CMP_QUEUE_SIZE0 + 10))
+#define MAX_CMP_QUEUE_LEN	(1ULL << (CMP_QUEUE_SIZE6 + 10))
+#define CMP_QUEUE_CQE_THRESH	(NAPI_POLL_WEIGHT / 2)
+#define CMP_QUEUE_TIMER_THRESH	80 /* ~2usec */
+
+/* No of CQEs that might anyway gets used by HW due to pipelining
+ * effects irrespective of PASS/DROP/LEVELS being configured
+ */
+#define CMP_QUEUE_PIPELINE_RSVD 544
+
+#define RBDR_SIZE		RBDR_SIZE0
+#define RCV_BUF_COUNT		(1ULL << (RBDR_SIZE + 13))
+#define MAX_RCV_BUF_COUNT	(1ULL << (RBDR_SIZE6 + 13))
+#define RBDR_THRESH		(RCV_BUF_COUNT / 2)
+#define DMA_BUFFER_LEN		1536 /* In multiples of 128bytes */
+#define RCV_FRAG_LEN	 (SKB_DATA_ALIGN(DMA_BUFFER_LEN + NET_SKB_PAD) + \
+			 SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))
+
+#define MAX_CQES_FOR_TX		((SND_QUEUE_LEN / MIN_SQ_DESC_PER_PKT_XMIT) * \
+				 MAX_CQE_PER_PKT_XMIT)
+
+/* RED and Backpressure levels of CQ for pkt reception
+ * For CQ, level is a measure of emptiness i.e 0x0 means full
+ * eg: For CQ of size 4K, and for pass/drop levels of 160/144
+ * HW accepts pkt if unused CQE >= 2560
+ * RED accepts pkt if unused CQE < 2304 & >= 2560
+ * DROPs pkts if unused CQE < 2304
+ */
+#define RQ_PASS_CQ_LVL         224ULL
+#define RQ_DROP_CQ_LVL         216ULL
+
+/* RED and Backpressure levels of RBDR for pkt reception
+ * For RBDR, level is a measure of fullness i.e 0x0 means empty
+ * eg: For RBDR of size 8K, and for pass/drop levels of 4/0
+ * HW accepts pkt if unused RBs >= 256
+ * RED accepts pkt if unused RBs < 256 & >= 0
+ * DROPs pkts if unused RBs < 0
+ */
+#define RQ_PASS_RBDR_LVL	8ULL
+#define RQ_DROP_RBDR_LVL	0ULL
+
+/* Descriptor size in bytes */
+#define SND_QUEUE_DESC_SIZE	16
+#define CMP_QUEUE_DESC_SIZE	512
+
+/* Buffer / descriptor alignments */
+#define NICVF_RCV_BUF_ALIGN		7
+#define NICVF_RCV_BUF_ALIGN_BYTES	(1ULL << NICVF_RCV_BUF_ALIGN)
+#define NICVF_CQ_BASE_ALIGN_BYTES	512  /* 9 bits */
+#define NICVF_SQ_BASE_ALIGN_BYTES	128  /* 7 bits */
+
+#define NICVF_ALIGNED_ADDR(ADDR, ALIGN_BYTES)	ALIGN(ADDR, ALIGN_BYTES)
+
+/* Queue enable/disable */
+#define NICVF_SQ_EN		BIT_ULL(19)
+
+/* Queue reset */
+#define NICVF_CQ_RESET		BIT_ULL(41)
+#define NICVF_SQ_RESET		BIT_ULL(17)
+#define NICVF_RBDR_RESET	BIT_ULL(43)
+
+enum CQ_RX_ERRLVL_E {
+	CQ_ERRLVL_MAC,
+	CQ_ERRLVL_L2,
+	CQ_ERRLVL_L3,
+	CQ_ERRLVL_L4,
+};
+
+enum CQ_RX_ERROP_E {
+	CQ_RX_ERROP_RE_NONE = 0x0,
+	CQ_RX_ERROP_RE_PARTIAL = 0x1,
+	CQ_RX_ERROP_RE_JABBER = 0x2,
+	CQ_RX_ERROP_RE_FCS = 0x7,
+	CQ_RX_ERROP_RE_TERMINATE = 0x9,
+	CQ_RX_ERROP_RE_RX_CTL = 0xb,
+	CQ_RX_ERROP_PREL2_ERR = 0x1f,
+	CQ_RX_ERROP_L2_FRAGMENT = 0x20,
+	CQ_RX_ERROP_L2_OVERRUN = 0x21,
+	CQ_RX_ERROP_L2_PFCS = 0x22,
+	CQ_RX_ERROP_L2_PUNY = 0x23,
+	CQ_RX_ERROP_L2_MAL = 0x24,
+	CQ_RX_ERROP_L2_OVERSIZE = 0x25,
+	CQ_RX_ERROP_L2_UNDERSIZE = 0x26,
+	CQ_RX_ERROP_L2_LENMISM = 0x27,
+	CQ_RX_ERROP_L2_PCLP = 0x28,
+	CQ_RX_ERROP_IP_NOT = 0x41,
+	CQ_RX_ERROP_IP_CSUM_ERR = 0x42,
+	CQ_RX_ERROP_IP_MAL = 0x43,
+	CQ_RX_ERROP_IP_MALD = 0x44,
+	CQ_RX_ERROP_IP_HOP = 0x45,
+	CQ_RX_ERROP_L3_ICRC = 0x46,
+	CQ_RX_ERROP_L3_PCLP = 0x47,
+	CQ_RX_ERROP_L4_MAL = 0x61,
+	CQ_RX_ERROP_L4_CHK = 0x62,
+	CQ_RX_ERROP_UDP_LEN = 0x63,
+	CQ_RX_ERROP_L4_PORT = 0x64,
+	CQ_RX_ERROP_TCP_FLAG = 0x65,
+	CQ_RX_ERROP_TCP_OFFSET = 0x66,
+	CQ_RX_ERROP_L4_PCLP = 0x67,
+	CQ_RX_ERROP_RBDR_TRUNC = 0x70,
+};
+
+enum CQ_TX_ERROP_E {
+	CQ_TX_ERROP_GOOD = 0x0,
+	CQ_TX_ERROP_DESC_FAULT = 0x10,
+	CQ_TX_ERROP_HDR_CONS_ERR = 0x11,
+	CQ_TX_ERROP_SUBDC_ERR = 0x12,
+	CQ_TX_ERROP_MAX_SIZE_VIOL = 0x13,
+	CQ_TX_ERROP_IMM_SIZE_OFLOW = 0x80,
+	CQ_TX_ERROP_DATA_SEQUENCE_ERR = 0x81,
+	CQ_TX_ERROP_MEM_SEQUENCE_ERR = 0x82,
+	CQ_TX_ERROP_LOCK_VIOL = 0x83,
+	CQ_TX_ERROP_DATA_FAULT = 0x84,
+	CQ_TX_ERROP_TSTMP_CONFLICT = 0x85,
+	CQ_TX_ERROP_TSTMP_TIMEOUT = 0x86,
+	CQ_TX_ERROP_MEM_FAULT = 0x87,
+	CQ_TX_ERROP_CK_OVERLAP = 0x88,
+	CQ_TX_ERROP_CK_OFLOW = 0x89,
+	CQ_TX_ERROP_ENUM_LAST = 0x8a,
+};
+
+enum RQ_SQ_STATS {
+	RQ_SQ_STATS_OCTS,
+	RQ_SQ_STATS_PKTS,
+};
+
+struct rx_tx_queue_stats {
+	u64	bytes;
+	u64	pkts;
+} ____cacheline_aligned_in_smp;
+
+struct q_desc_mem {
+	dma_addr_t	dma;
+	u64		size;
+	u32		q_len;
+	dma_addr_t	phys_base;
+	void		*base;
+	void		*unalign_base;
+};
+
+struct pgcache {
+	struct page	*page;
+	int		ref_count;
+	u64		dma_addr;
+};
+
+struct rbdr {
+	bool		enable;
+	u32		dma_size;
+	u32		frag_len;
+	u32		thresh;		/* Threshold level for interrupt */
+	void		*desc;
+	u32		head;
+	u32		tail;
+	struct q_desc_mem   dmem;
+	bool		is_xdp;
+
+	/* For page recycling */
+	int		pgidx;
+	int		pgcnt;
+	int		pgalloc;
+	struct pgcache	*pgcache;
+} ____cacheline_aligned_in_smp;
+
+struct rcv_queue {
+	bool		enable;
+	struct	rbdr	*rbdr_start;
+	struct	rbdr	*rbdr_cont;
+	bool		en_tcp_reassembly;
+	u8		cq_qs;  /* CQ's QS to which this RQ is assigned */
+	u8		cq_idx; /* CQ index (0 to 7) in the QS */
+	u8		cont_rbdr_qs;      /* Continue buffer ptrs - QS num */
+	u8		cont_qs_rbdr_idx;  /* RBDR idx in the cont QS */
+	u8		start_rbdr_qs;     /* First buffer ptrs - QS num */
+	u8		start_qs_rbdr_idx; /* RBDR idx in the above QS */
+	u8		caching;
+	struct		rx_tx_queue_stats stats;
+	struct xdp_rxq_info xdp_rxq;
+} ____cacheline_aligned_in_smp;
+
+struct cmp_queue {
+	bool		enable;
+	u16		thresh;
+	spinlock_t	lock;  /* lock to serialize processing CQEs */
+	void		*desc;
+	struct q_desc_mem   dmem;
+	int		irq;
+} ____cacheline_aligned_in_smp;
+
+struct snd_queue {
+	bool		enable;
+	u8		cq_qs;  /* CQ's QS to which this SQ is pointing */
+	u8		cq_idx; /* CQ index (0 to 7) in the above QS */
+	u16		thresh;
+	atomic_t	free_cnt;
+	u32		head;
+	u32		tail;
+	u64		*skbuff;
+	void		*desc;
+	u64		*xdp_page;
+	u16		xdp_desc_cnt;
+	u16		xdp_free_cnt;
+	bool		is_xdp;
+
+	/* For TSO segment's header */
+	char		*tso_hdrs;
+	dma_addr_t	tso_hdrs_phys;
+
+	cpumask_t	affinity_mask;
+	struct q_desc_mem   dmem;
+	struct rx_tx_queue_stats stats;
+} ____cacheline_aligned_in_smp;
+
+struct queue_set {
+	bool		enable;
+	bool		be_en;
+	u8		vnic_id;
+	u8		rq_cnt;
+	u8		cq_cnt;
+	u64		cq_len;
+	u8		sq_cnt;
+	u64		sq_len;
+	u8		rbdr_cnt;
+	u64		rbdr_len;
+	struct	rcv_queue	rq[MAX_RCV_QUEUES_PER_QS];
+	struct	cmp_queue	cq[MAX_CMP_QUEUES_PER_QS];
+	struct	snd_queue	sq[MAX_SND_QUEUES_PER_QS];
+	struct	rbdr		rbdr[MAX_RCV_BUF_DESC_RINGS_PER_QS];
+} ____cacheline_aligned_in_smp;
+
+#define GET_RBDR_DESC(RING, idx)\
+		(&(((struct rbdr_entry_t *)((RING)->desc))[idx]))
+#define GET_SQ_DESC(RING, idx)\
+		(&(((struct sq_hdr_subdesc *)((RING)->desc))[idx]))
+#define GET_CQ_DESC(RING, idx)\
+		(&(((union cq_desc_t *)((RING)->desc))[idx]))
+
+/* CQ status bits */
+#define	CQ_WR_FULL	BIT(26)
+#define	CQ_WR_DISABLE	BIT(25)
+#define	CQ_WR_FAULT	BIT(24)
+#define	CQ_CQE_COUNT	(0xFFFF << 0)
+
+#define	CQ_ERR_MASK	(CQ_WR_FULL | CQ_WR_DISABLE | CQ_WR_FAULT)
+
+static inline u64 nicvf_iova_to_phys(struct nicvf *nic, dma_addr_t dma_addr)
+{
+	/* Translation is installed only when IOMMU is present */
+	if (nic->iommu_domain)
+		return iommu_iova_to_phys(nic->iommu_domain, dma_addr);
+	return dma_addr;
+}
+
+void nicvf_unmap_sndq_buffers(struct nicvf *nic, struct snd_queue *sq,
+			      int hdr_sqe, u8 subdesc_cnt);
+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);
+void nicvf_cmp_queue_config(struct nicvf *nic, struct queue_set *qs,
+			    int qidx, bool enable);
+
+void nicvf_sq_enable(struct nicvf *nic, struct snd_queue *sq, int qidx);
+void nicvf_sq_disable(struct nicvf *nic, int qidx);
+void nicvf_put_sq_desc(struct snd_queue *sq, int desc_cnt);
+void nicvf_sq_free_used_descs(struct net_device *netdev,
+			      struct snd_queue *sq, int qidx);
+int nicvf_sq_append_skb(struct nicvf *nic, struct snd_queue *sq,
+			struct sk_buff *skb, u8 sq_num);
+int nicvf_xdp_sq_append_pkt(struct nicvf *nic, struct snd_queue *sq,
+			    u64 bufaddr, u64 dma_addr, u16 len);
+void nicvf_xdp_sq_doorbell(struct nicvf *nic, struct snd_queue *sq, int sq_num);
+
+struct sk_buff *nicvf_get_rcv_skb(struct nicvf *nic,
+				  struct cqe_rx_t *cqe_rx, bool xdp);
+void nicvf_rbdr_task(struct tasklet_struct *t);
+void nicvf_rbdr_work(struct work_struct *work);
+
+void nicvf_enable_intr(struct nicvf *nic, int int_type, int q_idx);
+void nicvf_disable_intr(struct nicvf *nic, int int_type, int q_idx);
+void nicvf_clear_intr(struct nicvf *nic, int int_type, int q_idx);
+int nicvf_is_intr_enabled(struct nicvf *nic, int int_type, int q_idx);
+
+/* Register access APIs */
+void nicvf_reg_write(struct nicvf *nic, u64 offset, u64 val);
+u64  nicvf_reg_read(struct nicvf *nic, u64 offset);
+void nicvf_qset_reg_write(struct nicvf *nic, u64 offset, u64 val);
+u64 nicvf_qset_reg_read(struct nicvf *nic, u64 offset);
+void nicvf_queue_reg_write(struct nicvf *nic, u64 offset,
+			   u64 qidx, u64 val);
+u64  nicvf_queue_reg_read(struct nicvf *nic,
+			  u64 offset, u64 qidx);
+
+/* Stats */
+void nicvf_update_rq_stats(struct nicvf *nic, int rq_idx);
+void nicvf_update_sq_stats(struct nicvf *nic, int sq_idx);
+int nicvf_check_cqe_rx_errs(struct nicvf *nic, struct cqe_rx_t *cqe_rx);
+int nicvf_check_cqe_tx_errs(struct nicvf *nic, struct cqe_send_t *cqe_tx);
+#endif /* NICVF_QUEUES_H */
diff --git a/drivers/net/ethernet/cavium/thunder/q_struct.h b/drivers/net/ethernet/cavium/thunder/q_struct.h
new file mode 100644
index 000000000..0df115d42
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/q_struct.h
@@ -0,0 +1,696 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * This file contains HW queue descriptor formats, config register
+ * structures etc
+ *
+ * Copyright (C) 2015 Cavium, Inc.
+ */
+
+#ifndef Q_STRUCT_H
+#define Q_STRUCT_H
+
+/* Load transaction types for reading segment bytes specified by
+ * NIC_SEND_GATHER_S[LD_TYPE].
+ */
+enum nic_send_ld_type_e {
+	NIC_SEND_LD_TYPE_E_LDD = 0x0,
+	NIC_SEND_LD_TYPE_E_LDT = 0x1,
+	NIC_SEND_LD_TYPE_E_LDWB = 0x2,
+	NIC_SEND_LD_TYPE_E_ENUM_LAST = 0x3,
+};
+
+enum ether_type_algorithm {
+	ETYPE_ALG_NONE = 0x0,
+	ETYPE_ALG_SKIP = 0x1,
+	ETYPE_ALG_ENDPARSE = 0x2,
+	ETYPE_ALG_VLAN = 0x3,
+	ETYPE_ALG_VLAN_STRIP = 0x4,
+};
+
+enum layer3_type {
+	L3TYPE_NONE = 0x00,
+	L3TYPE_GRH = 0x01,
+	L3TYPE_IPV4 = 0x04,
+	L3TYPE_IPV4_OPTIONS = 0x05,
+	L3TYPE_IPV6 = 0x06,
+	L3TYPE_IPV6_OPTIONS = 0x07,
+	L3TYPE_ET_STOP = 0x0D,
+	L3TYPE_OTHER = 0x0E,
+};
+
+enum layer4_type {
+	L4TYPE_NONE = 0x00,
+	L4TYPE_IPSEC_ESP = 0x01,
+	L4TYPE_IPFRAG = 0x02,
+	L4TYPE_IPCOMP = 0x03,
+	L4TYPE_TCP = 0x04,
+	L4TYPE_UDP = 0x05,
+	L4TYPE_SCTP = 0x06,
+	L4TYPE_GRE = 0x07,
+	L4TYPE_ROCE_BTH = 0x08,
+	L4TYPE_OTHER = 0x0E,
+};
+
+/* CPI and RSSI configuration */
+enum cpi_algorithm_type {
+	CPI_ALG_NONE = 0x0,
+	CPI_ALG_VLAN = 0x1,
+	CPI_ALG_VLAN16 = 0x2,
+	CPI_ALG_DIFF = 0x3,
+};
+
+enum rss_algorithm_type {
+	RSS_ALG_NONE = 0x00,
+	RSS_ALG_PORT = 0x01,
+	RSS_ALG_IP = 0x02,
+	RSS_ALG_TCP_IP = 0x03,
+	RSS_ALG_UDP_IP = 0x04,
+	RSS_ALG_SCTP_IP = 0x05,
+	RSS_ALG_GRE_IP = 0x06,
+	RSS_ALG_ROCE = 0x07,
+};
+
+enum rss_hash_cfg {
+	RSS_HASH_L2ETC = 0x00,
+	RSS_HASH_IP = 0x01,
+	RSS_HASH_TCP = 0x02,
+	RSS_HASH_TCP_SYN_DIS = 0x03,
+	RSS_HASH_UDP = 0x04,
+	RSS_HASH_L4ETC = 0x05,
+	RSS_HASH_ROCE = 0x06,
+	RSS_L3_BIDI = 0x07,
+	RSS_L4_BIDI = 0x08,
+};
+
+/* Completion queue entry types */
+enum cqe_type {
+	CQE_TYPE_INVALID = 0x0,
+	CQE_TYPE_RX = 0x2,
+	CQE_TYPE_RX_SPLIT = 0x3,
+	CQE_TYPE_RX_TCP = 0x4,
+	CQE_TYPE_SEND = 0x8,
+	CQE_TYPE_SEND_PTP = 0x9,
+};
+
+enum cqe_rx_tcp_status {
+	CQE_RX_STATUS_VALID_TCP_CNXT = 0x00,
+	CQE_RX_STATUS_INVALID_TCP_CNXT = 0x0F,
+};
+
+enum cqe_send_status {
+	CQE_SEND_STATUS_GOOD = 0x00,
+	CQE_SEND_STATUS_DESC_FAULT = 0x01,
+	CQE_SEND_STATUS_HDR_CONS_ERR = 0x11,
+	CQE_SEND_STATUS_SUBDESC_ERR = 0x12,
+	CQE_SEND_STATUS_IMM_SIZE_OFLOW = 0x80,
+	CQE_SEND_STATUS_CRC_SEQ_ERR = 0x81,
+	CQE_SEND_STATUS_DATA_SEQ_ERR = 0x82,
+	CQE_SEND_STATUS_MEM_SEQ_ERR = 0x83,
+	CQE_SEND_STATUS_LOCK_VIOL = 0x84,
+	CQE_SEND_STATUS_LOCK_UFLOW = 0x85,
+	CQE_SEND_STATUS_DATA_FAULT = 0x86,
+	CQE_SEND_STATUS_TSTMP_CONFLICT = 0x87,
+	CQE_SEND_STATUS_TSTMP_TIMEOUT = 0x88,
+	CQE_SEND_STATUS_MEM_FAULT = 0x89,
+	CQE_SEND_STATUS_CSUM_OVERLAP = 0x8A,
+	CQE_SEND_STATUS_CSUM_OVERFLOW = 0x8B,
+};
+
+enum cqe_rx_tcp_end_reason {
+	CQE_RX_TCP_END_FIN_FLAG_DET = 0,
+	CQE_RX_TCP_END_INVALID_FLAG = 1,
+	CQE_RX_TCP_END_TIMEOUT = 2,
+	CQE_RX_TCP_END_OUT_OF_SEQ = 3,
+	CQE_RX_TCP_END_PKT_ERR = 4,
+	CQE_RX_TCP_END_QS_DISABLED = 0x0F,
+};
+
+/* Packet protocol level error enumeration */
+enum cqe_rx_err_level {
+	CQE_RX_ERRLVL_RE = 0x0,
+	CQE_RX_ERRLVL_L2 = 0x1,
+	CQE_RX_ERRLVL_L3 = 0x2,
+	CQE_RX_ERRLVL_L4 = 0x3,
+};
+
+/* Packet protocol level error type enumeration */
+enum cqe_rx_err_opcode {
+	CQE_RX_ERR_RE_NONE = 0x0,
+	CQE_RX_ERR_RE_PARTIAL = 0x1,
+	CQE_RX_ERR_RE_JABBER = 0x2,
+	CQE_RX_ERR_RE_FCS = 0x7,
+	CQE_RX_ERR_RE_TERMINATE = 0x9,
+	CQE_RX_ERR_RE_RX_CTL = 0xb,
+	CQE_RX_ERR_PREL2_ERR = 0x1f,
+	CQE_RX_ERR_L2_FRAGMENT = 0x20,
+	CQE_RX_ERR_L2_OVERRUN = 0x21,
+	CQE_RX_ERR_L2_PFCS = 0x22,
+	CQE_RX_ERR_L2_PUNY = 0x23,
+	CQE_RX_ERR_L2_MAL = 0x24,
+	CQE_RX_ERR_L2_OVERSIZE = 0x25,
+	CQE_RX_ERR_L2_UNDERSIZE = 0x26,
+	CQE_RX_ERR_L2_LENMISM = 0x27,
+	CQE_RX_ERR_L2_PCLP = 0x28,
+	CQE_RX_ERR_IP_NOT = 0x41,
+	CQE_RX_ERR_IP_CHK = 0x42,
+	CQE_RX_ERR_IP_MAL = 0x43,
+	CQE_RX_ERR_IP_MALD = 0x44,
+	CQE_RX_ERR_IP_HOP = 0x45,
+	CQE_RX_ERR_L3_ICRC = 0x46,
+	CQE_RX_ERR_L3_PCLP = 0x47,
+	CQE_RX_ERR_L4_MAL = 0x61,
+	CQE_RX_ERR_L4_CHK = 0x62,
+	CQE_RX_ERR_UDP_LEN = 0x63,
+	CQE_RX_ERR_L4_PORT = 0x64,
+	CQE_RX_ERR_TCP_FLAG = 0x65,
+	CQE_RX_ERR_TCP_OFFSET = 0x66,
+	CQE_RX_ERR_L4_PCLP = 0x67,
+	CQE_RX_ERR_RBDR_TRUNC = 0x70,
+};
+
+struct cqe_rx_t {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64   cqe_type:4; /* W0 */
+	u64   stdn_fault:1;
+	u64   rsvd0:1;
+	u64   rq_qs:7;
+	u64   rq_idx:3;
+	u64   rsvd1:12;
+	u64   rss_alg:4;
+	u64   rsvd2:4;
+	u64   rb_cnt:4;
+	u64   vlan_found:1;
+	u64   vlan_stripped:1;
+	u64   vlan2_found:1;
+	u64   vlan2_stripped:1;
+	u64   l4_type:4;
+	u64   l3_type:4;
+	u64   l2_present:1;
+	u64   err_level:3;
+	u64   err_opcode:8;
+
+	u64   pkt_len:16; /* W1 */
+	u64   l2_ptr:8;
+	u64   l3_ptr:8;
+	u64   l4_ptr:8;
+	u64   cq_pkt_len:8;
+	u64   align_pad:3;
+	u64   rsvd3:1;
+	u64   chan:12;
+
+	u64   rss_tag:32; /* W2 */
+	u64   vlan_tci:16;
+	u64   vlan_ptr:8;
+	u64   vlan2_ptr:8;
+
+	u64   rb3_sz:16; /* W3 */
+	u64   rb2_sz:16;
+	u64   rb1_sz:16;
+	u64   rb0_sz:16;
+
+	u64   rb7_sz:16; /* W4 */
+	u64   rb6_sz:16;
+	u64   rb5_sz:16;
+	u64   rb4_sz:16;
+
+	u64   rb11_sz:16; /* W5 */
+	u64   rb10_sz:16;
+	u64   rb9_sz:16;
+	u64   rb8_sz:16;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64   err_opcode:8;
+	u64   err_level:3;
+	u64   l2_present:1;
+	u64   l3_type:4;
+	u64   l4_type:4;
+	u64   vlan2_stripped:1;
+	u64   vlan2_found:1;
+	u64   vlan_stripped:1;
+	u64   vlan_found:1;
+	u64   rb_cnt:4;
+	u64   rsvd2:4;
+	u64   rss_alg:4;
+	u64   rsvd1:12;
+	u64   rq_idx:3;
+	u64   rq_qs:7;
+	u64   rsvd0:1;
+	u64   stdn_fault:1;
+	u64   cqe_type:4; /* W0 */
+	u64   chan:12;
+	u64   rsvd3:1;
+	u64   align_pad:3;
+	u64   cq_pkt_len:8;
+	u64   l4_ptr:8;
+	u64   l3_ptr:8;
+	u64   l2_ptr:8;
+	u64   pkt_len:16; /* W1 */
+	u64   vlan2_ptr:8;
+	u64   vlan_ptr:8;
+	u64   vlan_tci:16;
+	u64   rss_tag:32; /* W2 */
+	u64   rb0_sz:16;
+	u64   rb1_sz:16;
+	u64   rb2_sz:16;
+	u64   rb3_sz:16; /* W3 */
+	u64   rb4_sz:16;
+	u64   rb5_sz:16;
+	u64   rb6_sz:16;
+	u64   rb7_sz:16; /* W4 */
+	u64   rb8_sz:16;
+	u64   rb9_sz:16;
+	u64   rb10_sz:16;
+	u64   rb11_sz:16; /* W5 */
+#endif
+	u64   rb0_ptr:64;
+	u64   rb1_ptr:64;
+	u64   rb2_ptr:64;
+	u64   rb3_ptr:64;
+	u64   rb4_ptr:64;
+	u64   rb5_ptr:64;
+	u64   rb6_ptr:64;
+	u64   rb7_ptr:64;
+	u64   rb8_ptr:64;
+	u64   rb9_ptr:64;
+	u64   rb10_ptr:64;
+	u64   rb11_ptr:64;
+};
+
+struct cqe_rx_tcp_err_t {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64   cqe_type:4; /* W0 */
+	u64   rsvd0:60;
+
+	u64   rsvd1:4; /* W1 */
+	u64   partial_first:1;
+	u64   rsvd2:27;
+	u64   rbdr_bytes:8;
+	u64   rsvd3:24;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64   rsvd0:60;
+	u64   cqe_type:4;
+
+	u64   rsvd3:24;
+	u64   rbdr_bytes:8;
+	u64   rsvd2:27;
+	u64   partial_first:1;
+	u64   rsvd1:4;
+#endif
+};
+
+struct cqe_rx_tcp_t {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64   cqe_type:4; /* W0 */
+	u64   rsvd0:52;
+	u64   cq_tcp_status:8;
+
+	u64   rsvd1:32; /* W1 */
+	u64   tcp_cntx_bytes:8;
+	u64   rsvd2:8;
+	u64   tcp_err_bytes:16;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64   cq_tcp_status:8;
+	u64   rsvd0:52;
+	u64   cqe_type:4; /* W0 */
+
+	u64   tcp_err_bytes:16;
+	u64   rsvd2:8;
+	u64   tcp_cntx_bytes:8;
+	u64   rsvd1:32; /* W1 */
+#endif
+};
+
+struct cqe_send_t {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64   cqe_type:4; /* W0 */
+	u64   rsvd0:4;
+	u64   sqe_ptr:16;
+	u64   rsvd1:4;
+	u64   rsvd2:10;
+	u64   sq_qs:7;
+	u64   sq_idx:3;
+	u64   rsvd3:8;
+	u64   send_status:8;
+
+	u64   ptp_timestamp:64; /* W1 */
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64   send_status:8;
+	u64   rsvd3:8;
+	u64   sq_idx:3;
+	u64   sq_qs:7;
+	u64   rsvd2:10;
+	u64   rsvd1:4;
+	u64   sqe_ptr:16;
+	u64   rsvd0:4;
+	u64   cqe_type:4; /* W0 */
+
+	u64   ptp_timestamp:64; /* W1 */
+#endif
+};
+
+union cq_desc_t {
+	u64    u[64];
+	struct cqe_send_t snd_hdr;
+	struct cqe_rx_t rx_hdr;
+	struct cqe_rx_tcp_t rx_tcp_hdr;
+	struct cqe_rx_tcp_err_t rx_tcp_err_hdr;
+};
+
+struct rbdr_entry_t {
+	u64   buf_addr;
+};
+
+/* TCP reassembly context */
+struct rbe_tcp_cnxt_t {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64   tcp_pkt_cnt:12;
+	u64   rsvd1:4;
+	u64   align_hdr_bytes:4;
+	u64   align_ptr_bytes:4;
+	u64   ptr_bytes:16;
+	u64   rsvd2:24;
+	u64   cqe_type:4;
+	u64   rsvd0:54;
+	u64   tcp_end_reason:2;
+	u64   tcp_status:4;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64   tcp_status:4;
+	u64   tcp_end_reason:2;
+	u64   rsvd0:54;
+	u64   cqe_type:4;
+	u64   rsvd2:24;
+	u64   ptr_bytes:16;
+	u64   align_ptr_bytes:4;
+	u64   align_hdr_bytes:4;
+	u64   rsvd1:4;
+	u64   tcp_pkt_cnt:12;
+#endif
+};
+
+/* Always Big endian */
+struct rx_hdr_t {
+	u64   opaque:32;
+	u64   rss_flow:8;
+	u64   skip_length:6;
+	u64   disable_rss:1;
+	u64   disable_tcp_reassembly:1;
+	u64   nodrop:1;
+	u64   dest_alg:2;
+	u64   rsvd0:2;
+	u64   dest_rq:11;
+};
+
+enum send_l4_csum_type {
+	SEND_L4_CSUM_DISABLE = 0x00,
+	SEND_L4_CSUM_UDP = 0x01,
+	SEND_L4_CSUM_TCP = 0x02,
+	SEND_L4_CSUM_SCTP = 0x03,
+};
+
+enum send_crc_alg {
+	SEND_CRCALG_CRC32 = 0x00,
+	SEND_CRCALG_CRC32C = 0x01,
+	SEND_CRCALG_ICRC = 0x02,
+};
+
+enum send_load_type {
+	SEND_LD_TYPE_LDD = 0x00,
+	SEND_LD_TYPE_LDT = 0x01,
+	SEND_LD_TYPE_LDWB = 0x02,
+};
+
+enum send_mem_alg_type {
+	SEND_MEMALG_SET = 0x00,
+	SEND_MEMALG_ADD = 0x08,
+	SEND_MEMALG_SUB = 0x09,
+	SEND_MEMALG_ADDLEN = 0x0A,
+	SEND_MEMALG_SUBLEN = 0x0B,
+};
+
+enum send_mem_dsz_type {
+	SEND_MEMDSZ_B64 = 0x00,
+	SEND_MEMDSZ_B32 = 0x01,
+	SEND_MEMDSZ_B8 = 0x03,
+};
+
+enum sq_subdesc_type {
+	SQ_DESC_TYPE_INVALID = 0x00,
+	SQ_DESC_TYPE_HEADER = 0x01,
+	SQ_DESC_TYPE_CRC = 0x02,
+	SQ_DESC_TYPE_IMMEDIATE = 0x03,
+	SQ_DESC_TYPE_GATHER = 0x04,
+	SQ_DESC_TYPE_MEMORY = 0x05,
+};
+
+struct sq_crc_subdesc {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64    rsvd1:32;
+	u64    crc_ival:32;
+	u64    subdesc_type:4;
+	u64    crc_alg:2;
+	u64    rsvd0:10;
+	u64    crc_insert_pos:16;
+	u64    hdr_start:16;
+	u64    crc_len:16;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64    crc_len:16;
+	u64    hdr_start:16;
+	u64    crc_insert_pos:16;
+	u64    rsvd0:10;
+	u64    crc_alg:2;
+	u64    subdesc_type:4;
+	u64    crc_ival:32;
+	u64    rsvd1:32;
+#endif
+};
+
+struct sq_gather_subdesc {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64    subdesc_type:4; /* W0 */
+	u64    ld_type:2;
+	u64    rsvd0:42;
+	u64    size:16;
+
+	u64    rsvd1:15; /* W1 */
+	u64    addr:49;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64    size:16;
+	u64    rsvd0:42;
+	u64    ld_type:2;
+	u64    subdesc_type:4; /* W0 */
+
+	u64    addr:49;
+	u64    rsvd1:15; /* W1 */
+#endif
+};
+
+/* SQ immediate subdescriptor */
+struct sq_imm_subdesc {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64    subdesc_type:4; /* W0 */
+	u64    rsvd0:46;
+	u64    len:14;
+
+	u64    data:64; /* W1 */
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64    len:14;
+	u64    rsvd0:46;
+	u64    subdesc_type:4; /* W0 */
+
+	u64    data:64; /* W1 */
+#endif
+};
+
+struct sq_mem_subdesc {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64    subdesc_type:4; /* W0 */
+	u64    mem_alg:4;
+	u64    mem_dsz:2;
+	u64    wmem:1;
+	u64    rsvd0:21;
+	u64    offset:32;
+
+	u64    rsvd1:15; /* W1 */
+	u64    addr:49;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64    offset:32;
+	u64    rsvd0:21;
+	u64    wmem:1;
+	u64    mem_dsz:2;
+	u64    mem_alg:4;
+	u64    subdesc_type:4; /* W0 */
+
+	u64    addr:49;
+	u64    rsvd1:15; /* W1 */
+#endif
+};
+
+struct sq_hdr_subdesc {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64    subdesc_type:4;
+	u64    tso:1;
+	u64    post_cqe:1; /* Post CQE on no error also */
+	u64    dont_send:1;
+	u64    tstmp:1;
+	u64    subdesc_cnt:8;
+	u64    csum_l4:2;
+	u64    csum_l3:1;
+	u64    csum_inner_l4:2;
+	u64    csum_inner_l3:1;
+	u64    rsvd0:2;
+	u64    l4_offset:8;
+	u64    l3_offset:8;
+	u64    rsvd1:4;
+	u64    tot_len:20; /* W0 */
+
+	u64    rsvd2:24;
+	u64    inner_l4_offset:8;
+	u64    inner_l3_offset:8;
+	u64    tso_start:8;
+	u64    rsvd3:2;
+	u64    tso_max_paysize:14; /* W1 */
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64    tot_len:20;
+	u64    rsvd1:4;
+	u64    l3_offset:8;
+	u64    l4_offset:8;
+	u64    rsvd0:2;
+	u64    csum_inner_l3:1;
+	u64    csum_inner_l4:2;
+	u64    csum_l3:1;
+	u64    csum_l4:2;
+	u64    subdesc_cnt:8;
+	u64    tstmp:1;
+	u64    dont_send:1;
+	u64    post_cqe:1; /* Post CQE on no error also */
+	u64    tso:1;
+	u64    subdesc_type:4; /* W0 */
+
+	u64    tso_max_paysize:14;
+	u64    rsvd3:2;
+	u64    tso_start:8;
+	u64    inner_l3_offset:8;
+	u64    inner_l4_offset:8;
+	u64    rsvd2:24; /* W1 */
+#endif
+};
+
+/* Queue config register formats */
+struct rq_cfg {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64 reserved_2_63:62;
+	u64 ena:1;
+	u64 tcp_ena:1;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64 tcp_ena:1;
+	u64 ena:1;
+	u64 reserved_2_63:62;
+#endif
+};
+
+struct cq_cfg {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64 reserved_43_63:21;
+	u64 ena:1;
+	u64 reset:1;
+	u64 caching:1;
+	u64 reserved_35_39:5;
+	u64 qsize:3;
+	u64 reserved_25_31:7;
+	u64 avg_con:9;
+	u64 reserved_0_15:16;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64 reserved_0_15:16;
+	u64 avg_con:9;
+	u64 reserved_25_31:7;
+	u64 qsize:3;
+	u64 reserved_35_39:5;
+	u64 caching:1;
+	u64 reset:1;
+	u64 ena:1;
+	u64 reserved_43_63:21;
+#endif
+};
+
+struct sq_cfg {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64 reserved_32_63:32;
+	u64 cq_limit:8;
+	u64 reserved_20_23:4;
+	u64 ena:1;
+	u64 reserved_18_18:1;
+	u64 reset:1;
+	u64 ldwb:1;
+	u64 reserved_11_15:5;
+	u64 qsize:3;
+	u64 reserved_3_7:5;
+	u64 tstmp_bgx_intf:3;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64 tstmp_bgx_intf:3;
+	u64 reserved_3_7:5;
+	u64 qsize:3;
+	u64 reserved_11_15:5;
+	u64 ldwb:1;
+	u64 reset:1;
+	u64 reserved_18_18:1;
+	u64 ena:1;
+	u64 reserved_20_23:4;
+	u64 cq_limit:8;
+	u64 reserved_32_63:32;
+#endif
+};
+
+struct rbdr_cfg {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64 reserved_45_63:19;
+	u64 ena:1;
+	u64 reset:1;
+	u64 ldwb:1;
+	u64 reserved_36_41:6;
+	u64 qsize:4;
+	u64 reserved_25_31:7;
+	u64 avg_con:9;
+	u64 reserved_12_15:4;
+	u64 lines:12;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64 lines:12;
+	u64 reserved_12_15:4;
+	u64 avg_con:9;
+	u64 reserved_25_31:7;
+	u64 qsize:4;
+	u64 reserved_36_41:6;
+	u64 ldwb:1;
+	u64 reset:1;
+	u64 ena: 1;
+	u64 reserved_45_63:19;
+#endif
+};
+
+struct qs_cfg {
+#if defined(__BIG_ENDIAN_BITFIELD)
+	u64 reserved_32_63:32;
+	u64 ena:1;
+	u64 reserved_27_30:4;
+	u64 sq_ins_ena:1;
+	u64 sq_ins_pos:6;
+	u64 lock_ena:1;
+	u64 lock_viol_cqe_ena:1;
+	u64 send_tstmp_ena:1;
+	u64 be:1;
+	u64 reserved_7_15:9;
+	u64 vnic:7;
+#elif defined(__LITTLE_ENDIAN_BITFIELD)
+	u64 vnic:7;
+	u64 reserved_7_15:9;
+	u64 be:1;
+	u64 send_tstmp_ena:1;
+	u64 lock_viol_cqe_ena:1;
+	u64 lock_ena:1;
+	u64 sq_ins_pos:6;
+	u64 sq_ins_ena:1;
+	u64 reserved_27_30:4;
+	u64 ena:1;
+	u64 reserved_32_63:32;
+#endif
+};
+
+#endif /* Q_STRUCT_H */
diff --git a/drivers/net/ethernet/cavium/thunder/thunder_bgx.c b/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
new file mode 100644
index 000000000..7eb2ddbe9
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/thunder_bgx.c
@@ -0,0 +1,1718 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2015 Cavium, Inc.
+ */
+
+#include <linux/acpi.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/phy.h>
+#include <linux/of.h>
+#include <linux/of_mdio.h>
+#include <linux/of_net.h>
+
+#include "nic_reg.h"
+#include "nic.h"
+#include "thunder_bgx.h"
+
+#define DRV_NAME	"thunder_bgx"
+#define DRV_VERSION	"1.0"
+
+/* RX_DMAC_CTL configuration */
+enum MCAST_MODE {
+		MCAST_MODE_REJECT = 0x0,
+		MCAST_MODE_ACCEPT = 0x1,
+		MCAST_MODE_CAM_FILTER = 0x2,
+		RSVD = 0x3
+};
+
+#define BCAST_ACCEPT      BIT(0)
+#define CAM_ACCEPT        BIT(3)
+#define MCAST_MODE_MASK   0x3
+#define BGX_MCAST_MODE(x) (x << 1)
+
+struct dmac_map {
+	u64                     vf_map;
+	u64                     dmac;
+};
+
+struct lmac {
+	struct bgx		*bgx;
+	/* actual number of DMACs configured */
+	u8			dmacs_cfg;
+	/* overal number of possible DMACs could be configured per LMAC */
+	u8                      dmacs_count;
+	struct dmac_map         *dmacs; /* DMAC:VFs tracking filter array */
+	u8			mac[ETH_ALEN];
+	u8                      lmac_type;
+	u8                      lane_to_sds;
+	bool                    use_training;
+	bool                    autoneg;
+	bool			link_up;
+	int			lmacid; /* ID within BGX */
+	int			lmacid_bd; /* ID on board */
+	struct net_device       netdev;
+	struct phy_device       *phydev;
+	unsigned int            last_duplex;
+	unsigned int            last_link;
+	unsigned int            last_speed;
+	bool			is_sgmii;
+	struct delayed_work	dwork;
+	struct workqueue_struct *check_link;
+};
+
+struct bgx {
+	u8			bgx_id;
+	struct	lmac		lmac[MAX_LMAC_PER_BGX];
+	u8			lmac_count;
+	u8			max_lmac;
+	u8                      acpi_lmac_idx;
+	void __iomem		*reg_base;
+	struct pci_dev		*pdev;
+	bool                    is_dlm;
+	bool                    is_rgx;
+};
+
+static struct bgx *bgx_vnic[MAX_BGX_THUNDER];
+static int lmac_count; /* Total no of LMACs in system */
+
+static int bgx_xaui_check_link(struct lmac *lmac);
+
+/* Supported devices */
+static const struct pci_device_id bgx_id_table[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_CAVIUM, PCI_DEVICE_ID_THUNDER_BGX) },
+	{ PCI_DEVICE(PCI_VENDOR_ID_CAVIUM, PCI_DEVICE_ID_THUNDER_RGX) },
+	{ 0, }  /* end of table */
+};
+
+MODULE_AUTHOR("Cavium Inc");
+MODULE_DESCRIPTION("Cavium Thunder BGX/MAC Driver");
+MODULE_LICENSE("GPL v2");
+MODULE_VERSION(DRV_VERSION);
+MODULE_DEVICE_TABLE(pci, bgx_id_table);
+
+/* The Cavium ThunderX network controller can *only* be found in SoCs
+ * containing the ThunderX ARM64 CPU implementation.  All accesses to the device
+ * registers on this platform are implicitly strongly ordered with respect
+ * to memory accesses. So writeq_relaxed() and readq_relaxed() are safe to use
+ * with no memory barriers in this driver.  The readq()/writeq() functions add
+ * explicit ordering operation which in this case are redundant, and only
+ * add overhead.
+ */
+
+/* Register read/write APIs */
+static u64 bgx_reg_read(struct bgx *bgx, u8 lmac, u64 offset)
+{
+	void __iomem *addr = bgx->reg_base + ((u32)lmac << 20) + offset;
+
+	return readq_relaxed(addr);
+}
+
+static void bgx_reg_write(struct bgx *bgx, u8 lmac, u64 offset, u64 val)
+{
+	void __iomem *addr = bgx->reg_base + ((u32)lmac << 20) + offset;
+
+	writeq_relaxed(val, addr);
+}
+
+static void bgx_reg_modify(struct bgx *bgx, u8 lmac, u64 offset, u64 val)
+{
+	void __iomem *addr = bgx->reg_base + ((u32)lmac << 20) + offset;
+
+	writeq_relaxed(val | readq_relaxed(addr), addr);
+}
+
+static int bgx_poll_reg(struct bgx *bgx, u8 lmac, u64 reg, u64 mask, bool zero)
+{
+	int timeout = 100;
+	u64 reg_val;
+
+	while (timeout) {
+		reg_val = bgx_reg_read(bgx, lmac, reg);
+		if (zero && !(reg_val & mask))
+			return 0;
+		if (!zero && (reg_val & mask))
+			return 0;
+		usleep_range(1000, 2000);
+		timeout--;
+	}
+	return 1;
+}
+
+static int max_bgx_per_node;
+static void set_max_bgx_per_node(struct pci_dev *pdev)
+{
+	u16 sdevid;
+
+	if (max_bgx_per_node)
+		return;
+
+	pci_read_config_word(pdev, PCI_SUBSYSTEM_ID, &sdevid);
+	switch (sdevid) {
+	case PCI_SUBSYS_DEVID_81XX_BGX:
+	case PCI_SUBSYS_DEVID_81XX_RGX:
+		max_bgx_per_node = MAX_BGX_PER_CN81XX;
+		break;
+	case PCI_SUBSYS_DEVID_83XX_BGX:
+		max_bgx_per_node = MAX_BGX_PER_CN83XX;
+		break;
+	case PCI_SUBSYS_DEVID_88XX_BGX:
+	default:
+		max_bgx_per_node = MAX_BGX_PER_CN88XX;
+		break;
+	}
+}
+
+static struct bgx *get_bgx(int node, int bgx_idx)
+{
+	int idx = (node * max_bgx_per_node) + bgx_idx;
+
+	return bgx_vnic[idx];
+}
+
+/* Return number of BGX present in HW */
+unsigned bgx_get_map(int node)
+{
+	int i;
+	unsigned map = 0;
+
+	for (i = 0; i < max_bgx_per_node; i++) {
+		if (bgx_vnic[(node * max_bgx_per_node) + i])
+			map |= (1 << i);
+	}
+
+	return map;
+}
+EXPORT_SYMBOL(bgx_get_map);
+
+/* Return number of LMAC configured for this BGX */
+int bgx_get_lmac_count(int node, int bgx_idx)
+{
+	struct bgx *bgx;
+
+	bgx = get_bgx(node, bgx_idx);
+	if (bgx)
+		return bgx->lmac_count;
+
+	return 0;
+}
+EXPORT_SYMBOL(bgx_get_lmac_count);
+
+/* Returns the current link status of LMAC */
+void bgx_get_lmac_link_state(int node, int bgx_idx, int lmacid, void *status)
+{
+	struct bgx_link_status *link = (struct bgx_link_status *)status;
+	struct bgx *bgx;
+	struct lmac *lmac;
+
+	bgx = get_bgx(node, bgx_idx);
+	if (!bgx)
+		return;
+
+	lmac = &bgx->lmac[lmacid];
+	link->mac_type = lmac->lmac_type;
+	link->link_up = lmac->link_up;
+	link->duplex = lmac->last_duplex;
+	link->speed = lmac->last_speed;
+}
+EXPORT_SYMBOL(bgx_get_lmac_link_state);
+
+const u8 *bgx_get_lmac_mac(int node, int bgx_idx, int lmacid)
+{
+	struct bgx *bgx = get_bgx(node, bgx_idx);
+
+	if (bgx)
+		return bgx->lmac[lmacid].mac;
+
+	return NULL;
+}
+EXPORT_SYMBOL(bgx_get_lmac_mac);
+
+void bgx_set_lmac_mac(int node, int bgx_idx, int lmacid, const u8 *mac)
+{
+	struct bgx *bgx = get_bgx(node, bgx_idx);
+
+	if (!bgx)
+		return;
+
+	ether_addr_copy(bgx->lmac[lmacid].mac, mac);
+}
+EXPORT_SYMBOL(bgx_set_lmac_mac);
+
+static void bgx_flush_dmac_cam_filter(struct bgx *bgx, int lmacid)
+{
+	struct lmac *lmac = NULL;
+	u8  idx = 0;
+
+	lmac = &bgx->lmac[lmacid];
+	/* reset CAM filters */
+	for (idx = 0; idx < lmac->dmacs_count; idx++)
+		bgx_reg_write(bgx, 0, BGX_CMR_RX_DMACX_CAM +
+			      ((lmacid * lmac->dmacs_count) + idx) *
+			      sizeof(u64), 0);
+}
+
+static void bgx_lmac_remove_filters(struct lmac *lmac, u8 vf_id)
+{
+	int i = 0;
+
+	if (!lmac)
+		return;
+
+	/* We've got reset filters request from some of attached VF, while the
+	 * others might want to keep their configuration. So in this case lets
+	 * iterate over all of configured filters and decrease number of
+	 * referencies. if some addresses get zero refs remove them from list
+	 */
+	for (i = lmac->dmacs_cfg - 1; i >= 0; i--) {
+		lmac->dmacs[i].vf_map &= ~BIT_ULL(vf_id);
+		if (!lmac->dmacs[i].vf_map) {
+			lmac->dmacs_cfg--;
+			lmac->dmacs[i].dmac = 0;
+			lmac->dmacs[i].vf_map = 0;
+		}
+	}
+}
+
+static int bgx_lmac_save_filter(struct lmac *lmac, u64 dmac, u8 vf_id)
+{
+	u8 i = 0;
+
+	if (!lmac)
+		return -1;
+
+	/* At the same time we could have several VFs 'attached' to some
+	 * particular LMAC, and each VF is represented as network interface
+	 * for kernel. So from user perspective it should be possible to
+	 * manipulate with its' (VF) receive modes. However from PF
+	 * driver perspective we need to keep track of filter configurations
+	 * for different VFs to prevent filter values dupes
+	 */
+	for (i = 0; i < lmac->dmacs_cfg; i++) {
+		if (lmac->dmacs[i].dmac == dmac) {
+			lmac->dmacs[i].vf_map |= BIT_ULL(vf_id);
+			return -1;
+		}
+	}
+
+	if (!(lmac->dmacs_cfg < lmac->dmacs_count))
+		return -1;
+
+	/* keep it for further tracking */
+	lmac->dmacs[lmac->dmacs_cfg].dmac = dmac;
+	lmac->dmacs[lmac->dmacs_cfg].vf_map = BIT_ULL(vf_id);
+	lmac->dmacs_cfg++;
+	return 0;
+}
+
+static int bgx_set_dmac_cam_filter_mac(struct bgx *bgx, int lmacid,
+				       u64 cam_dmac, u8 idx)
+{
+	struct lmac *lmac = NULL;
+	u64 cfg = 0;
+
+	/* skip zero addresses as meaningless */
+	if (!cam_dmac || !bgx)
+		return -1;
+
+	lmac = &bgx->lmac[lmacid];
+
+	/* configure DCAM filtering for designated LMAC */
+	cfg = RX_DMACX_CAM_LMACID(lmacid & LMAC_ID_MASK) |
+		RX_DMACX_CAM_EN | cam_dmac;
+	bgx_reg_write(bgx, 0, BGX_CMR_RX_DMACX_CAM +
+		      ((lmacid * lmac->dmacs_count) + idx) * sizeof(u64), cfg);
+	return 0;
+}
+
+void bgx_set_dmac_cam_filter(int node, int bgx_idx, int lmacid,
+			     u64 cam_dmac, u8 vf_id)
+{
+	struct bgx *bgx = get_bgx(node, bgx_idx);
+	struct lmac *lmac = NULL;
+
+	if (!bgx)
+		return;
+
+	lmac = &bgx->lmac[lmacid];
+
+	if (!cam_dmac)
+		cam_dmac = ether_addr_to_u64(lmac->mac);
+
+	/* since we might have several VFs attached to particular LMAC
+	 * and kernel could call mcast config for each of them with the
+	 * same MAC, check if requested MAC is already in filtering list and
+	 * updare/prepare list of MACs to be applied later to HW filters
+	 */
+	bgx_lmac_save_filter(lmac, cam_dmac, vf_id);
+}
+EXPORT_SYMBOL(bgx_set_dmac_cam_filter);
+
+void bgx_set_xcast_mode(int node, int bgx_idx, int lmacid, u8 mode)
+{
+	struct bgx *bgx = get_bgx(node, bgx_idx);
+	struct lmac *lmac = NULL;
+	u64 cfg = 0;
+	u8 i = 0;
+
+	if (!bgx)
+		return;
+
+	lmac = &bgx->lmac[lmacid];
+
+	cfg = bgx_reg_read(bgx, lmacid, BGX_CMRX_RX_DMAC_CTL);
+	if (mode & BGX_XCAST_BCAST_ACCEPT)
+		cfg |= BCAST_ACCEPT;
+	else
+		cfg &= ~BCAST_ACCEPT;
+
+	/* disable all MCASTs and DMAC filtering */
+	cfg &= ~(CAM_ACCEPT | BGX_MCAST_MODE(MCAST_MODE_MASK));
+
+	/* check requested bits and set filtergin mode appropriately */
+	if (mode & (BGX_XCAST_MCAST_ACCEPT)) {
+		cfg |= (BGX_MCAST_MODE(MCAST_MODE_ACCEPT));
+	} else if (mode & BGX_XCAST_MCAST_FILTER) {
+		cfg |= (BGX_MCAST_MODE(MCAST_MODE_CAM_FILTER) | CAM_ACCEPT);
+		for (i = 0; i < lmac->dmacs_cfg; i++)
+			bgx_set_dmac_cam_filter_mac(bgx, lmacid,
+						    lmac->dmacs[i].dmac, i);
+	}
+	bgx_reg_write(bgx, lmacid, BGX_CMRX_RX_DMAC_CTL, cfg);
+}
+EXPORT_SYMBOL(bgx_set_xcast_mode);
+
+void bgx_reset_xcast_mode(int node, int bgx_idx, int lmacid, u8 vf_id)
+{
+	struct bgx *bgx = get_bgx(node, bgx_idx);
+
+	if (!bgx)
+		return;
+
+	bgx_lmac_remove_filters(&bgx->lmac[lmacid], vf_id);
+	bgx_flush_dmac_cam_filter(bgx, lmacid);
+	bgx_set_xcast_mode(node, bgx_idx, lmacid,
+			   (BGX_XCAST_BCAST_ACCEPT | BGX_XCAST_MCAST_ACCEPT));
+}
+EXPORT_SYMBOL(bgx_reset_xcast_mode);
+
+void bgx_lmac_rx_tx_enable(int node, int bgx_idx, int lmacid, bool enable)
+{
+	struct bgx *bgx = get_bgx(node, bgx_idx);
+	struct lmac *lmac;
+	u64 cfg;
+
+	if (!bgx)
+		return;
+	lmac = &bgx->lmac[lmacid];
+
+	cfg = bgx_reg_read(bgx, lmacid, BGX_CMRX_CFG);
+	if (enable) {
+		cfg |= CMR_PKT_RX_EN | CMR_PKT_TX_EN;
+
+		/* enable TX FIFO Underflow interrupt */
+		bgx_reg_modify(bgx, lmacid, BGX_GMP_GMI_TXX_INT_ENA_W1S,
+			       GMI_TXX_INT_UNDFLW);
+	} else {
+		cfg &= ~(CMR_PKT_RX_EN | CMR_PKT_TX_EN);
+
+		/* Disable TX FIFO Underflow interrupt */
+		bgx_reg_modify(bgx, lmacid, BGX_GMP_GMI_TXX_INT_ENA_W1C,
+			       GMI_TXX_INT_UNDFLW);
+	}
+	bgx_reg_write(bgx, lmacid, BGX_CMRX_CFG, cfg);
+
+	if (bgx->is_rgx)
+		xcv_setup_link(enable ? lmac->link_up : 0, lmac->last_speed);
+}
+EXPORT_SYMBOL(bgx_lmac_rx_tx_enable);
+
+/* Enables or disables timestamp insertion by BGX for Rx packets */
+void bgx_config_timestamping(int node, int bgx_idx, int lmacid, bool enable)
+{
+	struct bgx *bgx = get_bgx(node, bgx_idx);
+	struct lmac *lmac;
+	u64 csr_offset, cfg;
+
+	if (!bgx)
+		return;
+
+	lmac = &bgx->lmac[lmacid];
+
+	if (lmac->lmac_type == BGX_MODE_SGMII ||
+	    lmac->lmac_type == BGX_MODE_QSGMII ||
+	    lmac->lmac_type == BGX_MODE_RGMII)
+		csr_offset = BGX_GMP_GMI_RXX_FRM_CTL;
+	else
+		csr_offset = BGX_SMUX_RX_FRM_CTL;
+
+	cfg = bgx_reg_read(bgx, lmacid, csr_offset);
+
+	if (enable)
+		cfg |= BGX_PKT_RX_PTP_EN;
+	else
+		cfg &= ~BGX_PKT_RX_PTP_EN;
+	bgx_reg_write(bgx, lmacid, csr_offset, cfg);
+}
+EXPORT_SYMBOL(bgx_config_timestamping);
+
+void bgx_lmac_get_pfc(int node, int bgx_idx, int lmacid, void *pause)
+{
+	struct pfc *pfc = (struct pfc *)pause;
+	struct bgx *bgx = get_bgx(node, bgx_idx);
+	struct lmac *lmac;
+	u64 cfg;
+
+	if (!bgx)
+		return;
+	lmac = &bgx->lmac[lmacid];
+	if (lmac->is_sgmii)
+		return;
+
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SMUX_CBFC_CTL);
+	pfc->fc_rx = cfg & RX_EN;
+	pfc->fc_tx = cfg & TX_EN;
+	pfc->autoneg = 0;
+}
+EXPORT_SYMBOL(bgx_lmac_get_pfc);
+
+void bgx_lmac_set_pfc(int node, int bgx_idx, int lmacid, void *pause)
+{
+	struct pfc *pfc = (struct pfc *)pause;
+	struct bgx *bgx = get_bgx(node, bgx_idx);
+	struct lmac *lmac;
+	u64 cfg;
+
+	if (!bgx)
+		return;
+	lmac = &bgx->lmac[lmacid];
+	if (lmac->is_sgmii)
+		return;
+
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SMUX_CBFC_CTL);
+	cfg &= ~(RX_EN | TX_EN);
+	cfg |= (pfc->fc_rx ? RX_EN : 0x00);
+	cfg |= (pfc->fc_tx ? TX_EN : 0x00);
+	bgx_reg_write(bgx, lmacid, BGX_SMUX_CBFC_CTL, cfg);
+}
+EXPORT_SYMBOL(bgx_lmac_set_pfc);
+
+static void bgx_sgmii_change_link_state(struct lmac *lmac)
+{
+	struct bgx *bgx = lmac->bgx;
+	u64 cmr_cfg;
+	u64 port_cfg = 0;
+	u64 misc_ctl = 0;
+	bool tx_en, rx_en;
+
+	cmr_cfg = bgx_reg_read(bgx, lmac->lmacid, BGX_CMRX_CFG);
+	tx_en = cmr_cfg & CMR_PKT_TX_EN;
+	rx_en = cmr_cfg & CMR_PKT_RX_EN;
+	cmr_cfg &= ~(CMR_PKT_RX_EN | CMR_PKT_TX_EN);
+	bgx_reg_write(bgx, lmac->lmacid, BGX_CMRX_CFG, cmr_cfg);
+
+	/* Wait for BGX RX to be idle */
+	if (bgx_poll_reg(bgx, lmac->lmacid, BGX_GMP_GMI_PRTX_CFG,
+			 GMI_PORT_CFG_RX_IDLE, false)) {
+		dev_err(&bgx->pdev->dev, "BGX%d LMAC%d GMI RX not idle\n",
+			bgx->bgx_id, lmac->lmacid);
+		return;
+	}
+
+	/* Wait for BGX TX to be idle */
+	if (bgx_poll_reg(bgx, lmac->lmacid, BGX_GMP_GMI_PRTX_CFG,
+			 GMI_PORT_CFG_TX_IDLE, false)) {
+		dev_err(&bgx->pdev->dev, "BGX%d LMAC%d GMI TX not idle\n",
+			bgx->bgx_id, lmac->lmacid);
+		return;
+	}
+
+	port_cfg = bgx_reg_read(bgx, lmac->lmacid, BGX_GMP_GMI_PRTX_CFG);
+	misc_ctl = bgx_reg_read(bgx, lmac->lmacid, BGX_GMP_PCS_MISCX_CTL);
+
+	if (lmac->link_up) {
+		misc_ctl &= ~PCS_MISC_CTL_GMX_ENO;
+		port_cfg &= ~GMI_PORT_CFG_DUPLEX;
+		port_cfg |=  (lmac->last_duplex << 2);
+	} else {
+		misc_ctl |= PCS_MISC_CTL_GMX_ENO;
+	}
+
+	switch (lmac->last_speed) {
+	case 10:
+		port_cfg &= ~GMI_PORT_CFG_SPEED; /* speed 0 */
+		port_cfg |= GMI_PORT_CFG_SPEED_MSB;  /* speed_msb 1 */
+		port_cfg &= ~GMI_PORT_CFG_SLOT_TIME; /* slottime 0 */
+		misc_ctl &= ~PCS_MISC_CTL_SAMP_PT_MASK;
+		misc_ctl |= 50; /* samp_pt */
+		bgx_reg_write(bgx, lmac->lmacid, BGX_GMP_GMI_TXX_SLOT, 64);
+		bgx_reg_write(bgx, lmac->lmacid, BGX_GMP_GMI_TXX_BURST, 0);
+		break;
+	case 100:
+		port_cfg &= ~GMI_PORT_CFG_SPEED; /* speed 0 */
+		port_cfg &= ~GMI_PORT_CFG_SPEED_MSB; /* speed_msb 0 */
+		port_cfg &= ~GMI_PORT_CFG_SLOT_TIME; /* slottime 0 */
+		misc_ctl &= ~PCS_MISC_CTL_SAMP_PT_MASK;
+		misc_ctl |= 5; /* samp_pt */
+		bgx_reg_write(bgx, lmac->lmacid, BGX_GMP_GMI_TXX_SLOT, 64);
+		bgx_reg_write(bgx, lmac->lmacid, BGX_GMP_GMI_TXX_BURST, 0);
+		break;
+	case 1000:
+		port_cfg |= GMI_PORT_CFG_SPEED; /* speed 1 */
+		port_cfg &= ~GMI_PORT_CFG_SPEED_MSB; /* speed_msb 0 */
+		port_cfg |= GMI_PORT_CFG_SLOT_TIME; /* slottime 1 */
+		misc_ctl &= ~PCS_MISC_CTL_SAMP_PT_MASK;
+		misc_ctl |= 1; /* samp_pt */
+		bgx_reg_write(bgx, lmac->lmacid, BGX_GMP_GMI_TXX_SLOT, 512);
+		if (lmac->last_duplex)
+			bgx_reg_write(bgx, lmac->lmacid,
+				      BGX_GMP_GMI_TXX_BURST, 0);
+		else
+			bgx_reg_write(bgx, lmac->lmacid,
+				      BGX_GMP_GMI_TXX_BURST, 8192);
+		break;
+	default:
+		break;
+	}
+	bgx_reg_write(bgx, lmac->lmacid, BGX_GMP_PCS_MISCX_CTL, misc_ctl);
+	bgx_reg_write(bgx, lmac->lmacid, BGX_GMP_GMI_PRTX_CFG, port_cfg);
+
+	/* Restore CMR config settings */
+	cmr_cfg |= (rx_en ? CMR_PKT_RX_EN : 0) | (tx_en ? CMR_PKT_TX_EN : 0);
+	bgx_reg_write(bgx, lmac->lmacid, BGX_CMRX_CFG, cmr_cfg);
+
+	if (bgx->is_rgx && (cmr_cfg & (CMR_PKT_RX_EN | CMR_PKT_TX_EN)))
+		xcv_setup_link(lmac->link_up, lmac->last_speed);
+}
+
+static void bgx_lmac_handler(struct net_device *netdev)
+{
+	struct lmac *lmac = container_of(netdev, struct lmac, netdev);
+	struct phy_device *phydev;
+	int link_changed = 0;
+
+	phydev = lmac->phydev;
+
+	if (!phydev->link && lmac->last_link)
+		link_changed = -1;
+
+	if (phydev->link &&
+	    (lmac->last_duplex != phydev->duplex ||
+	     lmac->last_link != phydev->link ||
+	     lmac->last_speed != phydev->speed)) {
+			link_changed = 1;
+	}
+
+	lmac->last_link = phydev->link;
+	lmac->last_speed = phydev->speed;
+	lmac->last_duplex = phydev->duplex;
+
+	if (!link_changed)
+		return;
+
+	if (link_changed > 0)
+		lmac->link_up = true;
+	else
+		lmac->link_up = false;
+
+	if (lmac->is_sgmii)
+		bgx_sgmii_change_link_state(lmac);
+	else
+		bgx_xaui_check_link(lmac);
+}
+
+u64 bgx_get_rx_stats(int node, int bgx_idx, int lmac, int idx)
+{
+	struct bgx *bgx;
+
+	bgx = get_bgx(node, bgx_idx);
+	if (!bgx)
+		return 0;
+
+	if (idx > 8)
+		lmac = 0;
+	return bgx_reg_read(bgx, lmac, BGX_CMRX_RX_STAT0 + (idx * 8));
+}
+EXPORT_SYMBOL(bgx_get_rx_stats);
+
+u64 bgx_get_tx_stats(int node, int bgx_idx, int lmac, int idx)
+{
+	struct bgx *bgx;
+
+	bgx = get_bgx(node, bgx_idx);
+	if (!bgx)
+		return 0;
+
+	return bgx_reg_read(bgx, lmac, BGX_CMRX_TX_STAT0 + (idx * 8));
+}
+EXPORT_SYMBOL(bgx_get_tx_stats);
+
+/* 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 = get_bgx(node, 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, struct lmac *lmac)
+{
+	int lmacid = lmac->lmacid;
+	u64 cfg;
+
+	bgx_reg_modify(bgx, lmacid, BGX_GMP_GMI_TXX_THRESH, 0x30);
+	/* max packet size */
+	bgx_reg_modify(bgx, lmacid, BGX_GMP_GMI_RXX_JABBER, MAX_FRAME_SIZE);
+
+	/* Disable frame alignment if using preamble */
+	cfg = bgx_reg_read(bgx, lmacid, BGX_GMP_GMI_TXX_APPEND);
+	if (cfg & 1)
+		bgx_reg_write(bgx, lmacid, BGX_GMP_GMI_TXX_SGMII_CTL, 0);
+
+	/* Enable lmac */
+	bgx_reg_modify(bgx, lmacid, BGX_CMRX_CFG, CMR_EN);
+
+	/* PCS reset */
+	bgx_reg_modify(bgx, lmacid, BGX_GMP_PCS_MRX_CTL, PCS_MRX_CTL_RESET);
+	if (bgx_poll_reg(bgx, lmacid, BGX_GMP_PCS_MRX_CTL,
+			 PCS_MRX_CTL_RESET, true)) {
+		dev_err(&bgx->pdev->dev, "BGX PCS reset not completed\n");
+		return -1;
+	}
+
+	/* power down, reset autoneg, autoneg enable */
+	cfg = bgx_reg_read(bgx, lmacid, BGX_GMP_PCS_MRX_CTL);
+	cfg &= ~PCS_MRX_CTL_PWR_DN;
+	cfg |= PCS_MRX_CTL_RST_AN;
+	if (lmac->phydev) {
+		cfg |= PCS_MRX_CTL_AN_EN;
+	} else {
+		/* In scenarios where PHY driver is not present or it's a
+		 * non-standard PHY, FW sets AN_EN to inform Linux driver
+		 * to do auto-neg and link polling or not.
+		 */
+		if (cfg & PCS_MRX_CTL_AN_EN)
+			lmac->autoneg = true;
+	}
+	bgx_reg_write(bgx, lmacid, BGX_GMP_PCS_MRX_CTL, cfg);
+
+	if (lmac->lmac_type == BGX_MODE_QSGMII) {
+		/* Disable disparity check for QSGMII */
+		cfg = bgx_reg_read(bgx, lmacid, BGX_GMP_PCS_MISCX_CTL);
+		cfg &= ~PCS_MISC_CTL_DISP_EN;
+		bgx_reg_write(bgx, lmacid, BGX_GMP_PCS_MISCX_CTL, cfg);
+		return 0;
+	}
+
+	if ((lmac->lmac_type == BGX_MODE_SGMII) && lmac->phydev) {
+		if (bgx_poll_reg(bgx, lmacid, BGX_GMP_PCS_MRX_STATUS,
+				 PCS_MRX_STATUS_AN_CPT, false)) {
+			dev_err(&bgx->pdev->dev, "BGX AN_CPT not completed\n");
+			return -1;
+		}
+	}
+
+	return 0;
+}
+
+static int bgx_lmac_xaui_init(struct bgx *bgx, struct lmac *lmac)
+{
+	u64 cfg;
+	int lmacid = lmac->lmacid;
+
+	/* Reset SPU */
+	bgx_reg_modify(bgx, lmacid, BGX_SPUX_CONTROL1, SPU_CTL_RESET);
+	if (bgx_poll_reg(bgx, lmacid, BGX_SPUX_CONTROL1, SPU_CTL_RESET, true)) {
+		dev_err(&bgx->pdev->dev, "BGX SPU reset not completed\n");
+		return -1;
+	}
+
+	/* Disable LMAC */
+	cfg = bgx_reg_read(bgx, lmacid, BGX_CMRX_CFG);
+	cfg &= ~CMR_EN;
+	bgx_reg_write(bgx, lmacid, BGX_CMRX_CFG, cfg);
+
+	bgx_reg_modify(bgx, lmacid, BGX_SPUX_CONTROL1, SPU_CTL_LOW_POWER);
+	/* Set interleaved running disparity for RXAUI */
+	if (lmac->lmac_type == BGX_MODE_RXAUI)
+		bgx_reg_modify(bgx, lmacid, BGX_SPUX_MISC_CONTROL,
+			       SPU_MISC_CTL_INTLV_RDISP);
+
+	/* Clear receive packet disable */
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SPUX_MISC_CONTROL);
+	cfg &= ~SPU_MISC_CTL_RX_DIS;
+	bgx_reg_write(bgx, lmacid, BGX_SPUX_MISC_CONTROL, cfg);
+
+	/* clear all interrupts */
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SMUX_RX_INT);
+	bgx_reg_write(bgx, lmacid, BGX_SMUX_RX_INT, cfg);
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SMUX_TX_INT);
+	bgx_reg_write(bgx, lmacid, BGX_SMUX_TX_INT, cfg);
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SPUX_INT);
+	bgx_reg_write(bgx, lmacid, BGX_SPUX_INT, cfg);
+
+	if (lmac->use_training) {
+		bgx_reg_write(bgx, lmacid, BGX_SPUX_BR_PMD_LP_CUP, 0x00);
+		bgx_reg_write(bgx, lmacid, BGX_SPUX_BR_PMD_LD_CUP, 0x00);
+		bgx_reg_write(bgx, lmacid, BGX_SPUX_BR_PMD_LD_REP, 0x00);
+		/* training enable */
+		bgx_reg_modify(bgx, lmacid,
+			       BGX_SPUX_BR_PMD_CRTL, SPU_PMD_CRTL_TRAIN_EN);
+	}
+
+	/* Append FCS to each packet */
+	bgx_reg_modify(bgx, lmacid, BGX_SMUX_TX_APPEND, SMU_TX_APPEND_FCS_D);
+
+	/* Disable forward error correction */
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SPUX_FEC_CONTROL);
+	cfg &= ~SPU_FEC_CTL_FEC_EN;
+	bgx_reg_write(bgx, lmacid, BGX_SPUX_FEC_CONTROL, cfg);
+
+	/* Disable autoneg */
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SPUX_AN_CONTROL);
+	cfg = cfg & ~(SPU_AN_CTL_AN_EN | SPU_AN_CTL_XNP_EN);
+	bgx_reg_write(bgx, lmacid, BGX_SPUX_AN_CONTROL, cfg);
+
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SPUX_AN_ADV);
+	if (lmac->lmac_type == BGX_MODE_10G_KR)
+		cfg |= (1 << 23);
+	else if (lmac->lmac_type == BGX_MODE_40G_KR)
+		cfg |= (1 << 24);
+	else
+		cfg &= ~((1 << 23) | (1 << 24));
+	cfg = cfg & (~((1ULL << 25) | (1ULL << 22) | (1ULL << 12)));
+	bgx_reg_write(bgx, lmacid, BGX_SPUX_AN_ADV, cfg);
+
+	cfg = bgx_reg_read(bgx, 0, BGX_SPU_DBG_CONTROL);
+	cfg &= ~SPU_DBG_CTL_AN_ARB_LINK_CHK_EN;
+	bgx_reg_write(bgx, 0, BGX_SPU_DBG_CONTROL, cfg);
+
+	/* Enable lmac */
+	bgx_reg_modify(bgx, lmacid, BGX_CMRX_CFG, CMR_EN);
+
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SPUX_CONTROL1);
+	cfg &= ~SPU_CTL_LOW_POWER;
+	bgx_reg_write(bgx, lmacid, BGX_SPUX_CONTROL1, cfg);
+
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SMUX_TX_CTL);
+	cfg &= ~SMU_TX_CTL_UNI_EN;
+	cfg |= SMU_TX_CTL_DIC_EN;
+	bgx_reg_write(bgx, lmacid, BGX_SMUX_TX_CTL, cfg);
+
+	/* Enable receive and transmission of pause frames */
+	bgx_reg_write(bgx, lmacid, BGX_SMUX_CBFC_CTL, ((0xffffULL << 32) |
+		      BCK_EN | DRP_EN | TX_EN | RX_EN));
+	/* Configure pause time and interval */
+	bgx_reg_write(bgx, lmacid,
+		      BGX_SMUX_TX_PAUSE_PKT_TIME, DEFAULT_PAUSE_TIME);
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SMUX_TX_PAUSE_PKT_INTERVAL);
+	cfg &= ~0xFFFFull;
+	bgx_reg_write(bgx, lmacid, BGX_SMUX_TX_PAUSE_PKT_INTERVAL,
+		      cfg | (DEFAULT_PAUSE_TIME - 0x1000));
+	bgx_reg_write(bgx, lmacid, BGX_SMUX_TX_PAUSE_ZERO, 0x01);
+
+	/* take lmac_count into account */
+	bgx_reg_modify(bgx, lmacid, BGX_SMUX_TX_THRESH, (0x100 - 1));
+	/* max packet size */
+	bgx_reg_modify(bgx, lmacid, BGX_SMUX_RX_JABBER, MAX_FRAME_SIZE);
+
+	return 0;
+}
+
+static int bgx_xaui_check_link(struct lmac *lmac)
+{
+	struct bgx *bgx = lmac->bgx;
+	int lmacid = lmac->lmacid;
+	int lmac_type = lmac->lmac_type;
+	u64 cfg;
+
+	if (lmac->use_training) {
+		cfg = bgx_reg_read(bgx, lmacid, BGX_SPUX_INT);
+		if (!(cfg & (1ull << 13))) {
+			cfg = (1ull << 13) | (1ull << 14);
+			bgx_reg_write(bgx, lmacid, BGX_SPUX_INT, cfg);
+			cfg = bgx_reg_read(bgx, lmacid, BGX_SPUX_BR_PMD_CRTL);
+			cfg |= (1ull << 0);
+			bgx_reg_write(bgx, lmacid, BGX_SPUX_BR_PMD_CRTL, cfg);
+			return -1;
+		}
+	}
+
+	/* wait for PCS to come out of reset */
+	if (bgx_poll_reg(bgx, lmacid, BGX_SPUX_CONTROL1, SPU_CTL_RESET, true)) {
+		dev_err(&bgx->pdev->dev, "BGX SPU reset not completed\n");
+		return -1;
+	}
+
+	if ((lmac_type == BGX_MODE_10G_KR) || (lmac_type == BGX_MODE_XFI) ||
+	    (lmac_type == BGX_MODE_40G_KR) || (lmac_type == BGX_MODE_XLAUI)) {
+		if (bgx_poll_reg(bgx, lmacid, BGX_SPUX_BR_STATUS1,
+				 SPU_BR_STATUS_BLK_LOCK, false)) {
+			dev_err(&bgx->pdev->dev,
+				"SPU_BR_STATUS_BLK_LOCK not completed\n");
+			return -1;
+		}
+	} else {
+		if (bgx_poll_reg(bgx, lmacid, BGX_SPUX_BX_STATUS,
+				 SPU_BX_STATUS_RX_ALIGN, false)) {
+			dev_err(&bgx->pdev->dev,
+				"SPU_BX_STATUS_RX_ALIGN not completed\n");
+			return -1;
+		}
+	}
+
+	/* Clear rcvflt bit (latching high) and read it back */
+	if (bgx_reg_read(bgx, lmacid, BGX_SPUX_STATUS2) & SPU_STATUS2_RCVFLT)
+		bgx_reg_modify(bgx, lmacid,
+			       BGX_SPUX_STATUS2, SPU_STATUS2_RCVFLT);
+	if (bgx_reg_read(bgx, lmacid, BGX_SPUX_STATUS2) & SPU_STATUS2_RCVFLT) {
+		dev_err(&bgx->pdev->dev, "Receive fault, retry training\n");
+		if (lmac->use_training) {
+			cfg = bgx_reg_read(bgx, lmacid, BGX_SPUX_INT);
+			if (!(cfg & (1ull << 13))) {
+				cfg = (1ull << 13) | (1ull << 14);
+				bgx_reg_write(bgx, lmacid, BGX_SPUX_INT, cfg);
+				cfg = bgx_reg_read(bgx, lmacid,
+						   BGX_SPUX_BR_PMD_CRTL);
+				cfg |= (1ull << 0);
+				bgx_reg_write(bgx, lmacid,
+					      BGX_SPUX_BR_PMD_CRTL, cfg);
+				return -1;
+			}
+		}
+		return -1;
+	}
+
+	/* Wait for BGX RX to be idle */
+	if (bgx_poll_reg(bgx, lmacid, BGX_SMUX_CTL, SMU_CTL_RX_IDLE, false)) {
+		dev_err(&bgx->pdev->dev, "SMU RX not idle\n");
+		return -1;
+	}
+
+	/* Wait for BGX TX to be idle */
+	if (bgx_poll_reg(bgx, lmacid, BGX_SMUX_CTL, SMU_CTL_TX_IDLE, false)) {
+		dev_err(&bgx->pdev->dev, "SMU TX not idle\n");
+		return -1;
+	}
+
+	/* Check for MAC RX faults */
+	cfg = bgx_reg_read(bgx, lmacid, BGX_SMUX_RX_CTL);
+	/* 0 - Link is okay, 1 - Local fault, 2 - Remote fault */
+	cfg &= SMU_RX_CTL_STATUS;
+	if (!cfg)
+		return 0;
+
+	/* Rx local/remote fault seen.
+	 * Do lmac reinit to see if condition recovers
+	 */
+	bgx_lmac_xaui_init(bgx, lmac);
+
+	return -1;
+}
+
+static void bgx_poll_for_sgmii_link(struct lmac *lmac)
+{
+	u64 pcs_link, an_result;
+	u8 speed;
+
+	pcs_link = bgx_reg_read(lmac->bgx, lmac->lmacid,
+				BGX_GMP_PCS_MRX_STATUS);
+
+	/*Link state bit is sticky, read it again*/
+	if (!(pcs_link & PCS_MRX_STATUS_LINK))
+		pcs_link = bgx_reg_read(lmac->bgx, lmac->lmacid,
+					BGX_GMP_PCS_MRX_STATUS);
+
+	if (bgx_poll_reg(lmac->bgx, lmac->lmacid, BGX_GMP_PCS_MRX_STATUS,
+			 PCS_MRX_STATUS_AN_CPT, false)) {
+		lmac->link_up = false;
+		lmac->last_speed = SPEED_UNKNOWN;
+		lmac->last_duplex = DUPLEX_UNKNOWN;
+		goto next_poll;
+	}
+
+	lmac->link_up = ((pcs_link & PCS_MRX_STATUS_LINK) != 0) ? true : false;
+	an_result = bgx_reg_read(lmac->bgx, lmac->lmacid,
+				 BGX_GMP_PCS_ANX_AN_RESULTS);
+
+	speed = (an_result >> 3) & 0x3;
+	lmac->last_duplex = (an_result >> 1) & 0x1;
+	switch (speed) {
+	case 0:
+		lmac->last_speed = SPEED_10;
+		break;
+	case 1:
+		lmac->last_speed = SPEED_100;
+		break;
+	case 2:
+		lmac->last_speed = SPEED_1000;
+		break;
+	default:
+		lmac->link_up = false;
+		lmac->last_speed = SPEED_UNKNOWN;
+		lmac->last_duplex = DUPLEX_UNKNOWN;
+		break;
+	}
+
+next_poll:
+
+	if (lmac->last_link != lmac->link_up) {
+		if (lmac->link_up)
+			bgx_sgmii_change_link_state(lmac);
+		lmac->last_link = lmac->link_up;
+	}
+
+	queue_delayed_work(lmac->check_link, &lmac->dwork, HZ * 3);
+}
+
+static void bgx_poll_for_link(struct work_struct *work)
+{
+	struct lmac *lmac;
+	u64 spu_link, smu_link;
+
+	lmac = container_of(work, struct lmac, dwork.work);
+	if (lmac->is_sgmii) {
+		bgx_poll_for_sgmii_link(lmac);
+		return;
+	}
+
+	/* Receive link is latching low. Force it high and verify it */
+	bgx_reg_modify(lmac->bgx, lmac->lmacid,
+		       BGX_SPUX_STATUS1, SPU_STATUS1_RCV_LNK);
+	bgx_poll_reg(lmac->bgx, lmac->lmacid, BGX_SPUX_STATUS1,
+		     SPU_STATUS1_RCV_LNK, false);
+
+	spu_link = bgx_reg_read(lmac->bgx, lmac->lmacid, BGX_SPUX_STATUS1);
+	smu_link = bgx_reg_read(lmac->bgx, lmac->lmacid, BGX_SMUX_RX_CTL);
+
+	if ((spu_link & SPU_STATUS1_RCV_LNK) &&
+	    !(smu_link & SMU_RX_CTL_STATUS)) {
+		lmac->link_up = true;
+		if (lmac->lmac_type == BGX_MODE_XLAUI)
+			lmac->last_speed = SPEED_40000;
+		else
+			lmac->last_speed = SPEED_10000;
+		lmac->last_duplex = DUPLEX_FULL;
+	} else {
+		lmac->link_up = false;
+		lmac->last_speed = SPEED_UNKNOWN;
+		lmac->last_duplex = DUPLEX_UNKNOWN;
+	}
+
+	if (lmac->last_link != lmac->link_up) {
+		if (lmac->link_up) {
+			if (bgx_xaui_check_link(lmac)) {
+				/* Errors, clear link_up state */
+				lmac->link_up = false;
+				lmac->last_speed = SPEED_UNKNOWN;
+				lmac->last_duplex = DUPLEX_UNKNOWN;
+			}
+		}
+		lmac->last_link = lmac->link_up;
+	}
+
+	queue_delayed_work(lmac->check_link, &lmac->dwork, HZ * 2);
+}
+
+static int phy_interface_mode(u8 lmac_type)
+{
+	if (lmac_type == BGX_MODE_QSGMII)
+		return PHY_INTERFACE_MODE_QSGMII;
+	if (lmac_type == BGX_MODE_RGMII)
+		return PHY_INTERFACE_MODE_RGMII_RXID;
+
+	return PHY_INTERFACE_MODE_SGMII;
+}
+
+static int bgx_lmac_enable(struct bgx *bgx, u8 lmacid)
+{
+	struct lmac *lmac;
+	u64 cfg;
+
+	lmac = &bgx->lmac[lmacid];
+	lmac->bgx = bgx;
+
+	if ((lmac->lmac_type == BGX_MODE_SGMII) ||
+	    (lmac->lmac_type == BGX_MODE_QSGMII) ||
+	    (lmac->lmac_type == BGX_MODE_RGMII)) {
+		lmac->is_sgmii = true;
+		if (bgx_lmac_sgmii_init(bgx, lmac))
+			return -1;
+	} else {
+		lmac->is_sgmii = false;
+		if (bgx_lmac_xaui_init(bgx, lmac))
+			return -1;
+	}
+
+	if (lmac->is_sgmii) {
+		cfg = bgx_reg_read(bgx, lmacid, BGX_GMP_GMI_TXX_APPEND);
+		cfg |= ((1ull << 2) | (1ull << 1)); /* FCS and PAD */
+		bgx_reg_modify(bgx, lmacid, BGX_GMP_GMI_TXX_APPEND, cfg);
+		bgx_reg_write(bgx, lmacid, BGX_GMP_GMI_TXX_MIN_PKT, 60 - 1);
+	} else {
+		cfg = bgx_reg_read(bgx, lmacid, BGX_SMUX_TX_APPEND);
+		cfg |= ((1ull << 2) | (1ull << 1)); /* FCS and PAD */
+		bgx_reg_modify(bgx, lmacid, BGX_SMUX_TX_APPEND, cfg);
+		bgx_reg_write(bgx, lmacid, BGX_SMUX_TX_MIN_PKT, 60 + 4);
+	}
+
+	/* actual number of filters available to exact LMAC */
+	lmac->dmacs_count = (RX_DMAC_COUNT / bgx->lmac_count);
+	lmac->dmacs = kcalloc(lmac->dmacs_count, sizeof(*lmac->dmacs),
+			      GFP_KERNEL);
+	if (!lmac->dmacs)
+		return -ENOMEM;
+
+	/* Enable lmac */
+	bgx_reg_modify(bgx, lmacid, BGX_CMRX_CFG, CMR_EN);
+
+	/* Restore default cfg, incase low level firmware changed it */
+	bgx_reg_write(bgx, lmacid, BGX_CMRX_RX_DMAC_CTL, 0x03);
+
+	if ((lmac->lmac_type != BGX_MODE_XFI) &&
+	    (lmac->lmac_type != BGX_MODE_XLAUI) &&
+	    (lmac->lmac_type != BGX_MODE_40G_KR) &&
+	    (lmac->lmac_type != BGX_MODE_10G_KR)) {
+		if (!lmac->phydev) {
+			if (lmac->autoneg) {
+				bgx_reg_write(bgx, lmacid,
+					      BGX_GMP_PCS_LINKX_TIMER,
+					      PCS_LINKX_TIMER_COUNT);
+				goto poll;
+			} else {
+				/* Default to below link speed and duplex */
+				lmac->link_up = true;
+				lmac->last_speed = SPEED_1000;
+				lmac->last_duplex = DUPLEX_FULL;
+				bgx_sgmii_change_link_state(lmac);
+				return 0;
+			}
+		}
+		lmac->phydev->dev_flags = 0;
+
+		if (phy_connect_direct(&lmac->netdev, lmac->phydev,
+				       bgx_lmac_handler,
+				       phy_interface_mode(lmac->lmac_type)))
+			return -ENODEV;
+
+		phy_start(lmac->phydev);
+		return 0;
+	}
+
+poll:
+	lmac->check_link = alloc_workqueue("check_link", WQ_UNBOUND |
+					   WQ_MEM_RECLAIM, 1);
+	if (!lmac->check_link)
+		return -ENOMEM;
+	INIT_DELAYED_WORK(&lmac->dwork, bgx_poll_for_link);
+	queue_delayed_work(lmac->check_link, &lmac->dwork, 0);
+
+	return 0;
+}
+
+static void bgx_lmac_disable(struct bgx *bgx, u8 lmacid)
+{
+	struct lmac *lmac;
+	u64 cfg;
+
+	lmac = &bgx->lmac[lmacid];
+	if (lmac->check_link) {
+		/* Destroy work queue */
+		cancel_delayed_work_sync(&lmac->dwork);
+		destroy_workqueue(lmac->check_link);
+	}
+
+	/* Disable packet reception */
+	cfg = bgx_reg_read(bgx, lmacid, BGX_CMRX_CFG);
+	cfg &= ~CMR_PKT_RX_EN;
+	bgx_reg_write(bgx, lmacid, BGX_CMRX_CFG, cfg);
+
+	/* Give chance for Rx/Tx FIFO to get drained */
+	bgx_poll_reg(bgx, lmacid, BGX_CMRX_RX_FIFO_LEN, (u64)0x1FFF, true);
+	bgx_poll_reg(bgx, lmacid, BGX_CMRX_TX_FIFO_LEN, (u64)0x3FFF, true);
+
+	/* Disable packet transmission */
+	cfg = bgx_reg_read(bgx, lmacid, BGX_CMRX_CFG);
+	cfg &= ~CMR_PKT_TX_EN;
+	bgx_reg_write(bgx, lmacid, BGX_CMRX_CFG, cfg);
+
+	/* Disable serdes lanes */
+        if (!lmac->is_sgmii)
+                bgx_reg_modify(bgx, lmacid,
+                               BGX_SPUX_CONTROL1, SPU_CTL_LOW_POWER);
+        else
+                bgx_reg_modify(bgx, lmacid,
+                               BGX_GMP_PCS_MRX_CTL, PCS_MRX_CTL_PWR_DN);
+
+	/* Disable LMAC */
+	cfg = bgx_reg_read(bgx, lmacid, BGX_CMRX_CFG);
+	cfg &= ~CMR_EN;
+	bgx_reg_write(bgx, lmacid, BGX_CMRX_CFG, cfg);
+
+	bgx_flush_dmac_cam_filter(bgx, lmacid);
+	kfree(lmac->dmacs);
+
+	if ((lmac->lmac_type != BGX_MODE_XFI) &&
+	    (lmac->lmac_type != BGX_MODE_XLAUI) &&
+	    (lmac->lmac_type != BGX_MODE_40G_KR) &&
+	    (lmac->lmac_type != BGX_MODE_10G_KR) && lmac->phydev)
+		phy_disconnect(lmac->phydev);
+
+	lmac->phydev = NULL;
+}
+
+static void bgx_init_hw(struct bgx *bgx)
+{
+	int i;
+	struct lmac *lmac;
+
+	bgx_reg_modify(bgx, 0, BGX_CMR_GLOBAL_CFG, CMR_GLOBAL_CFG_FCS_STRIP);
+	if (bgx_reg_read(bgx, 0, BGX_CMR_BIST_STATUS))
+		dev_err(&bgx->pdev->dev, "BGX%d BIST failed\n", bgx->bgx_id);
+
+	/* Set lmac type and lane2serdes mapping */
+	for (i = 0; i < bgx->lmac_count; i++) {
+		lmac = &bgx->lmac[i];
+		bgx_reg_write(bgx, i, BGX_CMRX_CFG,
+			      (lmac->lmac_type << 8) | lmac->lane_to_sds);
+		bgx->lmac[i].lmacid_bd = lmac_count;
+		lmac_count++;
+	}
+
+	bgx_reg_write(bgx, 0, BGX_CMR_TX_LMACS, bgx->lmac_count);
+	bgx_reg_write(bgx, 0, BGX_CMR_RX_LMACS, bgx->lmac_count);
+
+	/* Set the backpressure AND mask */
+	for (i = 0; i < bgx->lmac_count; i++)
+		bgx_reg_modify(bgx, 0, BGX_CMR_CHAN_MSK_AND,
+			       ((1ULL << MAX_BGX_CHANS_PER_LMAC) - 1) <<
+			       (i * MAX_BGX_CHANS_PER_LMAC));
+
+	/* Disable all MAC filtering */
+	for (i = 0; i < RX_DMAC_COUNT; i++)
+		bgx_reg_write(bgx, 0, BGX_CMR_RX_DMACX_CAM + (i * 8), 0x00);
+
+	/* Disable MAC steering (NCSI traffic) */
+	for (i = 0; i < RX_TRAFFIC_STEER_RULE_COUNT; i++)
+		bgx_reg_write(bgx, 0, BGX_CMR_RX_STEERING + (i * 8), 0x00);
+}
+
+static u8 bgx_get_lane2sds_cfg(struct bgx *bgx, struct lmac *lmac)
+{
+	return (u8)(bgx_reg_read(bgx, lmac->lmacid, BGX_CMRX_CFG) & 0xFF);
+}
+
+static void bgx_print_qlm_mode(struct bgx *bgx, u8 lmacid)
+{
+	struct device *dev = &bgx->pdev->dev;
+	struct lmac *lmac;
+	char str[27];
+
+	if (!bgx->is_dlm && lmacid)
+		return;
+
+	lmac = &bgx->lmac[lmacid];
+	if (!bgx->is_dlm)
+		sprintf(str, "BGX%d QLM mode", bgx->bgx_id);
+	else
+		sprintf(str, "BGX%d LMAC%d mode", bgx->bgx_id, lmacid);
+
+	switch (lmac->lmac_type) {
+	case BGX_MODE_SGMII:
+		dev_info(dev, "%s: SGMII\n", (char *)str);
+		break;
+	case BGX_MODE_XAUI:
+		dev_info(dev, "%s: XAUI\n", (char *)str);
+		break;
+	case BGX_MODE_RXAUI:
+		dev_info(dev, "%s: RXAUI\n", (char *)str);
+		break;
+	case BGX_MODE_XFI:
+		if (!lmac->use_training)
+			dev_info(dev, "%s: XFI\n", (char *)str);
+		else
+			dev_info(dev, "%s: 10G_KR\n", (char *)str);
+		break;
+	case BGX_MODE_XLAUI:
+		if (!lmac->use_training)
+			dev_info(dev, "%s: XLAUI\n", (char *)str);
+		else
+			dev_info(dev, "%s: 40G_KR4\n", (char *)str);
+		break;
+	case BGX_MODE_QSGMII:
+		dev_info(dev, "%s: QSGMII\n", (char *)str);
+		break;
+	case BGX_MODE_RGMII:
+		dev_info(dev, "%s: RGMII\n", (char *)str);
+		break;
+	case BGX_MODE_INVALID:
+		/* Nothing to do */
+		break;
+	}
+}
+
+static void lmac_set_lane2sds(struct bgx *bgx, struct lmac *lmac)
+{
+	switch (lmac->lmac_type) {
+	case BGX_MODE_SGMII:
+	case BGX_MODE_XFI:
+		lmac->lane_to_sds = lmac->lmacid;
+		break;
+	case BGX_MODE_XAUI:
+	case BGX_MODE_XLAUI:
+	case BGX_MODE_RGMII:
+		lmac->lane_to_sds = 0xE4;
+		break;
+	case BGX_MODE_RXAUI:
+		lmac->lane_to_sds = (lmac->lmacid) ? 0xE : 0x4;
+		break;
+	case BGX_MODE_QSGMII:
+		/* There is no way to determine if DLM0/2 is QSGMII or
+		 * DLM1/3 is configured to QSGMII as bootloader will
+		 * configure all LMACs, so take whatever is configured
+		 * by low level firmware.
+		 */
+		lmac->lane_to_sds = bgx_get_lane2sds_cfg(bgx, lmac);
+		break;
+	default:
+		lmac->lane_to_sds = 0;
+		break;
+	}
+}
+
+static void lmac_set_training(struct bgx *bgx, struct lmac *lmac, int lmacid)
+{
+	if ((lmac->lmac_type != BGX_MODE_10G_KR) &&
+	    (lmac->lmac_type != BGX_MODE_40G_KR)) {
+		lmac->use_training = false;
+		return;
+	}
+
+	lmac->use_training = bgx_reg_read(bgx, lmacid, BGX_SPUX_BR_PMD_CRTL) &
+							SPU_PMD_CRTL_TRAIN_EN;
+}
+
+static void bgx_set_lmac_config(struct bgx *bgx, u8 idx)
+{
+	struct lmac *lmac;
+	u64 cmr_cfg;
+	u8 lmac_type;
+	u8 lane_to_sds;
+
+	lmac = &bgx->lmac[idx];
+
+	if (!bgx->is_dlm || bgx->is_rgx) {
+		/* Read LMAC0 type to figure out QLM mode
+		 * This is configured by low level firmware
+		 */
+		cmr_cfg = bgx_reg_read(bgx, 0, BGX_CMRX_CFG);
+		lmac->lmac_type = (cmr_cfg >> 8) & 0x07;
+		if (bgx->is_rgx)
+			lmac->lmac_type = BGX_MODE_RGMII;
+		lmac_set_training(bgx, lmac, 0);
+		lmac_set_lane2sds(bgx, lmac);
+		return;
+	}
+
+	/* For DLMs or SLMs on 80/81/83xx so many lane configurations
+	 * are possible and vary across boards. Also Kernel doesn't have
+	 * any way to identify board type/info and since firmware does,
+	 * just take lmac type and serdes lane config as is.
+	 */
+	cmr_cfg = bgx_reg_read(bgx, idx, BGX_CMRX_CFG);
+	lmac_type = (u8)((cmr_cfg >> 8) & 0x07);
+	lane_to_sds = (u8)(cmr_cfg & 0xFF);
+	/* Check if config is reset value */
+	if ((lmac_type == 0) && (lane_to_sds == 0xE4))
+		lmac->lmac_type = BGX_MODE_INVALID;
+	else
+		lmac->lmac_type = lmac_type;
+	lmac->lane_to_sds = lane_to_sds;
+	lmac_set_training(bgx, lmac, lmac->lmacid);
+}
+
+static void bgx_get_qlm_mode(struct bgx *bgx)
+{
+	struct lmac *lmac;
+	u8  idx;
+
+	/* Init all LMAC's type to invalid */
+	for (idx = 0; idx < bgx->max_lmac; idx++) {
+		lmac = &bgx->lmac[idx];
+		lmac->lmacid = idx;
+		lmac->lmac_type = BGX_MODE_INVALID;
+		lmac->use_training = false;
+	}
+
+	/* It is assumed that low level firmware sets this value */
+	bgx->lmac_count = bgx_reg_read(bgx, 0, BGX_CMR_RX_LMACS) & 0x7;
+	if (bgx->lmac_count > bgx->max_lmac)
+		bgx->lmac_count = bgx->max_lmac;
+
+	for (idx = 0; idx < bgx->lmac_count; idx++) {
+		bgx_set_lmac_config(bgx, idx);
+		bgx_print_qlm_mode(bgx, idx);
+	}
+}
+
+#ifdef CONFIG_ACPI
+
+static int acpi_get_mac_address(struct device *dev, struct acpi_device *adev,
+				u8 *dst)
+{
+	u8 mac[ETH_ALEN];
+	int ret;
+
+	ret = fwnode_get_mac_address(acpi_fwnode_handle(adev), mac);
+	if (ret) {
+		dev_err(dev, "MAC address invalid: %pM\n", mac);
+		return -EINVAL;
+	}
+
+	dev_info(dev, "MAC address set to: %pM\n", mac);
+
+	ether_addr_copy(dst, mac);
+	return 0;
+}
+
+/* Currently only sets the MAC address. */
+static acpi_status bgx_acpi_register_phy(acpi_handle handle,
+					 u32 lvl, void *context, void **rv)
+{
+	struct bgx *bgx = context;
+	struct device *dev = &bgx->pdev->dev;
+	struct acpi_device *adev;
+
+	adev = acpi_fetch_acpi_dev(handle);
+	if (!adev)
+		goto out;
+
+	acpi_get_mac_address(dev, adev, bgx->lmac[bgx->acpi_lmac_idx].mac);
+
+	SET_NETDEV_DEV(&bgx->lmac[bgx->acpi_lmac_idx].netdev, dev);
+
+	bgx->lmac[bgx->acpi_lmac_idx].lmacid = bgx->acpi_lmac_idx;
+	bgx->acpi_lmac_idx++; /* move to next LMAC */
+out:
+	return AE_OK;
+}
+
+static acpi_status bgx_acpi_match_id(acpi_handle handle, u32 lvl,
+				     void *context, void **ret_val)
+{
+	struct acpi_buffer string = { ACPI_ALLOCATE_BUFFER, NULL };
+	struct bgx *bgx = context;
+	char bgx_sel[5];
+
+	snprintf(bgx_sel, 5, "BGX%d", bgx->bgx_id);
+	if (ACPI_FAILURE(acpi_get_name(handle, ACPI_SINGLE_NAME, &string))) {
+		pr_warn("Invalid link device\n");
+		return AE_OK;
+	}
+
+	if (strncmp(string.pointer, bgx_sel, 4)) {
+		kfree(string.pointer);
+		return AE_OK;
+	}
+
+	acpi_walk_namespace(ACPI_TYPE_DEVICE, handle, 1,
+			    bgx_acpi_register_phy, NULL, bgx, NULL);
+
+	kfree(string.pointer);
+	return AE_CTRL_TERMINATE;
+}
+
+static int bgx_init_acpi_phy(struct bgx *bgx)
+{
+	acpi_get_devices(NULL, bgx_acpi_match_id, bgx, (void **)NULL);
+	return 0;
+}
+
+#else
+
+static int bgx_init_acpi_phy(struct bgx *bgx)
+{
+	return -ENODEV;
+}
+
+#endif /* CONFIG_ACPI */
+
+#if IS_ENABLED(CONFIG_OF_MDIO)
+
+static int bgx_init_of_phy(struct bgx *bgx)
+{
+	struct fwnode_handle *fwn;
+	struct device_node *node = NULL;
+	u8 lmac = 0;
+
+	device_for_each_child_node(&bgx->pdev->dev, fwn) {
+		struct phy_device *pd;
+		struct device_node *phy_np;
+
+		/* Should always be an OF node.  But if it is not, we
+		 * cannot handle it, so exit the loop.
+		 */
+		node = to_of_node(fwn);
+		if (!node)
+			break;
+
+		of_get_mac_address(node, bgx->lmac[lmac].mac);
+
+		SET_NETDEV_DEV(&bgx->lmac[lmac].netdev, &bgx->pdev->dev);
+		bgx->lmac[lmac].lmacid = lmac;
+
+		phy_np = of_parse_phandle(node, "phy-handle", 0);
+		/* If there is no phy or defective firmware presents
+		 * this cortina phy, for which there is no driver
+		 * support, ignore it.
+		 */
+		if (phy_np &&
+		    !of_device_is_compatible(phy_np, "cortina,cs4223-slice")) {
+			/* Wait until the phy drivers are available */
+			pd = of_phy_find_device(phy_np);
+			if (!pd)
+				goto defer;
+			bgx->lmac[lmac].phydev = pd;
+		}
+
+		lmac++;
+		if (lmac == bgx->max_lmac) {
+			of_node_put(node);
+			break;
+		}
+	}
+	return 0;
+
+defer:
+	/* We are bailing out, try not to leak device reference counts
+	 * for phy devices we may have already found.
+	 */
+	while (lmac) {
+		if (bgx->lmac[lmac].phydev) {
+			put_device(&bgx->lmac[lmac].phydev->mdio.dev);
+			bgx->lmac[lmac].phydev = NULL;
+		}
+		lmac--;
+	}
+	of_node_put(node);
+	return -EPROBE_DEFER;
+}
+
+#else
+
+static int bgx_init_of_phy(struct bgx *bgx)
+{
+	return -ENODEV;
+}
+
+#endif /* CONFIG_OF_MDIO */
+
+static int bgx_init_phy(struct bgx *bgx)
+{
+	if (!acpi_disabled)
+		return bgx_init_acpi_phy(bgx);
+
+	return bgx_init_of_phy(bgx);
+}
+
+static irqreturn_t bgx_intr_handler(int irq, void *data)
+{
+	struct bgx *bgx = (struct bgx *)data;
+	u64 status, val;
+	int lmac;
+
+	for (lmac = 0; lmac < bgx->lmac_count; lmac++) {
+		status = bgx_reg_read(bgx, lmac, BGX_GMP_GMI_TXX_INT);
+		if (status & GMI_TXX_INT_UNDFLW) {
+			pci_err(bgx->pdev, "BGX%d lmac%d UNDFLW\n",
+				bgx->bgx_id, lmac);
+			val = bgx_reg_read(bgx, lmac, BGX_CMRX_CFG);
+			val &= ~CMR_EN;
+			bgx_reg_write(bgx, lmac, BGX_CMRX_CFG, val);
+			val |= CMR_EN;
+			bgx_reg_write(bgx, lmac, BGX_CMRX_CFG, val);
+		}
+		/* clear interrupts */
+		bgx_reg_write(bgx, lmac, BGX_GMP_GMI_TXX_INT, status);
+	}
+
+	return IRQ_HANDLED;
+}
+
+static void bgx_register_intr(struct pci_dev *pdev)
+{
+	struct bgx *bgx = pci_get_drvdata(pdev);
+	int ret;
+
+	ret = pci_alloc_irq_vectors(pdev, BGX_LMAC_VEC_OFFSET,
+				    BGX_LMAC_VEC_OFFSET, PCI_IRQ_ALL_TYPES);
+	if (ret < 0) {
+		pci_err(pdev, "Req for #%d msix vectors failed\n",
+			BGX_LMAC_VEC_OFFSET);
+		return;
+	}
+	ret = pci_request_irq(pdev, GMPX_GMI_TX_INT, bgx_intr_handler, NULL,
+			      bgx, "BGX%d", bgx->bgx_id);
+	if (ret)
+		pci_free_irq(pdev, GMPX_GMI_TX_INT, bgx);
+}
+
+static int bgx_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	int err;
+	struct device *dev = &pdev->dev;
+	struct bgx *bgx = NULL;
+	u8 lmac;
+	u16 sdevid;
+
+	bgx = devm_kzalloc(dev, sizeof(*bgx), GFP_KERNEL);
+	if (!bgx)
+		return -ENOMEM;
+	bgx->pdev = pdev;
+
+	pci_set_drvdata(pdev, bgx);
+
+	err = pcim_enable_device(pdev);
+	if (err) {
+		pci_set_drvdata(pdev, NULL);
+		return dev_err_probe(dev, err, "Failed to enable PCI device\n");
+	}
+
+	err = pci_request_regions(pdev, DRV_NAME);
+	if (err) {
+		dev_err(dev, "PCI request regions failed 0x%x\n", err);
+		goto err_disable_device;
+	}
+
+	/* MAP configuration registers */
+	bgx->reg_base = pcim_iomap(pdev, PCI_CFG_REG_BAR_NUM, 0);
+	if (!bgx->reg_base) {
+		dev_err(dev, "BGX: Cannot map CSR memory space, aborting\n");
+		err = -ENOMEM;
+		goto err_release_regions;
+	}
+
+	set_max_bgx_per_node(pdev);
+
+	pci_read_config_word(pdev, PCI_DEVICE_ID, &sdevid);
+	if (sdevid != PCI_DEVICE_ID_THUNDER_RGX) {
+		bgx->bgx_id = (pci_resource_start(pdev,
+			PCI_CFG_REG_BAR_NUM) >> 24) & BGX_ID_MASK;
+		bgx->bgx_id += nic_get_node_id(pdev) * max_bgx_per_node;
+		bgx->max_lmac = MAX_LMAC_PER_BGX;
+		bgx_vnic[bgx->bgx_id] = bgx;
+	} else {
+		bgx->is_rgx = true;
+		bgx->max_lmac = 1;
+		bgx->bgx_id = MAX_BGX_PER_CN81XX - 1;
+		bgx_vnic[bgx->bgx_id] = bgx;
+		xcv_init_hw();
+	}
+
+	/* On 81xx all are DLMs and on 83xx there are 3 BGX QLMs and one
+	 * BGX i.e BGX2 can be split across 2 DLMs.
+	 */
+	pci_read_config_word(pdev, PCI_SUBSYSTEM_ID, &sdevid);
+	if ((sdevid == PCI_SUBSYS_DEVID_81XX_BGX) ||
+	    ((sdevid == PCI_SUBSYS_DEVID_83XX_BGX) && (bgx->bgx_id == 2)))
+		bgx->is_dlm = true;
+
+	bgx_get_qlm_mode(bgx);
+
+	err = bgx_init_phy(bgx);
+	if (err)
+		goto err_enable;
+
+	bgx_init_hw(bgx);
+
+	bgx_register_intr(pdev);
+
+	/* Enable all LMACs */
+	for (lmac = 0; lmac < bgx->lmac_count; lmac++) {
+		err = bgx_lmac_enable(bgx, lmac);
+		if (err) {
+			dev_err(dev, "BGX%d failed to enable lmac%d\n",
+				bgx->bgx_id, lmac);
+			while (lmac)
+				bgx_lmac_disable(bgx, --lmac);
+			goto err_enable;
+		}
+	}
+
+	return 0;
+
+err_enable:
+	bgx_vnic[bgx->bgx_id] = NULL;
+	pci_free_irq(pdev, GMPX_GMI_TX_INT, bgx);
+err_release_regions:
+	pci_release_regions(pdev);
+err_disable_device:
+	pci_disable_device(pdev);
+	pci_set_drvdata(pdev, NULL);
+	return err;
+}
+
+static void bgx_remove(struct pci_dev *pdev)
+{
+	struct bgx *bgx = pci_get_drvdata(pdev);
+	u8 lmac;
+
+	/* Disable all LMACs */
+	for (lmac = 0; lmac < bgx->lmac_count; lmac++)
+		bgx_lmac_disable(bgx, lmac);
+
+	pci_free_irq(pdev, GMPX_GMI_TX_INT, bgx);
+
+	bgx_vnic[bgx->bgx_id] = NULL;
+	pci_release_regions(pdev);
+	pci_disable_device(pdev);
+	pci_set_drvdata(pdev, NULL);
+}
+
+static struct pci_driver bgx_driver = {
+	.name = DRV_NAME,
+	.id_table = bgx_id_table,
+	.probe = bgx_probe,
+	.remove = bgx_remove,
+};
+
+static int __init bgx_init_module(void)
+{
+	pr_info("%s, ver %s\n", DRV_NAME, DRV_VERSION);
+
+	return pci_register_driver(&bgx_driver);
+}
+
+static void __exit bgx_cleanup_module(void)
+{
+	pci_unregister_driver(&bgx_driver);
+}
+
+module_init(bgx_init_module);
+module_exit(bgx_cleanup_module);
diff --git a/drivers/net/ethernet/cavium/thunder/thunder_bgx.h b/drivers/net/ethernet/cavium/thunder/thunder_bgx.h
new file mode 100644
index 000000000..cdea49392
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/thunder_bgx.h
@@ -0,0 +1,263 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (C) 2015 Cavium, Inc.
+ */
+
+#ifndef THUNDER_BGX_H
+#define THUNDER_BGX_H
+
+/* PCI device ID */
+#define	PCI_DEVICE_ID_THUNDER_BGX		0xA026
+#define	PCI_DEVICE_ID_THUNDER_RGX		0xA054
+
+/* Subsystem device IDs */
+#define PCI_SUBSYS_DEVID_88XX_BGX		0xA126
+#define PCI_SUBSYS_DEVID_81XX_BGX		0xA226
+#define PCI_SUBSYS_DEVID_81XX_RGX		0xA254
+#define PCI_SUBSYS_DEVID_83XX_BGX		0xA326
+
+#define    MAX_BGX_THUNDER			8 /* Max 2 nodes, 4 per node */
+#define    MAX_BGX_PER_CN88XX			2
+#define    MAX_BGX_PER_CN81XX			3 /* 2 BGXs + 1 RGX */
+#define    MAX_BGX_PER_CN83XX			4
+#define    MAX_LMAC_PER_BGX			4
+#define    MAX_BGX_CHANS_PER_LMAC		16
+#define    MAX_DMAC_PER_LMAC			8
+#define    MAX_FRAME_SIZE			9216
+#define    DEFAULT_PAUSE_TIME			0xFFFF
+
+#define	   BGX_ID_MASK				0x3
+#define	   LMAC_ID_MASK				0x3
+
+#define    MAX_DMAC_PER_LMAC_TNS_BYPASS_MODE	2
+
+/* Registers */
+#define BGX_CMRX_CFG			0x00
+#define  CMR_PKT_TX_EN				BIT_ULL(13)
+#define  CMR_PKT_RX_EN				BIT_ULL(14)
+#define  CMR_EN					BIT_ULL(15)
+#define BGX_CMR_GLOBAL_CFG		0x08
+#define  CMR_GLOBAL_CFG_FCS_STRIP		BIT_ULL(6)
+#define BGX_CMRX_RX_ID_MAP		0x60
+#define BGX_CMRX_RX_STAT0		0x70
+#define BGX_CMRX_RX_STAT1		0x78
+#define BGX_CMRX_RX_STAT2		0x80
+#define BGX_CMRX_RX_STAT3		0x88
+#define BGX_CMRX_RX_STAT4		0x90
+#define BGX_CMRX_RX_STAT5		0x98
+#define BGX_CMRX_RX_STAT6		0xA0
+#define BGX_CMRX_RX_STAT7		0xA8
+#define BGX_CMRX_RX_STAT8		0xB0
+#define BGX_CMRX_RX_STAT9		0xB8
+#define BGX_CMRX_RX_STAT10		0xC0
+#define BGX_CMRX_RX_BP_DROP		0xC8
+#define BGX_CMRX_RX_DMAC_CTL		0x0E8
+#define BGX_CMRX_RX_FIFO_LEN		0x108
+#define BGX_CMR_RX_DMACX_CAM		0x200
+#define  RX_DMACX_CAM_EN			BIT_ULL(48)
+#define  RX_DMACX_CAM_LMACID(x)			(((u64)x) << 49)
+#define  RX_DMAC_COUNT				32
+#define BGX_CMR_RX_STEERING		0x300
+#define  RX_TRAFFIC_STEER_RULE_COUNT		8
+#define BGX_CMR_CHAN_MSK_AND		0x450
+#define BGX_CMR_BIST_STATUS		0x460
+#define BGX_CMR_RX_LMACS		0x468
+#define BGX_CMRX_TX_FIFO_LEN		0x518
+#define BGX_CMRX_TX_STAT0		0x600
+#define BGX_CMRX_TX_STAT1		0x608
+#define BGX_CMRX_TX_STAT2		0x610
+#define BGX_CMRX_TX_STAT3		0x618
+#define BGX_CMRX_TX_STAT4		0x620
+#define BGX_CMRX_TX_STAT5		0x628
+#define BGX_CMRX_TX_STAT6		0x630
+#define BGX_CMRX_TX_STAT7		0x638
+#define BGX_CMRX_TX_STAT8		0x640
+#define BGX_CMRX_TX_STAT9		0x648
+#define BGX_CMRX_TX_STAT10		0x650
+#define BGX_CMRX_TX_STAT11		0x658
+#define BGX_CMRX_TX_STAT12		0x660
+#define BGX_CMRX_TX_STAT13		0x668
+#define BGX_CMRX_TX_STAT14		0x670
+#define BGX_CMRX_TX_STAT15		0x678
+#define BGX_CMRX_TX_STAT16		0x680
+#define BGX_CMRX_TX_STAT17		0x688
+#define BGX_CMR_TX_LMACS		0x1000
+
+#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)
+#define BGX_SPUX_STATUS2		0x10020
+#define  SPU_STATUS2_RCVFLT			BIT_ULL(10)
+#define BGX_SPUX_BX_STATUS		0x10028
+#define  SPU_BX_STATUS_RX_ALIGN			BIT_ULL(12)
+#define BGX_SPUX_BR_STATUS1		0x10030
+#define  SPU_BR_STATUS_BLK_LOCK			BIT_ULL(0)
+#define  SPU_BR_STATUS_RCV_LNK			BIT_ULL(12)
+#define BGX_SPUX_BR_PMD_CRTL		0x10068
+#define  SPU_PMD_CRTL_TRAIN_EN			BIT_ULL(1)
+#define BGX_SPUX_BR_PMD_LP_CUP		0x10078
+#define BGX_SPUX_BR_PMD_LD_CUP		0x10088
+#define BGX_SPUX_BR_PMD_LD_REP		0x10090
+#define BGX_SPUX_FEC_CONTROL		0x100A0
+#define  SPU_FEC_CTL_FEC_EN			BIT_ULL(0)
+#define  SPU_FEC_CTL_ERR_EN			BIT_ULL(1)
+#define BGX_SPUX_AN_CONTROL		0x100C8
+#define  SPU_AN_CTL_AN_EN			BIT_ULL(12)
+#define  SPU_AN_CTL_XNP_EN			BIT_ULL(13)
+#define BGX_SPUX_AN_ADV			0x100D8
+#define BGX_SPUX_MISC_CONTROL		0x10218
+#define  SPU_MISC_CTL_INTLV_RDISP		BIT_ULL(10)
+#define  SPU_MISC_CTL_RX_DIS			BIT_ULL(12)
+#define BGX_SPUX_INT			0x10220	/* +(0..3) << 20 */
+#define BGX_SPUX_INT_W1S		0x10228
+#define BGX_SPUX_INT_ENA_W1C		0x10230
+#define BGX_SPUX_INT_ENA_W1S		0x10238
+#define BGX_SPU_DBG_CONTROL		0x10300
+#define  SPU_DBG_CTL_AN_ARB_LINK_CHK_EN		BIT_ULL(18)
+#define  SPU_DBG_CTL_AN_NONCE_MCT_DIS		BIT_ULL(29)
+
+#define BGX_SMUX_RX_INT			0x20000
+#define BGX_SMUX_RX_FRM_CTL		0x20020
+#define  BGX_PKT_RX_PTP_EN			BIT_ULL(12)
+#define BGX_SMUX_RX_JABBER		0x20030
+#define BGX_SMUX_RX_CTL			0x20048
+#define  SMU_RX_CTL_STATUS			(3ull << 0)
+#define BGX_SMUX_TX_APPEND		0x20100
+#define  SMU_TX_APPEND_FCS_D			BIT_ULL(2)
+#define BGX_SMUX_TX_PAUSE_PKT_TIME	0x20110
+#define BGX_SMUX_TX_MIN_PKT		0x20118
+#define BGX_SMUX_TX_PAUSE_PKT_INTERVAL	0x20120
+#define BGX_SMUX_TX_PAUSE_ZERO		0x20138
+#define BGX_SMUX_TX_INT			0x20140
+#define BGX_SMUX_TX_CTL			0x20178
+#define  SMU_TX_CTL_DIC_EN			BIT_ULL(0)
+#define  SMU_TX_CTL_UNI_EN			BIT_ULL(1)
+#define  SMU_TX_CTL_LNK_STATUS			(3ull << 4)
+#define BGX_SMUX_TX_THRESH		0x20180
+#define BGX_SMUX_CTL			0x20200
+#define  SMU_CTL_RX_IDLE			BIT_ULL(0)
+#define  SMU_CTL_TX_IDLE			BIT_ULL(1)
+#define	BGX_SMUX_CBFC_CTL		0x20218
+#define	RX_EN					BIT_ULL(0)
+#define	TX_EN					BIT_ULL(1)
+#define	BCK_EN					BIT_ULL(2)
+#define	DRP_EN					BIT_ULL(3)
+
+#define BGX_GMP_PCS_MRX_CTL		0x30000
+#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_LINK			BIT_ULL(2)
+#define	 PCS_MRX_STATUS_AN_CPT			BIT_ULL(5)
+#define BGX_GMP_PCS_ANX_ADV		0x30010
+#define BGX_GMP_PCS_ANX_AN_RESULTS	0x30020
+#define BGX_GMP_PCS_LINKX_TIMER		0x30040
+#define PCS_LINKX_TIMER_COUNT			0x1E84
+#define BGX_GMP_PCS_SGM_AN_ADV		0x30068
+#define BGX_GMP_PCS_MISCX_CTL		0x30078
+#define  PCS_MISC_CTL_MODE			BIT_ULL(8)
+#define  PCS_MISC_CTL_DISP_EN			BIT_ULL(13)
+#define  PCS_MISC_CTL_GMX_ENO			BIT_ULL(11)
+#define  PCS_MISC_CTL_SAMP_PT_MASK	0x7Full
+#define BGX_GMP_GMI_PRTX_CFG		0x38020
+#define  GMI_PORT_CFG_SPEED			BIT_ULL(1)
+#define  GMI_PORT_CFG_DUPLEX			BIT_ULL(2)
+#define  GMI_PORT_CFG_SLOT_TIME			BIT_ULL(3)
+#define  GMI_PORT_CFG_SPEED_MSB			BIT_ULL(8)
+#define  GMI_PORT_CFG_RX_IDLE			BIT_ULL(12)
+#define  GMI_PORT_CFG_TX_IDLE			BIT_ULL(13)
+#define BGX_GMP_GMI_RXX_FRM_CTL		0x38028
+#define BGX_GMP_GMI_RXX_JABBER		0x38038
+#define BGX_GMP_GMI_TXX_THRESH		0x38210
+#define BGX_GMP_GMI_TXX_APPEND		0x38218
+#define BGX_GMP_GMI_TXX_SLOT		0x38220
+#define BGX_GMP_GMI_TXX_BURST		0x38228
+#define BGX_GMP_GMI_TXX_MIN_PKT		0x38240
+#define BGX_GMP_GMI_TXX_SGMII_CTL	0x38300
+#define BGX_GMP_GMI_TXX_INT		0x38500
+#define BGX_GMP_GMI_TXX_INT_W1S		0x38508
+#define BGX_GMP_GMI_TXX_INT_ENA_W1C	0x38510
+#define BGX_GMP_GMI_TXX_INT_ENA_W1S	0x38518
+#define  GMI_TXX_INT_PTP_LOST			BIT_ULL(4)
+#define  GMI_TXX_INT_LATE_COL			BIT_ULL(3)
+#define  GMI_TXX_INT_XSDEF			BIT_ULL(2)
+#define  GMI_TXX_INT_XSCOL			BIT_ULL(1)
+#define  GMI_TXX_INT_UNDFLW			BIT_ULL(0)
+
+#define BGX_MSIX_VEC_0_29_ADDR		0x400000 /* +(0..29) << 4 */
+#define BGX_MSIX_VEC_0_29_CTL		0x400008
+#define BGX_MSIX_PBA_0			0x4F0000
+
+/* MSI-X interrupts */
+#define BGX_MSIX_VECTORS	30
+#define BGX_LMAC_VEC_OFFSET	7
+#define BGX_MSIX_VEC_SHIFT	4
+
+#define CMRX_INT		0
+#define SPUX_INT		1
+#define SMUX_RX_INT		2
+#define SMUX_TX_INT		3
+#define GMPX_PCS_INT		4
+#define GMPX_GMI_RX_INT		5
+#define GMPX_GMI_TX_INT		6
+#define CMR_MEM_INT		28
+#define SPU_MEM_INT		29
+
+#define LMAC_INTR_LINK_UP	BIT(0)
+#define LMAC_INTR_LINK_DOWN	BIT(1)
+
+#define BGX_XCAST_BCAST_ACCEPT  BIT(0)
+#define BGX_XCAST_MCAST_ACCEPT  BIT(1)
+#define BGX_XCAST_MCAST_FILTER  BIT(2)
+
+void bgx_set_dmac_cam_filter(int node, int bgx_idx, int lmacid, u64 mac, u8 vf);
+void bgx_reset_xcast_mode(int node, int bgx_idx, int lmacid, u8 vf);
+void bgx_set_xcast_mode(int node, int bgx_idx, int lmacid, u8 mode);
+void octeon_mdiobus_force_mod_depencency(void);
+void bgx_lmac_rx_tx_enable(int node, int bgx_idx, int lmacid, bool enable);
+void bgx_add_dmac_addr(u64 dmac, int node, int bgx_idx, int lmac);
+unsigned bgx_get_map(int node);
+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);
+void bgx_config_timestamping(int node, int bgx_idx, int lmacid, bool enable);
+void bgx_lmac_get_pfc(int node, int bgx_idx, int lmacid, void *pause);
+void bgx_lmac_set_pfc(int node, int bgx_idx, int lmacid, void *pause);
+
+void xcv_init_hw(void);
+void xcv_setup_link(bool link_up, int link_speed);
+
+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
+#define BGX_TX_STATS_COUNT 18
+
+struct bgx_stats {
+	u64 rx_stats[BGX_RX_STATS_COUNT];
+	u64 tx_stats[BGX_TX_STATS_COUNT];
+};
+
+enum LMAC_TYPE {
+	BGX_MODE_SGMII = 0, /* 1 lane, 1.250 Gbaud */
+	BGX_MODE_XAUI = 1,  /* 4 lanes, 3.125 Gbaud */
+	BGX_MODE_DXAUI = 1, /* 4 lanes, 6.250 Gbaud */
+	BGX_MODE_RXAUI = 2, /* 2 lanes, 6.250 Gbaud */
+	BGX_MODE_XFI = 3,   /* 1 lane, 10.3125 Gbaud */
+	BGX_MODE_XLAUI = 4, /* 4 lanes, 10.3125 Gbaud */
+	BGX_MODE_10G_KR = 3,/* 1 lane, 10.3125 Gbaud */
+	BGX_MODE_40G_KR = 4,/* 4 lanes, 10.3125 Gbaud */
+	BGX_MODE_RGMII = 5,
+	BGX_MODE_QSGMII = 6,
+	BGX_MODE_INVALID = 7,
+};
+
+#endif /* THUNDER_BGX_H */
diff --git a/drivers/net/ethernet/cavium/thunder/thunder_xcv.c b/drivers/net/ethernet/cavium/thunder/thunder_xcv.c
new file mode 100644
index 000000000..3ebb93792
--- /dev/null
+++ b/drivers/net/ethernet/cavium/thunder/thunder_xcv.c
@@ -0,0 +1,231 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2016 Cavium, Inc.
+ */
+
+#include <linux/acpi.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/phy.h>
+#include <linux/of.h>
+#include <linux/of_mdio.h>
+#include <linux/of_net.h>
+
+#include "nic.h"
+#include "thunder_bgx.h"
+
+#define DRV_NAME	"thunder_xcv"
+#define DRV_VERSION	"1.0"
+
+/* Register offsets */
+#define XCV_RESET		0x00
+#define   PORT_EN		BIT_ULL(63)
+#define   CLK_RESET		BIT_ULL(15)
+#define   DLL_RESET		BIT_ULL(11)
+#define   COMP_EN		BIT_ULL(7)
+#define   TX_PKT_RESET		BIT_ULL(3)
+#define   TX_DATA_RESET		BIT_ULL(2)
+#define   RX_PKT_RESET		BIT_ULL(1)
+#define   RX_DATA_RESET		BIT_ULL(0)
+#define XCV_DLL_CTL		0x10
+#define   CLKRX_BYP		BIT_ULL(23)
+#define   CLKTX_BYP		BIT_ULL(15)
+#define XCV_COMP_CTL		0x20
+#define   DRV_BYP		BIT_ULL(63)
+#define XCV_CTL			0x30
+#define XCV_INT			0x40
+#define XCV_INT_W1S		0x48
+#define XCV_INT_ENA_W1C		0x50
+#define XCV_INT_ENA_W1S		0x58
+#define XCV_INBND_STATUS	0x80
+#define XCV_BATCH_CRD_RET	0x100
+
+struct xcv {
+	void __iomem		*reg_base;
+	struct pci_dev		*pdev;
+};
+
+static struct xcv *xcv;
+
+/* Supported devices */
+static const struct pci_device_id xcv_id_table[] = {
+	{ PCI_DEVICE(PCI_VENDOR_ID_CAVIUM, 0xA056) },
+	{ 0, }  /* end of table */
+};
+
+MODULE_AUTHOR("Cavium Inc");
+MODULE_DESCRIPTION("Cavium Thunder RGX/XCV Driver");
+MODULE_LICENSE("GPL v2");
+MODULE_VERSION(DRV_VERSION);
+MODULE_DEVICE_TABLE(pci, xcv_id_table);
+
+void xcv_init_hw(void)
+{
+	u64  cfg;
+
+	/* Take DLL out of reset */
+	cfg = readq_relaxed(xcv->reg_base + XCV_RESET);
+	cfg &= ~DLL_RESET;
+	writeq_relaxed(cfg, xcv->reg_base + XCV_RESET);
+
+	/* Take clock tree out of reset */
+	cfg = readq_relaxed(xcv->reg_base + XCV_RESET);
+	cfg &= ~CLK_RESET;
+	writeq_relaxed(cfg, xcv->reg_base + XCV_RESET);
+	/* Wait for DLL to lock */
+	msleep(1);
+
+	/* Configure DLL - enable or bypass
+	 * TX no bypass, RX bypass
+	 */
+	cfg = readq_relaxed(xcv->reg_base + XCV_DLL_CTL);
+	cfg &= ~0xFF03;
+	cfg |= CLKRX_BYP;
+	writeq_relaxed(cfg, xcv->reg_base + XCV_DLL_CTL);
+
+	/* Enable compensation controller and force the
+	 * write to be visible to HW by readig back.
+	 */
+	cfg = readq_relaxed(xcv->reg_base + XCV_RESET);
+	cfg |= COMP_EN;
+	writeq_relaxed(cfg, xcv->reg_base + XCV_RESET);
+	readq_relaxed(xcv->reg_base + XCV_RESET);
+	/* Wait for compensation state machine to lock */
+	msleep(10);
+
+	/* enable the XCV block */
+	cfg = readq_relaxed(xcv->reg_base + XCV_RESET);
+	cfg |= PORT_EN;
+	writeq_relaxed(cfg, xcv->reg_base + XCV_RESET);
+
+	cfg = readq_relaxed(xcv->reg_base + XCV_RESET);
+	cfg |= CLK_RESET;
+	writeq_relaxed(cfg, xcv->reg_base + XCV_RESET);
+}
+EXPORT_SYMBOL(xcv_init_hw);
+
+void xcv_setup_link(bool link_up, int link_speed)
+{
+	u64  cfg;
+	int speed = 2;
+
+	if (!xcv) {
+		pr_err("XCV init not done, probe may have failed\n");
+		return;
+	}
+
+	if (link_speed == 100)
+		speed = 1;
+	else if (link_speed == 10)
+		speed = 0;
+
+	if (link_up) {
+		/* set operating speed */
+		cfg = readq_relaxed(xcv->reg_base + XCV_CTL);
+		cfg &= ~0x03;
+		cfg |= speed;
+		writeq_relaxed(cfg, xcv->reg_base + XCV_CTL);
+
+		/* Reset datapaths */
+		cfg = readq_relaxed(xcv->reg_base + XCV_RESET);
+		cfg |= TX_DATA_RESET | RX_DATA_RESET;
+		writeq_relaxed(cfg, xcv->reg_base + XCV_RESET);
+
+		/* Enable the packet flow */
+		cfg = readq_relaxed(xcv->reg_base + XCV_RESET);
+		cfg |= TX_PKT_RESET | RX_PKT_RESET;
+		writeq_relaxed(cfg, xcv->reg_base + XCV_RESET);
+
+		/* Return credits to RGX */
+		writeq_relaxed(0x01, xcv->reg_base + XCV_BATCH_CRD_RET);
+	} else {
+		/* Disable packet flow */
+		cfg = readq_relaxed(xcv->reg_base + XCV_RESET);
+		cfg &= ~(TX_PKT_RESET | RX_PKT_RESET);
+		writeq_relaxed(cfg, xcv->reg_base + XCV_RESET);
+		readq_relaxed(xcv->reg_base + XCV_RESET);
+	}
+}
+EXPORT_SYMBOL(xcv_setup_link);
+
+static int xcv_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
+{
+	int err;
+	struct device *dev = &pdev->dev;
+
+	xcv = devm_kzalloc(dev, sizeof(struct xcv), GFP_KERNEL);
+	if (!xcv)
+		return -ENOMEM;
+	xcv->pdev = pdev;
+
+	pci_set_drvdata(pdev, xcv);
+
+	err = pci_enable_device(pdev);
+	if (err) {
+		dev_err(dev, "Failed to enable PCI device\n");
+		goto err_kfree;
+	}
+
+	err = pci_request_regions(pdev, DRV_NAME);
+	if (err) {
+		dev_err(dev, "PCI request regions failed 0x%x\n", err);
+		goto err_disable_device;
+	}
+
+	/* MAP configuration registers */
+	xcv->reg_base = pcim_iomap(pdev, PCI_CFG_REG_BAR_NUM, 0);
+	if (!xcv->reg_base) {
+		dev_err(dev, "XCV: Cannot map CSR memory space, aborting\n");
+		err = -ENOMEM;
+		goto err_release_regions;
+	}
+
+	return 0;
+
+err_release_regions:
+	pci_release_regions(pdev);
+err_disable_device:
+	pci_disable_device(pdev);
+err_kfree:
+	devm_kfree(dev, xcv);
+	xcv = NULL;
+	return err;
+}
+
+static void xcv_remove(struct pci_dev *pdev)
+{
+	struct device *dev = &pdev->dev;
+
+	if (xcv) {
+		devm_kfree(dev, xcv);
+		xcv = NULL;
+	}
+
+	pci_release_regions(pdev);
+	pci_disable_device(pdev);
+}
+
+static struct pci_driver xcv_driver = {
+	.name = DRV_NAME,
+	.id_table = xcv_id_table,
+	.probe = xcv_probe,
+	.remove = xcv_remove,
+};
+
+static int __init xcv_init_module(void)
+{
+	pr_info("%s, ver %s\n", DRV_NAME, DRV_VERSION);
+
+	return pci_register_driver(&xcv_driver);
+}
+
+static void __exit xcv_cleanup_module(void)
+{
+	pci_unregister_driver(&xcv_driver);
+}
+
+module_init(xcv_init_module);
+module_exit(xcv_cleanup_module);